Carey: “The rules” in Morocco? This work is licensed under the Creative Commons | © Matthew Carey. Attribution-NonCommercial-NoDerivs 3.0 Unported. ISSN 2049-1115 (Online) “The rules” in Morocco?: Pragmatic approaches to flirtation and lying Matthew Carey, Collège de France This article explores the social contexts and strategies of flirtation, seduction, and relationship-building in Southern Morocco. It examines the epistemological constraints of different movements within the drama of seduction, and focuses on the ways in which people actively seek to unsettle or opacify such interactions so as to further their social ends. Uncertainty, it suggests, is not merely a social obstacle, but also a social tool. It further uses this investigation of seduction as an opportunity to explore some of the methodological shortcomings of the wider pragmatic project and adumbrate a potential remedy. Keywords: flirtation, seduction, Morocco, opacity, pragmatism In 1995, Ellen Fein and Sherrie Schneider published their best-selling self-help bible, The rules: Time-tested secrets for capturing the heart of Mr. Right. The book, which encouraged women to play cat and mouse with potential suitors so as to ensure their affections, was variously decried as outdated, radically antifeminist, and coldly utilitarian. While rejecting the first two criticisms, the authors implicitly embraced the latter, arguing that their recommendations were not making any necessary claim about the nature of men and women; they were merely highlighting what sorts of strategies worked. Their approach was, in other words, pragmatic, both in its everyday sense of focusing on what is effective, rather than what is right or true, and in its simultaneously more precise and yet somehow much broader usage within the social sciences. This latter, social scientific usage can, in truth, be hard to formulate or pin down. Where, for instance, is the conceptual red thread connecting the ideal-typology of modes of communication pioneered by Boltanski and Thévenot (1991) with William Hanks’ (1990, 2009) phenomenological investigations of deixis, and Nicolas Dodier’s (1995) ethnography of human/nonhuman relations on the factory floor? What unity there is to this eclectic array can be attributed to a common desire to break with analyses that start from all-encompassing and fixed oppositions between, say, society and the individual or domination and resistance. Within these binary pairs, one term (society, culture, structure) operates as an abstract determining force that sets an agenda for action, while the other (the individual or agent) either partially submits to this agenda or, contrariwise, creatively struggles against it and seeks to carve out a space of freedom. Pragmatism, in contrast, jettisons agency and the question of freedom as objects of analysis in favor of a renewed emphasis on the context of action and communication. It then proceeds to pick out the formal properties of particular contexts, or the strategies deployed by various actors in such contexts, and uses these as a means of mapping and interpreting interactions that straddle and encompass multiple social and conceptual scales, rather than merely oscillating between the social macro and the individual micro. So, in the case of The rules, the context of flirtation displays certain characteristics and these require actors to use strategies that cast them in a particular light by virtue of their appeal to abstract orders of value: “If she snubs my phone call, that must mean she is a modest woman who would make an ideal wife—I shall pursue her all the harder.” In this example, as for more academic forms of pragmatism, perhaps the key methodological question is how to identify or define the context to be explored. This, I suggest, is also pragmatism’s principal analytical weakness, in that the context whose definition is the product of analysis is also, all too frequently, the initial object of study.1 Take, for example, ritual, which was both the analytical starting point for this collection of themed articles, and also a particularly clear example of the tendency toward circularity. So if we look, for instance, at Højbjerg’s (2002) discussion of ritual, he relies on the identification of a certain mode of reflexivity that then serves to distinguish it from other social contexts. Humphrey and Laidlaw’s (1994) analysis of the term depends on a similar device, although this time ritual’s definiens is a matter of opaque intentionality, rather than inner reflexivity. And much the same point could be made regarding the work of more explicitly pragmatic social scientists. Boltanski’s (2009a) recent analysis of that slippery sociological object, the institution, concludes that the ambiguity he identifies as its chief characteristic can be attributed to the fact that it exists in a perpetual oscillation between instantiation and reification, and, at the same time, this oscillation constitutes the very definition of what an institution is. In each case, the context of study—which is necessary for the elaboration of this study—is determined ex post by the study itself. Jörmungandr reigns.2 The purpose of this article is not, however, to offer a critique of such analyses, whose immense heuristic value I fully recognize (it is difficult to discuss the night-sky without first breaking it up into constellations) and whose methodological subtlety I cannot do justice to here.3 Nonetheless, the article takes a different approach to pragmatic analysis by explicitly renouncing the appeal to a particular context. In part, this is in an effort to avoid, insofar as possible, presupposing my object of study, but it is also a product of the limitations imposed by the nature of this object and my own position as observer. As is frequently the case for anthropologists, my sample is not wide enough to elaborate an ideal-type of the romantic “affair” in Morocco.4 Nor can I appeal to the sort of fine-grained linguistic and contextual analysis that anthropologists normally rely on in lieu of sociological breadth. One problem with exploring the development of a flirtatious affair is that much of it must necessarily occur outside the gaze of third parties. And one problem with exploring the development of a flirtatious affair in Morocco is that it is uncommonly hard to have access to both sides of the story—at least for heterosexual encounters. Above all, I do not wish to elaborate a one-eyed and androcentric ideal-type of what is, without doubt, a highly asymmetrical relationship.5 Instead, I appeal to an older, but also notably pragmatic approach to social analysis (and one whose aim was also, at least in part, to break with totalizing concepts of society): the case study pioneered by Max Gluckman in “The bridge” (1940). Working in an intellectual environment where the presumed context and object of analysis was society as a whole, Gluckman put forward the case study as an alternative analytical frame: one that did not presuppose its object, but allowed it to emerge from the material. So rather than assuming that there was such a thing as “Zulu society” and describing its characteristics, Gluckman related the inauguration of a bridge by “Zulus” and “Whites,” concentrating on the conflicts and tensions characteristic of interactions between members of the society described by the case study—one that included both Zulus and Whites. My focus in this article differs from that of Gluckman in that I am both perfectly uninterested in describing “Moroccan society” and far from certain that such a thing exists, but its methodological purpose is very similar. What follows is not, in other words, an ethnography of The rules in Morocco, nor of what a French pragmatist might call “la forme-flirt,”6 but rather a study of a flirtatious encounter and its unfolding over time. It proceeds by exploring the formal properties of certain junctures within the social drama and examining the strategies deployed by the two parties, situating them in the wider context (or contexts) of social interaction in Morocco. Particular attention is paid to the epistemological constraints present at different points in the story, as well as to the ways in which the protagonists actively unsettle and opacify social interactions so as to further their social ends. I thereby suggest that such interactional uncertainty is not merely something to be worked around, but rather something that people work with. Hicham and Suqayna7 My plane landed in Marrakesh shortly after five o’clock and I caught a bus to the city center, alighting at the main square in the old town—the Jâmi’ Unâ’, famous for its food-stalls, musicians, traditional storytellers, and, more recently, terrorist bombings. I found a table on one of the café terraces, ordered a “half-and-half” coffee and set to calling various friends and acquaintances. First on the list was Hicham, a nurse from Beni Mellal, who now worked in the mountains around Azilal, three hours by coach from Marrakesh. To my surprise, he responded to my greetings by asking where I was in Marrakesh and then precisely which café I was sat at. Ten minutes later, he pulled up a chair next to mine. Hicham, it transpired, was in Marrakesh to meet a girl, or rather had extended his stay in the city in hopes of meeting the girl in question. Their “relationship”—if we can call it that—stretched back five months to the New Year, when the miracle of modern technology had contrived their encounter. For a week over the holiday period, Morocco’s principal mobile phone provider (Itissalat Al Maghrib) had offered free text messages to all national numbers. Few Moroccans are sufficiently moneyed to spend much time talking and so texts are extremely popular, but even they add up, and the offer was much appreciated. As well as saving them money, it also provided the bored youth of Morocco’s periurban hinterland with something to while away their evenings. Several of the young men I knew had composed stock messages stating their name, age, and desire to strike up an acquaintance with practically anybody who was interested, and then proceeded to send these messages out to random numbers. They sat around each other’s houses, drank cups of tea, and laboriously typed out countless variations on the standard ten-digit mobile number in the hope of striking it lucky. And Hicham did strike it lucky, receiving no fewer than six responses to his innumerable text messages, though two of these were from men and so pointedly ignored. Of the four remaining respondees, three swiftly ruled themselves out on one ground or another (distance, class, etc.), but the fourth, a girl called Suqayna, claimed to live in Azilal, near to where Hicham worked. This point of convergence in their otherwise quite different life trajectories allowed the two of them to drive the acquaintanceship on, exchanging further texts before swiftly upgrading to MSN (Microsoft’s instant messaging service). They even went so far as to make a few brief telephone calls. In short, they hit it off and after a couple of months of electronic communication, they arranged to meet for a cup of coffee in Marrakesh—Azilal is too small and its gossips’ tongues too lively to allow for such public encounters. Since then, Hicham had been pressing her to come away with him, perhaps to the town of Essaouira, on the sea, where he hoped the change of air and exotic surroundings would work its magic. In this endeavor he had proved less lucky, her resistance having thus far proved quite equal to his persuasion. His efforts, however, had not been entirely in vain—as there he was, in Marrakesh, waiting for Suqayna to arrive. He had been expecting her that day, but a minor contretemps had apparently forced her to delay the journey, and she was now expected at nine o’clock the next morning. Hicham and I spent the evening with some friends and arose early to trudge through the spring rain to the coach station. There, we settled ourselves at the station café and waited. Nine o’clock came and went and at half past the hour, Hicham called her. She said that she had, once again, been delayed, but was on the coach now and should arrive at half past ten. And so we waited a little longer, not unduly concerned as we had plenty of catching up to do. But half past ten also came and went and there was still no sign of Suqayna. Hicham called her…no response. Again…no response. A third time…still nothing. At this juncture he paused and then asked for my phone, explaining that she was perhaps ignoring his calls because she had lied about being on the bus, but that she might take a call from a number she didn’t recognize. This ploy also proved unsuccessful, however, and the mystery remained unresolved. She might be on the coach and unable to hear her telephone, or she might be in a mobile black spot, or she might, as Hicham had initially surmised, be ducking our attempts at communication. We waited some more and then at one o’clock we gave up and went for lunch. This setback rather poisoned the rest of the day. We wandered around the city in a fug of intermittent drizzle and desultory conversation, punctuated by Hicham’s escalating philippics against Suqayna and her presumed treachery. We reached his friend’s flat, in one of the mushrooming suburbs that encircle Marrakesh, toward nightfall, and Hicham once more embarked on his sorry tale. Upon hearing it, the friend (who, I hasten to add, had only ever seen a single blurry photo of Suqayna, taken on a mobile phone) declared that he was quite sure he had spotted her in the neighborhood that very afternoon. There could be only one explanation: she had slipped past us at the coach station and gone to meet somebody else. The friend speculated ever more wildly about the reasons for her visit to this distant suburb: it couldn’t be an accident, as her sister (with whom she was to stay) lived much more centrally; so it must be a man she had met online. Acquiescing to the implacable logic of his friend’s exposé, Hicham grimly pulled out his mobile phone and methodically deleted every single text message he had received from Suqayna, before finally expunging even her contact details. Shortly afterwards, my phone rang. It was Suqayna, whom I greeted warmly, before handing the phone over to Hicham. Suqayna had that very instant reached her sister’s flat. She had left Demnate at the time she said, but the road had been cut off by the torrential rain, and so she had been forced to spend seven or eight hours in a small town along the way, waiting until the road was passable. During this time, her phone had been at the bottom of her bag and she hadn’t heard it ring. The story was wholly improbable, but also wholly unverifiable, and Hicham was in a forgiving mood. We decided that she probably hadn’t been visiting in the neighborhood and that Hicham’s friend must have been mistaken. Probably she had just elected to stay at home in Demnate until the rain cleared and hadn’t answered her phone as it would have revealed her earlier claim to be on the coach as a lie. Or perhaps not…. We would, in any event, never know. Quite why she had phoned me instead of Hicham remained unclear, but we tacitly chose not to rake over those particular coals. We set off the next morning at around the same time, walking to the ville nouvelle where we met Suqayna in a little modern café on one of the less frequented, and less attractive side streets. We sat and chatted for half an hour until I made my excuses and left. Later that afternoon, I ran into them in the souk, where Hicham was exposing Suqayna to the exotic delights of traditional Moroccan culture. Though she had often visited Marrakesh, she had never yet ventured into this notorious repair of thieves and tourists. We retired to a nearby park, in front of the Koutoubia mosque, and discussed her CV (Curriculum Vitae), which Hicham had been correcting and I was also to look over. She was in Marrakesh for a month’s work experience in an accounts department and she wanted to make a good impression. I made a few minor alterations and then they took a taxi back to the friend’s flat. A few hours later, I received a text message from Hicham telling me that it was okay to return to the flat. I phoned to make sure there was no confusion and then did as instructed. Upon arrival, I found Hicham and Suqayna sitting around the low table in the living room eating last night’s leftover kefta (meatballs). I joined them and we nattered away like old friends in the casual intimacy of commensality. Then as dusk fell, Suqayna declared that she had to return to her sister’s house, as it wouldn’t do for her to be out after dark. Hicham and I duly walked her home—they arm-in-arm, and I loitering a good few yards behind. Throughout, the atmosphere was good and the mood easy and relaxed. After half an hour’s stroll, we parted company at the end of her sister’s street, Hicham and I smoking a cigarette while we watched her to the door. On the walk home, Hicham asked whether I had seen the date of birth written on her CV. I hadn’t. It indicated she was twenty-six years old, whereas she had repeatedly told Hicham that she was twenty-two. He proceeded to point out another couple of discrepancies between her declared and revealed identities. His tone was not one of malice, or recrimination, but a thoughtful and rather methodical reflection on the nature of the tales “we Moroccans” tell one another, and perhaps also a slightly barbed commentary on her naivety in having given herself away so easily. The next day we caught the coach for his natal village near Beni Mellal, with the intention of spending several days there visiting a local moussem (festival) before he returned to his work and I proceeded to the Atlas Mountains and my principle fieldwork site. For three days, he exchanged text messages with Suqayna, but breathed not a word of their content, the evolution of their relationship, or the events of Marrakesh. Finally, the night before I was due to leave, he mentioned that Suqayna had suggested he visit her parents’ house, adding for my benefit that this amounted to an injunction to him to propose. A few minutes silence ensued, before I enquired as to whether he was intending to go. He laughed uncomfortably and shook his head: “I don’t want to marry yet”—by which I think he principally meant that he did not want to marry HER. “So what did you say to her?” I asked. “That I have to raise the matter with my family and that it’s delicate. I’m waiting for the right moment.” The conversation petered out. The following morning, we rose before dawn and he walked me a mile or so to the nearest main road. There we sat on the verge in silence, waiting for a bus to come by. The first two were full and didn’t stop, but after an hour, one finally drew to a halt in front of us. As I loaded my rucksack into the hold, Hicham took my sleeve and said: “You meet a girl, and everything’s okay, and it gradually develops, but as soon as you sleep with her, that’s where the lying starts. And you say to yourself: ‘Why did you do it? You could have been patient…you could have suffered a while yet.’”8 I met up with him a few weeks later and, in response to my questions, he informed me that he had told Suqayna that he was leaving to work in the mountains for three months and would have no access to telephone or Internet for that period. This, as far as I am aware, marked the end of their relationship. What, then, are we to make of the various forms of interaction on display in this case study? How to separate out the different contexts in which the relationship evolves? Once more, I propose to eschew a typological analysis of the events in question, distinguishing instead between three temps, or movements, that provide the social drama with chronological structure. These three movements are those of encounter, seduction, and rupture and I explore the particular interactional strategies and constraints present in each case, as well as the types of social effect they produce. They are, of course, neither more nor less arbitrary than the contextual ideal-types in whose role they stand as vicars, but they have the advantage of not presupposing their own heuristic extensibility across a given cultural space. In other words, I am not suggesting that there is necessarily a form “flirtatious encounter,” whose rudiments or subcontexts can be abstracted from the case at hand and used to identify other encounters elsewhere in Morocco. The movements I discuss are nothing more than narrative artifacts used to address questions of cultural repertoire and situational constraint characteristic of a range of different social contexts and interactions. Encounters—or prolonging opacity Hicham and Suqayna’s encounter (i.e., the means by which they established contact), when observed from the outside, looks like the product of two remarkably bold attempts to reach out beyond their immediate social confines and establish a bridge with the unknown. Over recent years, much anthropological analysis has been poured into efforts to explore networks of people and things whose “natural” proliferation must ultimately be arrested by the limiting action of “culture” in order for new entities or ideas to emerge. Marilyn Strathern argues that transactions such as marriage (1996: 529), or claims of, say, ownership of a patent (ibid.: 524) operate by imposing limits on potentially endlessly ramifying networks. In the case of a marriage in a context of exogamic patrilines, marriage defines the precise point at which networks of blood must be seen to come to an end and as regards the other, Strathern states that “social networks…are long, patenting truncates them” (ibid.). Networks, in short, must be cut in order to be productive. This vision of human interactions and exchanges depends on our accepting that these sorts of mycelial networks are the starting point for social action, and it may be that from the sort of analytical perspective advocated by actor network, or simply network, theorists, they are. But the analytical Grundpunkt of many small-town Moroccans is just the opposite. One starts out as a more or less atomized individual, embedded in limited and largely inert kin networks, and social action is the endeavor to make that network ramify and to link one network to another. This is not to suggest that kinship is ethnographically irrelevant (as we shall see, it functions as a useful idiom of relatedness), but Hicham’s “automatic” kinship ties—those people whose claims he cannot easily avoid and vice-versa—do not run very broad, encompassing perhaps twenty individuals all told. If he is to “make it” outside the narrow confines of this world of blood, then he must follow one of two paths: the impersonal machine of postcolonial French bureaucracy with its concours and its affectations, or the hyper personalized activity of making networks proliferate and hybridize. This is most evident in efforts to establish casual points of coincidence with significant figures who straddle multiple social worlds. Hicham’s village is built just at the point where the foothills of the Central High Atlas begin to rear up out of the fertile Tadla plain. But until thirty years ago, its inhabitants lived a mile or so farther up the mountainside, only moving to the new village in the 1980s when they were offered electricity and running water to do so. Some families still live on the old site, however, and many others keep a summer house to which they retreat on summer evenings, to escape the sultry heat of the plains and seek out the dusk-wind that blows at altitude. On one of these evening walks to the old house, undertaken during a previous visit, Hicham had shown me around the ruined palace of the village’s former caïd (government strongman),9 pointing out the cachot (black-hole) in which he had flung those who flouted his authority. The caïd himself, along with his family, had long since moved to the administrative capital, Rabat, but the shadow of his former residence still hung over the village. Then, on a subsequent visit, when I was lounging around at home, Hicham suddenly emerged in the doorway, insisting that I dress myself properly as there was somebody “interesting” I had to meet. He took me to an electrical goods shop in the center of the village, and then through to the back room where we had tea and biscuits with an unveiled women of fifty or so, who spoke impeccable French (as she had been educated by nuns in Casablanca) and declared herself to be an artist—something of a sociological oddity in the immediate context. We stayed for twenty or so minutes, during which time Hicham said not a word and I bemusedly played my appointed role as art-loving Frenchman. It subsequently transpired that she was the wife of a descendant of the former caïd, who still owned significant tracts of land and properties in the village and surrounding area. When I asked Hicham why it had been so vital that I meet her, he merely replied that one never knew when such relations might come in handy. This encounter could easily be read as a classic opening gambit of a patron-client relationship, and with time it might yet reveal itself to have been one. On a number of levels, however, it fails to conform to the ideal-type of such exchanges. For one thing, no services were offered or asked for, we made no attempt to show our potential value to the dame patronesse and, above all, the connection established was most definitely not dyadic (contra Hammoudi).10 Instead, Hicham had merely caused part of his network to coincide with the network of which the caïd’s wife was a “condensation” or metonym. Not I suspect, in any great hope or expectation—he was only too aware of the limited appeal of a stammering young academic as a social lure—but because such attempts are a standard aspect of social activity and, as he rightly pointed out, one never knew. This fairly canonical example of social promiscuity is, however, merely the most obviously instrumental of a whole range of similar network-ramifying activities, many of which are nowadays conducted via modern communication technology, whose web-like properties lend themselves admirably to such endeavors. And no opportunity for establishing social contact is too trifling to be seized upon. For instance, wrong numbers are a regular feature of Moroccan existence as everybody has a mobile phone, but as I mentioned, nobody can afford to call from one. Instead, people make calls from payphones and, in the process of copying numbers from chipped screen to damaged dial, digits are often confused. Even calls from payphones are expensive, though, and they rarely last longer than a few minutes. They are only made when it is absolutely imperative to speak to somebody, or to contact loved ones and potential lovers. And yet despite this, I have never seen anybody hang up on a wrong number. They try to strike up a conversation or prolong the interaction: implicitly letting on that it was the right number, but the wrong person answered it, or simply providing so little information that the caller is unsure of what to do. What matters is to draw the exchange out. “One never knows.” Indeed, one of the high points of my social integration was when I received a ribald text message from some young man to his sweetheart and immediately responded in a way that suggested that I was not the intended recipient, but gave no clue as to my gender, intentions, or potential openness to further exchanges. A friend leaned across, asked to see what I had written and then nodded approvingly, congratulating me on my having become a local, a “son of the soil” (sâfiy, ou tamazirt at tgit). Similar games are played with people’s instant messaging accounts. I have frequently seen friends, acquaintances, or mere strangers in cybercafés assume a friend or brother’s avatar (either because they know the password or because he has just left the room for five minutes) and to use it to strike up conversations with his contacts or indeed with random other people. The game consists of then pushing the encounter as far as one can without giving up any information about oneself, playing a complex form of social battleship. And then finally, there are the endless, aimless text messages sent out for free over the New Year period and whose purpose is ostensibly sexual, but which also hold out the broader possibility of bringing together perhaps entirely extraneous networks. What, then, unites these highly different instances of communication, beyond the simple fact that they all tend toward precipitating a relationship by manufacturing an encounter? It seems to me that their principal formal property is the extraordinarily minimal amount of information exchanged by the two parties. This is least clearly the case with the meeting in the electrical goods shop, but that encounter was fairly prolonged and, even there, Hicham distinguished himself by a tenacious silence, so much of the detail elicited from our host was the result of the fact that I am a nonproficient Moroccan social actor and asked polite questions to fill the gaps (although such an approach may well, in fact, have been better suited to her particular milieu). All the other examples—those which occurred by virtue of electronically mediated forms of communication—saw the participants give away so little about themselves that they remained little more than ciphers. The whole point of the conversations struck up with wrong numbers or via assumed messaging accounts is not to reveal anything about oneself, as to do so risks terminating the exchange. Indeed, what most surprised me about Hicham’s text message sent out to random numbers was how little it said. “My name is Hicham. I am twenty-eight years old. Would you like to get to know me?”11 Nothing more. In this sense, it was rather similar to the personal ads published in Moroccan magazines, many of which are also startling in their lack of descriptors. The following example is drawn from Morocco’s best-selling national weekly French-language magazine, Tel Quel, and though perhaps a little more Spartan than most, is fairly representative of the genre: “Young Moroccan man seeks woman for friendship.” In both cases, there is no mention of profession, interests, looks, what kind of person one wants to meet, or even where one lives. Tel Quel, I remind the reader, is a nationwide magazine. Contrast these interactions with equivalent ones in Western Europe or North America, where, as Schegloff (1979) notes in his ethnomethodology of telephone opening gambits, the initial moments are invariably devoted to mutual identification,12 and as even a cursory experience of dating websites will demonstrate, you will not get many responses if you reveal scant information about yourself. As regards the formal properties of this cultivation of opacity characteristic of the opening movement, it is worth stressing that it is necessarily temporally restricted—i.e., it can only be deployed at the precise point of encounter and thus is particularly evident in forms of communication where the exchange is not seamless and analogical, but compartmentalized into discrete pockets of information (or quanta), such as text messages. In more open-ended exchanges, this initial stance is quickly subsumed by subsequent communicative strategies. In any event, its effect is fairly clear: to hold open the possibility of…whatever. The point is, as Hicham repeatedly made clear, that one cannot know what might arise from an encounter, and so one must begin by enabling it. The minimal exchange of information maximally extends the field of potential partners. But in order to develop the relationship thus precipitated, one must cultivate it, and that requires different tactics—to wit, seduction. Seduction—or the cultivation of uncertainty The second movement of this social drama, which I am calling seduction, is far less clearly definable than the first. Granted, it does have fairly unproblematic natural boundaries, beginning shortly after the initial, and necessarily brief, act of encounter and encompassing everything leading up to the act of congress—be that deliberate wooing or simply the more general attempt to establish a relationship. It wanders, however, through a whole range of different interactional contexts (the more intimate of which I had no access to whatsoever) and these cannot simply be distilled into some “essence of seduction” whose formal properties are brought to the fore by the process of concentration. Instead, I wish to focus on one particular aspect of the movement: the manner in which the actors represent themselves and their actions to the other party, and the relational implications of the epistemological stances they adopt vis-à-vis of the “truth.” This process of self-representation begins just as soon as the initial encounter establishes the possibility of a relationship. Hicham sends out a text message; Suqayna responds; but to push things further, they must exchange information about one another that allows them to create social or biographical points of convergence, or “hooks” on which to hang the relationship. So Suqayna informs Hicham of her age (twenty-two), place of residence (her parent’s home in Azilal), professional goals (secretarial work in a larger city) and so forth, and Hicham does the same. In this way, each revealed a range of aspects of their persona that gave the other party an opportunity to situate them socially and established a shared idiom that underlined their similarity. In their case, this similarity was based on a common social position as reasonably well-educated members of the numerically dominant but culturally subaltern Berber-speaking community, which acts as a buffer between the Arab-speaking plains and the more fully Berber mountains. So Hicham and Suqayna wrote to one another in French, spoke Arabic on the phone, and frequently switched to Berber (Tamazight) when they met in person, reinforcing their shared sense of identity in the process. But this identity could equally have been expressed in terms of religious practice or family origins; what matters is that it be shared. Morocco is a country in which the ideological foundation of a (horizontal) relationship is equivalence or at least similarity. For instance, if one wants to compliment or merely be nice to the anthropologist, one calls him “son of [our shared] natal village” (ou tamazirt—glossed above as “son of the soil”), and if one wants to establish closeness to a girl, one can refer to her as sister (ultama). This general process by which similarity is established is one that has received a very great deal of attention in Moroccan anthropology, beginning with Geertz’s (1979: 142) work on the concept of nisba, which he singles out as a critical and distinguishing element of Moroccan sociality. The nisba is, at heart, simply a grammatical means of transforming a noun or proper noun into another term that is simultaneously noun and adjective. Thus, the city of Marrakesh (mrâkush) yields the nisba mrâkshiy, which is both adjective and a noun: a girl from Marrakesh is a bint mrâkshiyya or simply a mrâkshiyya. The same grammatical procedure can be used to transform a profession (e.g., carpenter) or even a physical characteristic (e.g., blindness) into a noun/adjective. What makes this so important, Geertz suggests, is that rather than attributing fixed identities to people—as members of such and such a tribe or family—it allows them to negotiate their identity by selectively highlighting different facets of it. As Nicolas Puig puts it, “elle classe sans définir,”—it categorizes without defining (2004: 265). Now such an approach to the production of social relations is far from being restricted to Morocco, and Geertz’s efforts to explain it with reference to a particular grammatical construction both smack a little of the sort of culturalism that pragmatic approaches within the social sciences have been so anxious to avoid and are undermined by the widespread existence of similar grammatical forms elsewhere.13 There is, in other words, nothing so very remarkable about the fact of highlighting different aspects of one’s personality in order to establish a relationship. It is, I suspect, something that we have all engaged in at one point or another. Two things, however, distinguish such encounters in Morocco. One is the aforementioned emphasis on establishing similarity or equivalence. And the other is the nature of the utterances made by the two parties. This is the subject of a book by Lawrence Rosen (1984). He suggests that statements made in the process of creating relationship are “bargaining positions” with no direct link to the truth (1984: 118). Instead, they are to be seen in much the same way as J. L. Austin (1962) describes “performatives” in English: the question of truth or falsehood is irrelevant to them; what matters is what they do. Rosen then proceeds to link this to wider elements of “Moroccan culture,” pointing out that perjury is not a crime in the Moroccan legal system and utterances made in court have no truth-value until validated or contradicted (1984: 124). This, I would suggest, is somewhat overstating the case. Perjury may not be a crime, but the purpose of a court case is to establish the truth-value of a series of statements, and part of the game of relationship-building, as Rosen himself admits, is to gather enough information about somebody so as to be able to evaluate their utterances. Nevertheless, it is helpful for understanding at least part of the process of seduction in Hicham and Suqayna’s case. Thus we can think of the early stages of Hicham and Suqayna’s relationship, the initial steps of seduction, in much the same way as Rosen and Geertz discuss the creation of relationships and relatedness in a market environment. The process of revealing different aspects of their social trajectories mentioned above served the dual purpose of situating them socially and allowing them to establish an idiom of common identity, principally based on shared geographical and linguistic characteristics. And Rosen’s observation that the statements made during these exchanges have no necessary relationship to the truth also holds true. They each expect the other to misrepresent themselves and these acts of dissembling are not considered to be full-blown lies. (Note that despite Hicham having pointed several of Suqayna’s misrepresentations out to me—about her age and so forth—he refrained from describing them as lies and even explicitly stated that “the lying” only begins after one sleeps with a girl.) However, Rosen is I think wrong to argue that notions of truth and falsehood are irrelevant to such statements. On the contrary, Hicham and I suspect also Suqayna went to great efforts to identify the other’s misrepresentations, comparing different statements made months apart and poring over apparent aporia in the stories related. In other words, they do have a truth-value, but it is one that cannot be assumed and that only exists in potential or suspended form as something to be discovered. The nature of the misrepresentations also went beyond any straightforward effort to establish a shared idiom of identity. In the case of Hicham and Suqayna, nothing it seemed was too trifling to be misrepresented. She, and also he, dissembled regarding their whereabouts, what they were doing, where they spent l’ayd n tifeska (or Eid al-Adha, the principal Muslim festival), and pretty much anything else. The crucial thing is that any statement made be possible and quite unverifiable. When Suqayna explained that she had arrived ten hours late in Marrakesh because the road had been flooded and hadn’t answered the phone because it was at the bottom of her bag, the story was potentially true. It was raining and flash floods do sometimes tear down mountain wadis and wash away those roads. It was also highly improbable: it wasn’t really that sort of rain and she would probably have used her phone to inform her sister of her late arrival. Although, if confronted, she could always have retorted that she had told her sister she would arrive in the evening so she could freely spend the day with Hicham. Either way, Hicham would never be able to know for sure and this, I suggest, is somehow the whole point of such deceptions. They are not merely instrumental acts of misrepresentation designed to manufacture a point of convergence or similarity; they also aim to produce a generalized sense of uncertainty. Had she been in Azilal, or in Marrakesh with another man, or even stuck halfway between the two as she claimed? How old was she? Where did she spend Eid? This uncertainty can be immensely personally frustrating for the individuals involved, but also extremely socially (and perhaps also emotionally) productive. By constant acts of misrepresentation or dissembling, one maintains the other in a state of suspense where a vast range of possibilities remain open. One does not know what to think of the other person, nor how to feel toward them and so nothing, or very little, is foreclosed. Thus, contrary to Boltanski’s general philosophical claim about social interaction that in “moments of practice” (which he contrasts with “moments of reflexivity”), social actors “actively cooperate to dispel the sense of disquiet that lurks in wait for them…by closing their eyes to differences in behavior that might introduce elements of uncertainty” (2009: 165), the practice of seduction engaged in by Hicham and Suqayna relies on their acting to produce a sense of disquiet regarding the other’s identity and actions and, thereby, to generate relationally productive uncertainty. The acts of misrepresentation and dissembling that perform this task are not full-blown lies because their principal goal is not to actively mislead the other, giving them an incorrect idea about the real, but simply to mystify, obfuscate, and generalize uncertainty. What though of actual lies that aim to mislead? For these, we must turn to the third movement: rupture. Rupture and resolution—or the battle over definition In Hicham and Suqayna’s case, the process of rupture begins quite soon after the period of seduction ends. There is no particular reason why this should be so. I am not suggesting that this is a typical aspect of Moroccan “affairs” and I have witnessed others where the processes of seduction and rupture were separated by a year or more of an intercalary movement we might call “unofficial coexistence” or “established relationship.” At some point, however, the production and cultivation of uncertainty will most likely be brought to a close and the affair resolved: either by being terminated or, conversely, by being socially recognized in the form of marriage. There are exceptions to this general rule, normally when one or both parties are either already or previously have been married. Then the affair can be pursued indefinitely without receiving explicit social sanction, as when a married man, for example, carries on with a local widow (cf. Carey 2010). Such cases, though, are the exception, rather than the rule, and for Hicham and Suqayna, there was never any doubt that the situation must be resolved one way or another. What concerns us here is by what means this resolution was achieved and what distinguishes the type of speech acts made in the movements of encounter and seduction from those of rupture. For Hicham, as we saw at the end of the case study, the postcoital period is characterized by lying (“as soon as you sleep with her, that’s where the lying starts”), but this simply raises the further question of what distinguishes the falsehoods told in the process of resolving the relationship from those integral to seduction. To answer this, we need to look more closely at what the two different kinds of speech act do. So where the misrepresentations of seduction can be seen as cultivating uncertainty in order to maximize possibility, here, I suggest, the opposite must occur: rupture or other kinds of closure (such as marriage) require that uncertainty be dispelled and that the situation and each actor’s motives and behavior be somehow defined. The speech acts that occur after sexual union are, in this case, about defining the act that occurred and one’s stance toward it. This is quite clear in Suqayna’s rapid clarification of her position. She presented the act as a first step in a logical sequence that would lead ineluctably to marriage. And in so doing, she implicitly placed both the act and herself within a particular moral framework where each serves to define the other: as I am not the sort of girl who would simply sleep with a man for the fun of it, what occurred can only have been a prelude to marriage, and as it was just such a prelude, so I have invited you to visit my parents. The circle is closed and the uncertainty regarding her disposition is eliminated. There is only one thing that Hicham can be permitted to think of Suqayna—that she is a good girl. The situation is much less clear-cut when it comes to Hicham’s statements. He initially says that he will speak to his parents and then shortly afterward informs Suqayna that he has not done so and declares he will be three months incommunicado in the mountains. It is far from obvious in what way this is a clarification of anything. In other words, to return to our initial question, how do these misrepresentations differ from those offered by Suqayna on the day she failed to arrive in Marrakesh and which I have described as intended to cultivate uncertainty? The short answer is that from an external perspective there is no difference between the two sets of statements. Nor can we merely say that the difference lies in the context of enunciation (seduction versus rupture), for that would be to reproduce the tautological approach to context that the article set out to avoid. To distinguish between them, we need to appeal to the two parties’ assumptions regarding the other’s interpretation of the statements made (and I, of course, only have access to Hicham’s). So, when Suqayna likely misrepresented her day spent cut off by the floodwaters, Hicham assumed that she (the deceiving party) did not intend for him to believe the falsehood—i.e., she was using it as a smokescreen that he would recognize as such, but nonetheless be unable to penetrate. When, in contrast, he said he would talk to his parents, he (the deceiving party) hoped that she would believe him; and when he subsequently declared that he had not in fact spoken to them and was going to be uncontactable for three months, he hoped that she would recognize this as a falsehood and be aware of the actual message it purported to conceal—that he had no intention of ever discussing it with his parents and considered the relationship over. In other words, what made them lies was that they were both false and were intended to induce a particular, definite impression in the recipient, rather than being false and merely aiming to instill a vague state of uncertainty. Here, just as with Suqayna’s invitation to visit, the aim of the utterance was to make a specific claim about the relationship. The sincerity or falsehood of the statement is, from the analyst’s perspective, quite secondary to their intended task: definition of the relationship. And on this note, it is worth pointing out that we only have Hicham’s word for it that Suqayna was, in fact, sincere in wanting to marry him. The ease with which she accepted his putting an end to the relationship suggests an alternative, and equally credible explanation: that she felt obliged to define the relationship as a prelude to marriage because if she had not done so, then Hicham could have assumed she was a loose woman and dragged her good name through the mud. She, because of her vastly more precarious social position as a sexually active unmarried woman, was obliged to play on his sentiments as insurance against her reputation being destroyed. Hicham assumed she was sincere because, as he put it, “deep down, all women want is to get married,” but it is, I would argue, rather more likely that she was simply a more proficient liar than he, and this because her social existence depended upon it. Conclusion Aside from the obvious ethnographic goal of describing the contours of an amorous interaction, charting its progression, and teasing out the social techniques deployed by the two parties at different strategic junctures, the aim of this article has been twofold. On a methodological level, it has sought to retain the central tenets of pragmatic sociology, with its overriding emphasis on situational analysis, while simultaneously problematizing the idea of context that underpins so much of the pragmatic conceptual edifice. Recourse to the case study is, of course, neither bold nor innovative—it has been part of the ethnographic repertoire for seventy years—but I hope to have shown that when redeployed in pragmatic analyses, it offers a potential solution to the tautology of context-driven approaches. Instead of predetermining the social and formal boundaries of particular sorts of event, it allows actors situationally to define the contours of their interactions and leaves more room for the exploration of the ways in which interactional strategies can migrate across contextual boundaries. This methodological plasticity is especially important when analyzing everyday forms of interaction, such as the creation and cultivation of relationships, as opposed to more formalized and easily identifiable forms of actions, such as ritual. On an analytical level, I have sought not to oppose, but perhaps to modulate the claims of authors such as Boltanski or Bakhtin that social action and communication (respectively) are first and foremost a matter of papering over the cracks of existential or semantic uncertainty that always lurk underneath interactions. As mentioned above, Boltanski argues that in moments of practice, social actors “cooperate…to dispel disquiet…and [ignore] uncertainty” (op. cit.), and in his discussion of heteroglossia, Bakhtin similarly contends that “a unitary language is not something that is given, but something that must be posited…[in opposition to] the realities of heteroglossia” (1992: xix, emphasis added). In other words, by behaving as if they meant the same thing by a word, speakers of a language ensure the ongoing possibility of communication. As far as I am concerned, these two related claims are—insofar as they place the management of uncertainty and potential multiplicity at the very heart of interaction—vital to our understanding of what it means to engage in society. Nonetheless, I would suggest that while this “as if” approach to uncertainty is undoubtedly the dominant mode, there is also space for the cultivation of uncertainty in contexts where it can be socially or emotionally productive. I have described one of these. References Austin, J. L. 1962. How to do things with words. Oxford: Clarendon. Bakhtin, Mikhail. 1992. The dialogic imagination: Four essays. Austin: University of Texas Press. Boltanski, Luc. 2009a. De la critique: Précis de sociologie de l’émancipation. Paris: Gallimard. ———. 2009b. “L’inquiétude sur ce qui est: Pratique, confirmation et critique comme modalités du traitement social de l’incertitude.” Cahiers d’Anthropologie Sociale 5: 163–79. Boltanski, Luc, and Laurent Thévenot. (1987) 1991. De la justification: Les économies de la grandeur. Paris: Gallimard. Bonhomme, Julien. 2011. “Les numéros de téléphone portable qui tuent: Epidémiologie culturelle d’une rumeur transnationale.” Tracés 21: 125–50. Carey, Matthew. 2010. “Entre rencontres et rendez-vous: stratégies marocaines de sexualité hors mariage.” L’Année du Maghreb VI: 165–81. Dodier, Nicolas. 1995. Les hommes et les machines: La conscience collective dans les sociétés contemporaines. Paris: Métailié. Fein, Ellen, and Sherrie Schneider. 1995. The rules: Time-tested secrets for capturing the heart of Mr. Right. New York: Warner Books. Geertz, Clifford, Hildred Geertz, and Lawrence Rosen. 1979. Meaning and order in Moroccan society: Three essays in cultural analysis. Cambridge: Cambridge University Press. Gluckman, Max. 1940. Analysis of a social situation in modern Zululand. Manchester: Manchester University Press. Hammoudi, Abdellah. 1997. Master and disciple: The cultural foundations of Moroccan authoritarianism. Chicago: The University of Chicago Press. Hanks, William F. 1990. Referential practice, language and lived space among the Maya. Chicago: The University of Chicago Press. ———. 2009. “Comment établir un terrain d’entente dans un rituel?” Cahiers d’Anthropologie Sociale 5: 87–114. Hojbjerg, Christian K. 2002. “Inner iconoclasm: Forms of reflexivity in Loma rituals of sacrifice.” Social Anthropology 10 (1): 57–75. Humphrey, Caroline, and James Laidlaw. 1994. The archetypal actions of ritual: A theory of ritual Illustrated by the Jain rite of worship. Oxford: Oxford University Press. Puig, Nicolas. 2004. Bédouins sédentaires et société citadine à Tozeur. Paris: Karthala. Rosen, Lawrence. 1984. Bargaining for reality: The construction of social relations in a Muslim community. Chicago: The University of Chicago Press. Schegloff, Emanuel. 1979. “Identification and recognition in telephone conversation openings.” In Everyday language: Studies in ethnomethodology, edited by George Psathas, 23–78. New York: Irvington Publishers. Strathern, Marilyn. 1996. “Cutting the network.” The Journal of the Royal Anthropological Institute 2 (3): 517–35. Mensonges et séduction au Maroc? : une approche pragmatique Résumé : Cet article propose une analyse des stratégies de drague, de séduction et de construction des relations dans le sud du Maroc, ainsi que des contextes sociaux dans lesquels elles de déroulent. Il explore les contraintes épistémologiques à l’œuvre pendant les différents mouvements du drame de la séduction, en se focalisant sur les façons dont les acteurs essaient de brouiller ou d’opacifier leurs interactions afin d’atteindre certains buts sociaux. Il s’appuie sur cette analyse de la séduction pour signaler certaines lacunes méthodologiques propres aux approches pragmatiques, ainsi que pour en esquisser des réponses possibles. Matthew Carey is currently a visiting scholar at the Collège de France. His research has explored the nature of institutionality among Ichelhiyn Berbers in Southern Morocco, as well as questions of domestic violence, mental illness, parental grief, and the perception of pain. Presently, he studies modalities of lying. ___________________ 1. This problem is not, of course, restricted to pragmatism; it is merely that the especial emphasis on defining the context or ground of interaction serves to underscore a general methodological problem within the social sciences. 2. Jormungandr, for those unfamiliar with Norse mythology, is the great world serpent or “sea-thread” which encircles Midgard and eats its own tail. It is rather similar to the Greek Ouroboros. 3. Notably the incorporation of actors’ own notions of context into the resulting sociological definition, a procedure that bypasses many of the problems of analytical circularity. This approach, though, is most prevalent in sociology where the frequent overlap between the analyst’s and the analysand’s lay conceptual framework further complicates the issue. 4. I make this admission advisedly, having previously developed an ideal-typology of two forms of the premarital or extramarital affair in Morocco. 5. The gender asymmetry of such relationships is something I have discussed at length elsewhere (Carey 2010) and, for reasons of space, will not dwell on here. 6. A good deal of French pragmatism is concerned with the identification of particular sociological “forms” or templates of social interactions or events—e.g., the scandal, the affair, or the catastrophe. 7. For obvious reasons, these are pseudonyms, and I have also altered a number of other significant biographical and geographical details. 8. The verb sbr, tellingly enough, has the dual meanings of “to be patient” and “to suffer.” 9. The exact institutional position covered by the term “caïd” has changed numerous times over the last 150 years. During the period in question, it was a combination of judicial, financial, and administrative roles usually occupied by the head of a significant (and above all wealthy) local family. 10. Hammoudi (1997: 5–14) argues that patron-client relations (and indeed all same-sex relations) in Morocco are hierarchical and dyadic, and are based on the cultural “diagram” of the master-disciple relationship in Sufi orders. The interaction under discussion here was, of course, cross-sex and so does not perfectly conform to Hammoudi’s ideal-type, but my central point is that it was a nondyadic form of a potential patron-client relationship. 11. Je m’appelle Hicham. J’ai 28 ans. Voulez-vous faire ma connaissance? The fact of writing in French was intended to exclude certain categories of recipient and appeal to others. 12. See also Bonhomme (2011) for a comparative discussion of the question. 13. Though English may be bereft of an equivalent to the nisba, most Latin languages have something similar. The French terms parisien/Parisien, policier, and aveugle are all derived noun/adjective combinations that people can wear as markers of identity. They may lack a common ending (unlike the Arabic -iy), but their usage is very nearly identical.

Allard: Bureaucratic Anxiety This work is licensed under the Creative Commons | © Olivier Allard. Attribution-NonCommercial-NoDerivs 3.0 Unported. ISSN 2049-1115 (Online) Bureaucratic Anxiety: Asymmetrical interactions and the role of documents in the Orinoco Delta, Venezuela Olivier Allard, Université de Picardie-Jules Verne This article, which focuses on relations between Warao Amerindians and nonnative agents, describes documents as an opaque medium of interaction whose meaning is far from being shared, and which also create doubts as to the identity of the people involved. It suggests that such uncertainty is integral to the Warao’s involvement in asymmetrical relations with outsiders, which are inextricably political, economic, and moral. The current situation largely results from past administration by Spanish Catholic missionaries, and from the later involvement of Venezuela’s native population in national politics. Nowadays, documents are used by the Warao to switch between appeals to personal compassion and claims to administrative rights, and bureaucratic dealings are primarily articulated around the performative nature of writing. This also accounts for the mix of hope and anxiety that pervades the Warao’s interactions with nonnatives, when they use the latter’s own technology to impinge on them. Keywords: Bureaucracy, documents, missionaries, patron-client, writing, South America, Warao During certain periods of my fieldwork in a Warao village, my hosts were constantly engaged in administrative processes: almost everyday, various government agents visited the villagers and had them fill in forms so that they could take part in official programs. Local politicians arrived to campaign for the coming elections, while some Warao traveled to bigger settlements and to the regional capital in order to seek out baptismal certificates from the missionaries or to check whether their names had appeared on the lists of state pensioners. Obviously I had no illusions of finding a pristine or isolated exotic tribe when I decided to work with the Warao—a lowland native South American population mostly living in the state of Delta Amacuro (i.e., the Orinoco Delta) in Venezuela—for I was well aware that their interactions with non-Indian outsiders dated at least from Raleigh’s 1595 expedition, and that patterns of local existence had been successively affected by Spanish Catholic missionaries since the 1920s, as well as by the current Venezuelan government’s social outreach programs.1 Yet I was surprised by the extent and intensity of the bureaucratic processes I witnessed, and also by the way the Warao dealt with them—granting documents and procedures an automatic efficacy, as if they were a form of magic. During my last visit to the Delta, in autumn 2011, some of my friends showed me a list with the full names, as well as ID card numbers, of themselves and their relatives. They had composed it after hearing that some politicians had come to Guayo (the neighboring mission-village and local administrative center), telling me that it was an easy and foolproof way to obtain gifts: all one had to do was to present such a list to the visiting political leaders, get them to read and sign it, and then go and collect free food from the local store. In this instance, my friends were unsuccessful because the politicians had left early, but the efficacy of the list itself was never doubted. Of course, the local system of political clientelism helped foster such attitudes, but my friends did not believe that they were more likely to receive gifts during periods of electoral campaigning; nor did they consider that the applications they were filing for grants and funds would need to be assessed and approved, rather than being automatically and directly efficacious. Although the conceptions held by the Warao are not completely at odds with official logics of administration, their practices rely heavily on an assumption of unmediated potency: procedures and documents are not merely necessary, but are deemed sufficient, to produce the desired results. My experiences tally with evidence recorded in other settings, which have given rise to divergent interpretations: some authors analyze these events in terms of a dialectics of state control and local resistance (for a discussion, see for instance Hawkins 2002: 12; Reed 2006: 158–59), whereas others attempt to locate writing and textual objects within native ontological categories (see for instance Guzmán-Gallegos 2009: 228–30; Miller 2007: 172).2 However interesting such perspectives are, both suppress or eclipse the uncertainty that pervades bureaucratic interactions: the former because it subsumes them under a macrosocial logic, the latter because it reduces them to (new) instances of a preexisting conceptual universe. By contrast, I argue that it is possible to do justice to such uncertainty by recognizing that documents are the site of equivocal adjustments and partial compatibilities. Moreover, their opaque nature is in fact constitutive of bureaucratic interactions in the Orinoco Delta; it is doubtful whether the Warao could reach any agreement with Creoles3 and foreigners regarding what exactly they were doing in their appeals to bureaucracy. Rather, they attempt to impinge on others through recourse to a technology that does not need to have a shared meaning in order to be effective. In this paper, I try to account for the power that the Warao grant to documents. More importantly, I analyze the role documents play in asymmetrical interactions that are always negotiated and whose outcome is always uncertain. Indeed, bureaucracy can be seen as one mode of action among others, a way of influencing or coercing people that can be used as an alternative to shamanic rituals or ceremonial dialogues. Bureaucracy’s capacity to generate extended and unfamiliar social relations is a chief source of its uncertainty: like mobile phones (see Bonhomme; Carey, this volume), paperwork introduces mediated interactions between people who do not necessarily know each other. The Warao sometimes question who exactly is trying to impinge on whom, who are the authors or the recipients of documents—is it local Creole politicians, president Chávez, or the administration? Although different individuals have unequal access to administration and its techniques, the question is therefore not just one of the state controlling its population, or of the dominants coercing the dominated, because even marginalized people can play the bureaucratic game (see Hull 2008). My contribution should, however, not be seen as a case study in the “anthropology of bureaucracy.” The very lack of a “common ground” between Creole bureaucrats and their Warao clients implies that formalism is not the main trope used by both groups to justify their behavior (Herzfeld 1992: 15). While bureaucracy is a complex form of organization, defined by respect for formal procedures, and intended to achieve impersonal decision-making, only one of its features is crucial in my analysis and in the process of partial compatibility I am describing: the use of documents, and more precisely the act of writing them. There has been much research into the specificity of documents as textual artifacts, most notably the work of Goody (1977, 1986) who stresses that lists, accounts, and legal treaties dominate the documentation we have about those societies that invented written scripts. However, rather than study the quasi-universal cognitive and social effects of writing and documents, I want to adopt an approach inspired by pragmatics. Documents are often performative texts, endowed with a force that stems from their linguistic and graphic properties (Fraenkel 2008; Hull 2003), as well as from the procedures that have to be followed for their production (see Garfinkel 1967: 186–207). In other words, “writing acts” are often as important as the written documents they generate, to use a term coined by Fraenkel (2007) in a creative appropriation of Austin (1962). This is why they can generate “considerable anxiety,” for instance in situations when they engender—and not merely record—the responsibility of their authors (Hull 2003: 289). I will try to show how this Western “social magic” (Bourdieu 1982: 62) enables productive misunderstandings when it is used by (or on) Amerindian people. Writing is indeed the activity through which contemporary Venezuelan (Creole) and Warao practices are articulated, and the Warao have taken it, especially the writing of names and ID numbers, as acts that entail commitment—although maybe not for the same reasons or with the same consequences as their Creole counterparts. This is why my exploration of bureaucracy will eventually lead into a discussion of the literacy practices of the Warao, and especially the way they articulate speech and writing. Although documents share some general features and a specific logic, we can distinguish in them various “written genres” (Hull 2003: 291); I will also consider several discrete levels of interaction, since documents mediate between agents of varying size (from individuals to institutions) whose encounters are rarely face-to-face. This article thus departs from the canonical research conducted within the fields of interactionist sociology and anthropological linguistics, but is still driven by the certainty that detailed contextual analysis is the most fruitful way of understanding what happens when the limits of a given situation are thrown into question (see Bonhomme; Carey, this volume). The broader social and political context is not only the background that constrains or shapes microsociological phenomena: through the administration and writing of documents, institutions are involved as ambiguous interlocutors in interpersonal interactions whose understanding is at the core of this paper. In order to analyze the ambivalent effects the missionaries had on the Warao, I shall begin with a historical excursus that shows how they bound them through “letters of appointment” that established positions of authority, but also provided them with an administrative identity through baptismal records. The latter represented less a form of control than the very means that made possible the involvement of the Warao in national politics from the 1960s onward. A second step of my argument will be devoted to describing this phenomenon and the appearance of lists of political supporters, and of ID and electoral cards. Documents became a key to obtaining wealth and favors, but their use was still embedded in relations of personal patronage. Finally, I will try to analyze the changes, regarding both literacy practices and political relations, that result from Hugo Chávez’ Bolivarian revolution. Nowadays, the Warao use the power of documents to switch between appeals to personal compassion (drawing on a logic similar to that which would bind them to their patrons) and claims to administrative rights that are impersonal or imputed to the abstract benevolence of Chávez. It does not mean that their claims are always granted; yet they deploy writing as a newly available mode of action, which entails both assertiveness and anxiety in their interactions with outsiders. Religious administration: Learning to be bureaucratic subjects During my period of fieldwork, Adventist missionaries came to the village where I was living, sparking some debate as to whether people should convert and join them. One argument put forward by a mature man particularly struck me: Catholic missionaries are preferable to Protestants, he said, because they “write paper [documents]” (karata habata-ya). This is an example of the contemporary importance attached to documents, but also reveals the historic role still granted to missionaries, echoing the administrative work of the Capuchin Order that settled in the Orinoco Delta in 1925. The missionaries ruled the native population of the area, in lieu of the Venezuelan government, and introduced documents in their dealings with the locals in order to formalize and legitimize their new authority. What is remarkable is that they did so in a very ambivalent way. On the one hand, they turned textual artifacts into potent objects per se (more than was the case before); but on the other hand, they grounded the efficacy of documents in the pragmatic features of writing that were beyond the grasp of the Warao. It was a way of forcing the Warao to submit to the power of documents while retaining a monopoly on their production. This ambivalence is characteristic of the control exerted by the missionaries: they did not really attempt to establish a shared and stable understanding about what documents were, but rather actively maintained and capitalized on a situation of “equivocal compatibility” (see Piña-Cabral 1999: 225), at least until new generations of “civilized” Warao had been raised in their boarding schools. Let us explore this situation by way of a brief historical contextualization. Until the arrival of the Capuchin missionaries the region was a frontier for Venezuela, and its native population were only nominally controlled by the Creoles, most of whom had arrived toward the end of the nineteenth century. The central government of Venezuela was well aware of its weak grip on the peripheries, especially when trying to protect the border with English Guyana, and therefore signed a treaty in 1923 with the Capuchin Order of Castile in Spain. The treaty authorized a Mission that would benefit both sides: the Spanish missionaries would turn the Warao into Venezuelan citizens, and in return would receive state support in their civilizing and evangelizing work. The missionaries were thus granted administrative and police jurisdiction over the native population, and were able to found several missionary centers. Their appearance caused a major change in the relationships between the Warao and outsiders: the missionaries behaved as bureaucrats, much more so than any other local actor had before, which is perhaps not surprising considering that the Catholic Church itself has been regarded as essentially a bureaucratic organization, administering “grace,” in an entirely institutional way (Weber 1978: 560).4 The primary effect of documents produced by the Capuchin missionaries was to bind the Warao into hierarchies of obligation: as part of their attempt at reorganizing the sociopolitical morphology of the Delta, they granted some local leaders authority over others by giving them appointment letters. They had instituted a neat pyramidal structure: the whole area was under the general authority of a Creole who had the title of “Captain Settler” (Capitan Poblador), and was divided into three zones headed by “General Governors,” each of which was split into several centers led by “Captains” consisting of settlements ruled by “Prosecutors” (fiscales) and “Policemen,” all the latter being Warao (EE. UU. de Venezuela 1942). It is difficult to measure the extent of the change introduced by the missionaries in the organization of the Delta, and the missionaries’ own reports seem to imply that they mostly appointed people who already enjoyed local legitimacy.5 But it is extremely significant that the Capuchins’ administrative system sacralized letters of appointment. A missionary describes one instance of a “patriotic-religious feast” that they organized in their mission of Araguaimujo, in order to name the chiefs who would rule their settlements for the year. Its aim was “infiltrating in the heart of these natives the sacred love of God, the Nation [Patria] and Authority” (Muñecas 1942a: 215).6 He records his amazement at the number of Warao who asked him when it would happen and promised to come, calling it “the ceremony of the document” (fiesta de la carata). At dawn, a great number of Warao arrived in canoes from their settlements and the missionaries organized a solemn mass celebrated by the Apostolic Vicar, followed by a procession along the river. The actual nomination process then commenced. Father Rodrigo read in loud voice the new Hierarchical-Organic Order of authorities, which was to exist in all the settlements. He announced the names of those who had been distinguished with the title of Governor, Captain or Prosecutor, warning each of them of their respective duties and rights, which had to be fulfilled and respected. He then gave to each of them the letter of nomination that recognised their office and authority over others…. [Other missionaries] meanwhile, were pinning on their chest, as an emblem of power, a medal of the Liberator, tied to a sash of the national flag. It is impossible to describe the transports of joy and intimate satisfaction felt by those men, at seeing they had been given authority over the other natives and reflecting on the trust that the missionaries had in them. Even the famous Aladdin with his magic lamp, was not as happy as our Guaraunos [Warao] with their “carata” [document] and medal. (Muñecas 1942b: 265–66) In such ceremonies, which took place annually for several years at least, the missionaries went even further than other introducers of writing and documents in South America. While traders and bosses have used account ledgers to bind their debtors since the nineteenth century (see Gow 1991; Hugh-Jones 1992, 2010), the Capuchin missionaries of the Orinoco Delta did not produce mere written artifacts: they mobilized the full potency of the institution (represented here by the Apostolic Vicar, the head of the Mission) to issue letters of appointment that were administrative artifacts and, by themselves, conferred power on some over others. Indeed, since respect for “Authority” is repeatedly described in their narratives as just as important as the fear of God, this fusion of a religious and a bureaucratic ritual is not surprising. But the missionaries, just like the traders, subjected the Warao to the efficacy of documents whose full mastery they kept for themselves. If some youths had already been educated in their boarding schools, mature men still did not know how to read and write, and when they occasionally pretended that they, too, could write letters to authorities, the missionaries always laughed at such implausible efforts at appropriation (Barral 1942: 294). Through theatrical ceremonies, the missionaries promulgated an idea of documents as material objects (so much more effective upon people who do not know how to read), and one maliciously noted that the Warao leaders kept all their letters of appointment as “treasures,” long after they had expired (Turrado Moreno 1945: 59).7 Yet this is only part of the story. For in other contexts the missionaries did precisely the opposite and rooted the potency of documents in their textual content and graphic properties, which are the essential vehicles of their illocutionary force (Fraenkel 2008; Hull 2003; Riles 2006), therefore switching the focus away from the material existence of the papers.8 This alternation between two uses of documents on the part of the missionaries was thus a way of controlling the Warao: the opacity of the devices they produced and deployed was a crucial source of their power rather than an obstacle to interethnic communication. Father Barral, who worked in the Delta from 1933 until 1988, tells how a shaman called Moreno once asked to be named chief, promising he would bring all the Warao out of the forest so they could be evangelized and civilized by the missionaries—but only if he received a letter of appointment (Barral 1972: 195). Barral apparently complied, writing something on a sheet of paper that he handed to the shaman and telling him that this was his document and that he was now a chief. The Warao went back to his settlement of Siawani, announced his appointment, and started giving orders to others. But this caused a split in the village, which already had a leader. The preexisting leader eventually decided to refer the matter to the Captain Settler. Moreno was asked to produce his document, but it took some time before anyone could decipher it and read it aloud (the Captain Settler was impeded by blindness, while others just did not know): “Risum teneatis, amici! [Have a laugh, friends!] Of all the tramps of the neighbourhood, of all the crooks, layabouts and slackers, of all the thugs and thieves, I name supreme and perpetual chief the friend Moreno. Made and signed on a moon crescent. General Mao-Tse-Tung” (includes a signature with flourish). (Barral 1972: 196) The priest ends the story with his hoax letter, and does not tell what happened next, although one can imagine that Moreno was ridiculed and that the other leader resumed his authority. This was clearly a way for Barral to stress that only he and other missionaries were able to produce authoritative documents, and that not all pieces of paper bearing some writing had an effect. Naming Moreno the foremost chief of tramps and thieves in the area already turns his letter into a parody. But his parody of the pragmatic features that ensure the efficacy of bureaucratic documents is in my view much more revealing, for inscribing one’s location, name, and signing are writing acts (Fraenkel 2007: 108–10; Hull 2003: 294) required to perform the “exercitive” act (Austin 1962: 154) of appointing someone to a post. By stating that the act had been produced at a place no one had been to at the time, and by inventing a caricature of signature that he imputed to someone else (even though the Warao were likely not to know who Mao Zedung was), Barral played on the Warao’s incapacity to control what makes a document effective. He made fun of the fact that he and his brothers had managed to subject the Warao to the rule of documents, while retaining the knowledge for themselves of where their real power lay. In this—admittedly exceptional—case, as in many other instances of bureaucratic interactions, documents are used so as to deny explanation and deliberation (Graeber 2006)9: it was particularly unsettling for the Warao, who were constantly outwitted. Letters of appointment are nevertheless only one among various written genres (Hull 2003), and instituting a political order was only one of the missionaries’ aims. During their first decades in the Delta, they also performed mass baptism in the settlements they visited, with the primary aim of “saving souls,” recalling the attitude of their forerunners in the colonial era. Nowadays, baptism is alternatively called in Warao “to put water” (ho aba-kitane) and “to put a name” (wai aba-kitane), and both acts are obviously at the core of the ritual. But, as my host who mentioned the production of documents by the missionaries was well aware, the administrative aspect of baptism is as important as holy water and name giving, and missionaries themselves stress that they recorded the personal data of the newly baptized in a register (Barral 1972: 241). The missionaries have been reluctant to baptize unprepared people for several decades, but it remains an important moment in the interactions between the Warao and the outside: parents now take the lead in having their children baptized, bringing them to the missionaries or complaining that the latter have stopped visiting them. And many people still regularly come to the mission in order to ask for written certificates, although most administrative responsibilities have been transferred to state agencies. Documents such as baptismal certificates, and nowadays ID cards, have become constitutive of the mixed identity proclaimed by the Warao: they are indeed “Warao,” but not “pure Indians” (warao-witu or Sp. indio). They live in villages on river banks rather than deep in the forest; they wear clothes rather than a mere loincloth (bua, the mark of savagery); they understand at least some Spanish and therefore do not look dumbstruck when they are talked to by Creoles; they eat store-bought food and not just moriche palm starch, etcetera. This modern Indian identity is materialized in their baptismal records and ID cards: both feature the term “native” (indigena) while being essentially Christian or Venezuelan objects, and they are both therefore important instruments for constituting oneself as a Warao that is not indio (cf. Gow 1991: 59). To a large extent, this situation is recognized as a result of the missionaries’ presence, and their action was therefore not unilateral: while subjecting the Warao to their bureaucracy, they also progressively gave their pupils the means of actively taking part in the Creole world. This reflects the oft-noted ambivalence of documents and writing as both a medium of control and a source of empowerment (Goody 1986: 87–88; Gordillo 2006: 173; Guzmán-Gallegos 2009: 218–19). The missionaries founded several boarding schools where the Warao were not merely taught the love of “God, the Nation and Authority,” but were also taught to read and educated in a Creole way. Some of their former pupils became schoolmasters in the new settlements created for those who had married (religiously) after completing their studies, while others obtained official responsibilities. The missionaries’ presence in fact touched most of the population, for the systematic recording of baptisms and births produced the documentation that later permitted the Warao to obtain ID cards and passports, to vote, and to claim old-age pensions, in a word to perform bureaucratic acts. At the same time that they were subjecting the Warao to the power of documents whose conditions of production they monopolized, the missionaries were also laying the foundations for the later incorporation of their flock into bureaucratic processes—although such a process only took place after the Capuchins had lost their jurisdiction in the 1960s and 1970s. Wealth, political clientelism, and mastery The period that follows the end of the delegation of state authority to the Mission is particularly interesting because it marks the introduction of national politics to the Delta, and the distribution of salaries and gifts that accompany it.10 I argue that documents began at that time to create extended—and therefore more unpredictable—forms of interaction, whereby the Warao were linked to unknown members of bureaucratic organizations and political parties. In Venezuela in general, the crucial date is 1958, which marks the end of the dictatorship of Perez Jiménez. Under the so-called “Punto Fijo Pact,” political parties pledged to recognize the result of future democratic elections, which in practice led to the establishment of a clientelist bipartisan political system. In the 1960s, AD (Acción Democrática), and later COPEI (Cooperativa Política Electoral Independiente), started to establish party houses in each town, where voters could negotiate the terms of their political support and the redistribution of material benefits they expected (Rodriguez 2011: 60). This process relied heavily on the “lists” drawn in each party house of those supporters who would receive gifts from winning candidates (ibid.: 112–13). The Warao were at first only marginally affected by the new system, although probably more than other Venezuelan indigenous groups (cf. Alès 2007 on the Yanomami); after all, at the end of the twentieth century, most Warao inhabiting remote or small settlements did not have ID cards. The first campaigns of registration had taken place in the 1970s, but mostly targeted the larger villages, which had often been the outgrowth of missionary activity. There, people were considered to form a reserve of votes easy to control in exchange for material gifts such as an outboard motor, or a position on the local government’s payroll for well-chosen leaders. I want to underline two features of this system. On the one hand, documents (lists drawn by political parties, ID and electoral cards) started to represent a key to wealth, in addition to their older status as devices of authority, or tokens of identity. On the other hand, their use was embedded in relations of political patronage. A narrative recorded by J. L. Rodriguez shows the full extent of this ambivalence. His main informant had become a supporter of AD candidate Emery Mata Millán, and went to the party house in Tucupita after the latter was elected governor of Delta Amacuro in 1992. “There they looked up [his] name in the forms” and, since “it showed up…they nominated [him]” (Rodriguez 2011: 118). Since that time, he has been employed by the regional state to work the electricity generator in his home village, receiving a monthly wage that he explicitly justifies by his presence on the payroll book (ibid.: 120), even though there were long periods without any generator in his charge. The presence of his name on administrative lists is a direct cause of his lucrative position. Rodriguez’ informant also narrates how the documents were an objectification of his personal relation with a political patron: when Emery Mata Millán quit AD, joined COPEI, and founded another party, he always followed him. But the politician was jailed for embezzlement in 1999, and, when he tried to run again for governor after a few years in jail, the Warao leader was torn between his feeling of allegiance and the intuition that a new era had started; he eventually joined the supporters of Chávez and became a member of the PSUV (Partido Socialista Unido de Venezuela). This form of political patronage is suggestive of the important research devoted to patron-client relationships, especially in Latin America and the Mediterranean area. Yet this is another case of equivocal compatibility between Warao and Creole concepts and practices, for it evokes the very Amazonian relationship between master or owner on the one hand, and subordinate or “pet” on the other hand, which is defined by control and care in equal measure (see Fausto 2008). The Warao indeed see political patronage as an instance of a general form of asymmetrical relations between a protector (master, foster parent, leader, shaman, etc.) and a dependent being (pet, foster child, followers, auxiliary spirits, etc.). The missionaries as well, who had initially appeared as predators when they came to forcefully “collect” children for their boarding schools, forcing the Warao to flee deep into the mangrove, eventually turned into bosses (aidamo), who strictly controlled their flock, but also fed and cared for the children. Some of my older informants who had been raised at the mission even applied to some of them an expression that stresses the affective nature of fosterage—“we grew into their hands” (a-moho eku ida-e)—and such a relationship “altered” the Warao, turning them into the Venezuelan Indians that they are nowadays (Allard 2010: 117–23; cf. also Bonilla [2005: 56] about the Brazilian Paumari). In the case of political patronage, there is admittedly no such educational aspect, but the relationship is one of the main methods available to the Warao to access wealth: political gifts are seen as reciprocal help for the votes given by constituents, but also as a sign that bosses care for their followers. The Warao often appeal to the compassionate benevolence of others in order to acquire commodities and stored foodstuffs (or the money to purchase such things). They also regularly have to engage in arduous work, but these two modes are far from contradictory: things and commodities are a means of alleviating suffering or of offering pleasurable consumption without effort, and are directly linked to a central concern for others’ well-being and pain. People therefore suffer (work) out of care for their loved ones, but they also try to evoke compassion in their interlocutors, who should take their suffering into account (see e.g., Alès 2000: 135; Walker 2012a: 148). Such notions echo anthropological discourses about Amazonian ideals of generosity and sharing (Overing and Passes 2000), but I have argued that these processes are better understood in terms of “moral performances” aimed at extracting a reaction from others, with the critical notion that the uncertain outcome of such performances is not just a material product, but the revelation of what one is (an influential person) and what the other is (a moral person) (Allard 2010: 28–33; cf. also Fausto 2008: 347; Kelly 2011: 101; Strathern 1988). The Warao often try to induce compassion in others as a means of obtaining gifts from them, and this is why they so frequently stress their poverty and suffering in their discourses and behavior. My closest friend in the field, for instance, emphatically stated that he would wear rags and go without shoes if Chávez was ever to visit, so that the president would be compelled to help him after making such a visual impression. Similarly, H. Dieter Heinen has published a compilation of narratives recorded with some of his Warao informants in the 1970s and 1980s, with a whole section entitled “We, the Warao, are poor” (Oko Warao sanera), which details a visit by some outsiders who take pity on the Warao villagers and promise they will help them (Heinen 1988).11 Such logic is also at the heart of the durable relationship between bosses and their subordinates, the latter of whom constantly attempt to maximize the care they receive from the former. In contexts of political patronage, forms and lists therefore play an important and yet relatively straightforward role: they essentially materialize a connection and a due, and are fully in the hands of politicians and Creole administrators. From such a perspective they could be compared to the account ledgers used by bosses and traders. Yet they differ from the latter in that they enable the extension of interaction in space and time. Whereas written accounts are only used by local bosses who show them to their clients in order to ascertain a debt, lists of political supporters have an effect on unknown people throughout the state of Delta Amacuro. The impersonal nature of bureaucracy is thus becoming evident for the Warao, with the Warao leader discovering that the presence of his name on a list has an effect on people to whom he is otherwise unknown, namely local members of the political party to whom his patron is affiliated. His main surprise is probably that everything turns out well, but it will give way to new forms of uncertainty when it becomes unclear whom exactly one is interacting with through bureaucratic documents. Administrative rights and claims The situation has changed radically with the election of Chávez as president of Venezuela in 1998, the new constitution ratified in 1999, and more generally the “Bolivarian revolution,” as Chávez has labeled his ambitious program of social and political change—sufficiently ambitious that it is beyond the scope of this paper to describe. Alès (2007) and Mansutti and Alès (2007) have critically argued that, in spite of a very progressive constitution, the situation of indigenous peoples has not improved in Venezuela: neither their territorial rights, nor their right to be governed according to their culture and customs, have been implemented in practice. All they note is a tendency toward politicization, which creates a gap between the new leaders and those they are supposed to represent, and accentuates the subjection to Creole ways of being. Similar processes are noticeable in Delta Amacuro, but I argue that the new policies have had other—unexpected—consequences on the Warao, offering them the possibility to behave in new ways in their interactions with Creoles. Targeting the most marginal sectors of the Venezuelan population, the “Bolivarian revolution” has entailed massive registration campaigns for ID cards and electoral lists from which hardly any Warao is now excluded; this has been accompanied by a proliferation of social programs designed to redistribute the benefits of the national oil wealth, and a partial renewal of the political personnel. The practice of clientelism has obviously not been abandoned, but rather transformed: alongside direct payments or gifts on election day, and appointment of followers to positions of regional or local employee, most of the social programs are also ways of distributing benefits to constituents. For instance, educational “missions,” as they are called,12 offer grants to some students, and a small salary to teaching assistants, while the housing mission both organizes the construction of decent houses and pays local residents for the work done. The institution organizing communal day nurseries pays the women who run them and provides food for children. The “communal councils,” through which the national government funds grassroots initiatives organized along collective lines, are also sources of wealth and commodities. The list is almost endless, with projects targeting specific issues and more general programs such as the old-age pension, which theoretically benefits anyone over sixty. These changes have not caused the disappearance of forms of asymmetrical relations, because most Warao men, at least in the settlements where I lived, have bosses for whom they work or to whom they sell their produce, while local politicians still act as patrons who negotiate material benefits in exchange for electoral support. Nevertheless, Chávez has provided ways to bypass direct bosses, both practically and ideologically. Whereas allegiance was previously directed to local politicians, nowadays those politicians are only links between the population and their supreme leader: they are legitimate only insofar as they support his policies and receive his backing. Pictures of local politicians posing with a smiling president (thanks to Photoshop) are displayed all over their constituencies, at least in the state of Delta Amacuro where the opposition is nonexistent. All the social programs find their origin in Chávez’ care and generosity, as state employees are prone to remind the beneficiaries, and he becomes the ultimate master or foster father—even though he has never been to the Delta, and the Warao have to use the mediation of bureaucracy or lesser politicians if they want to reach out to him. At the same time, Chávez has insisted since the start of his first mandate that the newly distributed wealth is a right of the Venezuelan people, from which they have been deprived by previous generations of politicians and profiteers. Hence a motto of the regime: “Oil now belongs to everyone.” This discourse, widely reported in the media and relayed by local politicians and government agents, has had a major impact on the whole population and especially on the Warao. Whereas commodities can been seen as foreign to indigenous people, as revealed in the myth concerning the origin of the difference between the Creoles and the Warao,13 commodities are also now a Warao entitlement insofar they have become Venezuelan citizens, and it changes radically the way they interact with wealthier outsiders. The Warao still try to elicit caring compassion in their interlocutors, but discourses about their right to national oil wealth have also offered them an alternative possibility: to claim their due. A comparable dual strategy of inducing compassion and asserting claims has been noted among lowland South Amerindians, but in other groups the latter often takes the form of “wild” Indians dressing as warriors and speaking “hard” or “without fear” in their own language (see Alès [2007] and Kelly [2011: 164–68] on the Venezuelan Yanomami and Gordon [2006: 210] on the Brazilian Xikrin). Conversely, the assertiveness of the Warao is essentially bureaucratic. They try to compel others to release wealth, not through fear, but because they have filled in the forms and drawn lists of names and ID numbers. Indeed, compassion and bureaucracy are sometimes intertwined. During my last visit to the Delta, I accompanied Amador Medina, a young Warao employed by a local radio station, who was conducting interviews with elders who were not receiving their pension. Their speech, later broadcast in a daily program, was very straightforward and alternated between two stances: “we the Warao are helpless, we have no money…” (oko warao sanera, ka burata ekida), and, “our names have been written down / taken away, the money must come now!” (ka wai abanae/konaruae, burata ehobo-kitane ha!). This practice in fact points to the Warao’s ambivalence toward paternalistic relations, which they consider desirable as a source of protection, care, and wealth (cf. Bonilla 2005; Walker 2012a), but which they also want to be able to avoid in other instances, such as when they act as autonomous people (cf. Gow 1991: 214)—that is, as Venezuelan citizens accustomed to modernity. They alternate between the two logics because the most effective way to trigger a release of wealth is never known in advance, and also because the two logics remain intertwined: is bureaucracy a technology that enables the indirect elicitation of Chávez’ compassion? Or does it trigger the release of a due? In the discourses that Amador Medina recorded, and that I heard on other occasions, what interests me is that the Warao in question always made a direct causal link between the manipulation of documents and the claims for relief: they had a right to their share of national wealth, but it could only be triggered by means of bureaucratic procedures. From the Warao’s point of view, the outcome of their applications for pensions, grants, or subsidies does not depend anymore on the benevolence of their patrons or interlocutors (the exception being the distant figure of Chávez), which, in a way, is in line with our own conception of the ideal bureaucracy. Whereas according to the Western ideal model of bureaucracy it is the use of formal procedures that frees decisions from being tainted by personal connections and personal interests, the Warao grant this function directly to the power of documents, independent of any accompanying bureaucratic ritual or procedure. This is why they insist on having filled in forms or written lists of names when they confront interlocutors. It can be seen as a legacy of past interactions with missionaries and politicians. Although belonging to different genres, forms and lists share, along with the letters of appointment produced by the Capuchins, and with the lists of political supporters posted outside party houses, the property of having automatic effects. Yet they also create new tensions. While political theorists treat bureaucratic procedures as key to rational organization, its users often have the experience of “hateful formalism” (Herzfeld 1992: 4), of an “arbitrary, secretive, and devious” system (ibid.: 140, see also Graeber [2006] on arbitrariness). Yet this does not fit what I witnessed in the Orinoco Delta. Indeed, the Warao are often disappointed when they travel to the regional capital in order to check whether their names have appeared on the lists of payees, in the same way that the outcome of their moral performances is never a given. Documents may have an automatic effect, but the Warao are far from the sources of wealth, on the margins of the country. Failures are not seen as a consequence of indifference and arbitrariness on the part of bureaucrats, but rather as the sign that others have misappropriated their due. This issue is admittedly acknowledged by most observers, and embezzlement is common in the Delta, but here again it seems that the Warao have a different take on the matter. I argue that their feeling of bureaucratic uncertainty stems from the involvement of unknown agents in the extended interactions enabled by documents (as with mobile phones), which requires us to look more closely at the written genres and literacy practices at work. Writing anxiety The Warao’s understanding of bureaucratic efficacy in fact turns the moment of writing into a critical and uncertain step. More than discussing what happens to their forms and lists of names once such documents have reached the administrative offices of the mainland, the Warao focus their anxiety on the moment their names are written down on sheets, either by themselves or, more importantly, by others. I want to describe this process briefly, before suggesting an interpretation of why writing is the crucial step in bureaucratic processes. The Warao sometimes write their own lists of names and ID numbers, as in my initial example, but very often Creole state agents directly record the data of their target population, and the Warao always comply with reluctance. What is in doubt is obviously not the Creoles’ technical knowledge, but rather their morality, which is continually assessed—similarly to interactions with doctors among the Yanomami, as described by Kelly (2011: 128). On one occasion, some agents of the literacy “mission” (Misión Robinson) came to have everyone fill out registration forms. This was recognized by the Warao as a potential regular source of income (much more than a way of learning how to read and write) thanks to the student grants and teaching assistant wages from which my host village had so far been excluded. Yet this was an ambivalent moment, for the excitement at being potential recipients was counterbalanced by the fear of being cheated. The Creole state agents acted in a highly patronizing way, making cheeky jokes at the expense of the Warao,14 bossing them around while they were asking for their personal data, which they inscribed on sheets pompously titled “Patriotic student registration form.” Everyone complied, adopting a shy and submissive attitude, apart from the village leader who showed he knew how to interact with Creoles by answering or relaying their jokes. About a week later, a Venezuelan engineer working for the social branch of the French oil company Total, visited the village and carried out a rudimentary census of the population. He had organized the installation of solar panels about six months before (so everyone already knew him) and was now preparing a drinking water system for which he required demographic data. Yet villagers were slightly anxious about why he was recording the name of each head of household and the number of inhabitants, until the village leader confronted him angrily, demanding an explanation. He was easily convinced that it was a harmless process, but concluded by saying that it was unacceptable to come and record people’s names without presenting oneself and stating one’s purpose. Finally, the following month, it had been the turn of employees of the “communal councils” program to come and take down everyone’s name. But on that occasion I learned that my host had left angrily when asked to give his name. Some told me that he had been foolish, but I believe that he acted upon an anxiety shared by many others, which had built up over the previous months. The Creoles were performing a bureaucratic act (i.e., writing) for the Warao, but also in place of the Warao, often acting in a patronizing way or without making any effort to explain their behavior. They therefore differ markedly from literacy mediators, who have been shown to play an important role in asymmetrical relations (see Papen 2010), and rather evoke other situations of bureaucratic violence. Graeber (2006) for instance mentions how his Malagasy informants were reluctant to give him any information on the size of each family and its holdings, since such questions appeared too similar to the enquiries of government agents trying to register taxable property. However, there is no threat of violence backing such bureaucratic interactions in the Orinoco Delta, but rather a fear of not receiving the share of the wealth available to other Venezuelan citizens. As the village leader had told me, too often the Warao are taken advantage of when their names are written down and manipulated without their knowledge. Being dispossessed of action was a source of uneasiness. It is interesting to interpret such concerns by freely adopting a Strathernian idiom (Strathern 1988: 268–305). The Warao act (drawing up lists or filling out forms) in order to be the cause of the Creoles’ acts (i.e., their release of a flow of wealth). This elicitation of the latter’s acts is proof of their efficacy—they have compelled others to act with them in mind—and of their being Venezuelan citizens capable of using bureaucratic techniques in order to do so. These are obviously moments of “uncertainty and anxiety” since “there is nothing automatic about elicitation” (Strathern 1988: 299), but even more so, insofar as it involves a mode of action the Warao recognize as foreign. The anxiety surrounding writing documents can therefore be read as the fear that, by being prevented from acting themselves, the Warao would be deprived of the opportunity to cause Creoles to act with regard for them—or rather they would be confined to previous technologies of persuasion, such as direct attempts to evoke compassion. I argue that such an idiom is apposite to the situation that I am describing because, in the Warao view of bureaucracy, acting is tantamount to writing, an equivalence that I will now try to analyze. The reluctance of the Warao to have their names written down by others, or on others’ forms, contrasts with their self-confidence when they draw up their own lists, which can be freestanding or sometimes support a communal council project. Writing down names and ID numbers in fact does not only entail a referential use of language, but could also be compared to signing: “Autographic writing is supposed to accompany, produce, or be action” (Hull 2003: 294, my italics; see also Fraenkel 2007). In the Orinoco Delta, it seems to hold true when someone else inscribes one’s name: it is an illocutionary act that commits people, which is why it is such a potent—and therefore dangerous—moment. Rather than discuss how such documents are indexes of identity and personhood (e.g., Guzmán-Gallegos [2009: 228–30] on the Peruvian Runa; Herzfeld [1992: 92–93] on Greece), I want to argue here that writing is a specific form of action, at the core of bureaucracy according to the Warao. Since names and ID numbers are those that appear on Venezuelan identification cards (cédula de identidad), lists and forms can also be understood as complex graphic genres that recontextualize other written artifacts—an equivalent of reported speech (Hull 2003: 295). Words and signs indeed circulate between various written artifacts, but they only exceptionally cross the line between writing and speech. Among the Warao, there seems to be a radical disjunction between both forms, which I shall explore further in order to show that the very act of writing is much more than the material recording of speech. Contrary to practices described for the Peruvian Amazon (see Déléage 2010; Gow 2001), among the Warao there is no integration of books into shamanic rituals and visions, nor are textual artifacts used as a support that would give authority to ritual speech. On the contrary, ritual speech and ritual writing are radically opposed and incompatible practices. Warao shamanism heavily relies on nomination (Olsen 1974): to name autonomous or master-less entities is a way to subdue them as auxiliary spirits for practitioners of hoa shamanism, while a curing shaman gets control over the pathogenic entities that have penetrated a patient’s body if he utters their right name, causing the entities to vibrate and allowing them to be extracted. On other occasions, some of my informants also mentioned that it is possible to sell someone’s soul to Creole sorcerers (Sp. curioso) by giving them a piece of paper bearing someone’s written name: death will then be instantaneous. Although both practices rely on a specific use of names, my informants systematically stressed the contrast between native shamanism, practiced through speech and thought, and the magic imputed to the Creoles, which relies on reading or writing texts. When it comes to administrative texts, it is also striking that documents are not read aloud to their recipients, as is the case among the Peruvian Urarina, where it creates effects through a “displaced voice” that echoes that of the shaman (Walker 2012b). Indeed, the archetypal textual artifacts among the Warao are the ID card, the baptismal certificate, or the list of names; that is, something which is written, exhibited, glanced at, or copied, but does not offer a narrative to be read. Goody has stressed the “non-syntactic” use of language that characterizes such documents (1986: 94; see also 1977: 78). Yet instead of trying to study the broad cognitive and social effects of such technological changes, I want to focus on a specific “practice of literacy” (Scribner and Cole 1981). Hispanic names have existed as purely written objects ever since the arrival of the Capuchins. The missionaries gave names to the Warao in order to provide them with a civilized identity and to perform bureaucratic acts. Most men already had Spanish first names because they were in regular contact with local Creoles, but this was far less common among women and children, and almost none had a family name. The priests therefore chose the first name for those they baptized and gave family names when enrolling children in their boarding schools (Barral 1972: 122, 231; cf. Alès [n.d.] for the Yanomami). In other instances, family names could also be given by Creole godparents.15 Indeed, what was at stake for the missionaries was also to be able to fill out the forms properly: the first entries of the baptismal registry that I consulted in Guayo were only half completed, and the progressive civilization and evangelization of their Warao flock was revealed by the fact that soon they were able to inscribe family names for the baptized and his or her parents.16 But those names are hardly ever uttered by the Warao to address or refer to someone. It is, of course, never the case with the full sequence of first name, family name, and number (as it appears on the ID card), but even first names are hardly ever spoken, for the Warao systematically use Spanish or Warao nicknames and kin terms, which in turn are never written.17 They therefore differ from the Yanomami, who are less reluctant to use Creole than native names, because, among the Warao, official Creole names are nowadays their “real names” (wai witu). There is not a proscription on speaking out loud the real name of one’s coresident, but people usually do it with unwillingness or hesitation, and the connection between writing and speaking is therefore made in rare and significant moments. Names are occasionally read, for instance when taking a roll call, and the manager of the literacy program told me that, in remote settlements, people were amazed to discover the official name of their coresidents when he did so. On other occasions it is a serious matter, especially when names are uttered in order to be written: the only time that a Warao spontaneously told me a full name was indeed a very tense moment. Several children had died in a row, and most rumors were targeting one of the shamans of the settlement, as people whispered his nickname to each other. The eldest and most respected shaman of the village then came to see me, and he very slowly and deliberately told me the full name of the accused, which I did not previously know. He stressed that this was a name I should write down—probably in contrast to the irrelevant data I was scribbling in my notebooks—in order to denounce him to the authorities and have him put away. The exceptional character of this event is further evidence that it is not harmless to speak names. Indeed, when my hosts were reluctant to have me take down their genealogies (including their names), during the first months of my fieldwork, I believe that they were not just wary of inquisitive outsiders, but were also skeptical that the writing of their names would be of no consequence: their assumptions about the performative nature of writing underpinned their hopes but also their fears. Conclusion In describing interactions that are about the production and use of documents, or that are mediated by such documents, I have tried to show that the introduction of an administrative logic has changed the way the Warao manage asymmetrical relations with outsiders. They have learned how to act as bureaucratic agents, and use documents in their attempts to obtain wealth without being confined to a subordinate position; by writing documents, they can cause others to act with regard for them. Having to elicit compassion or to work strenuously are not their only options when they desire money and commodities, since they can now be assertive and claim their rights. However, such an attitude rests on their conception of documents as producing automatic effects, of writing lists of names and ID numbers as being a powerful act in itself. Without this partial misunderstanding, this “equivocal compatibility,” the Warao would probably not dare confront Creole administrative agents and politicians, in spite of recent appeals to enhance the political agency of indigenous groups in Venezuela. Yet this also makes any bureaucratic interaction a particularly uncertain event and accounts for the mix of expectation and anxiety that I witnessed each time outsiders visited the settlements where I lived, or when my friends traveled upriver to local or regional centers. The Warao’s hold on administrative processes is tenuous and tangential, since Creoles are always prone to misappropriate their due, and yet they are dependent on these opaque agents and procedures if they want to obtain Venezuelan oil wealth. They therefore inevitably subscribe to most of the social programs they are presented with, but always try to avoid “being written” by others whenever they can act themselves. I could have studied the same topic adopting other perspectives, for instance focusing on the ontological status of paper documents in relation to personhood, or limiting myself to large-scale historical changes of power relations with bosses, missionaries, politicians, and bureaucrats. However, only by dealing with actual events taking place in asymmetrical interactions could I do justice to the uncertain nature of the processes involved, to the fact that, when the Warao try to coerce Creoles using their own logic, they engage in a foreign “magic” that often has unforeseen consequences. Acknowledgements This article is based on research I conducted in 2007–2008 for my PhD at Cambridge University, which was funded by a Knox and a Pre-Research Linguistic studentship from Trinity College, and by a fees-only PhD studentship from the ESRC (PTA-031-2006-00305), and supervised by Stephen Hugh-Jones. Additional trips to the field in 2009 and 2011 were made possible thanks to support from the Cambridge Department of Social Anthropology and from an Ecos-Nord project directed by Catherine Alès (“Diversité culturelle, participation et gouvernance,” EHESS/GSPM). Some of the ideas I develop here have been presented at the Ersipal Seminar at the IHEAL, chaired by Capucine Boidin, at the workshop “Démocratie, participation, droits autchtones,” organized by Catherine Alès at the EHESS, and at the symposium “Amerindian modernity and political ontology,” convened by Casey High and Harry Walker at the 54th International Congress of Americanists. I wish to thank all of them, as well as others who took the time to make valuable comments—especially François Berthomé, Julien Bonhomme, Sébastien Chauvin, Grégory Delaplace, and Juan Luis Rodriguez. I am also grateful to this journal’s anonymous reviewers, and to Tom Stammers and Jacob Copeman for their help in revising the final version. References Alès, Catherine. 2000. “Anger as a marker of love: The ethic of conviviality among the Yanomami.” In The anthropology of love and anger: The aesthetics of conviviality in native Amazonia, edited by Joanna Overing and Alan Passes, 133–51. London: Routledge. ———. 2007. “‘Le goût du miel’: Le nouvel ordre politique dans l’Amazonie vénézuelienne et la participation yanomami.” Journal de la Société des Américanistes 93: 227–56. ———. n.d. “Systems of naming and creolization: Authentic acculturation or authentic tradition? The Yanomami case.” Manuscript. Allard, Olivier. 2010. “Morality and emotion in the dynamics of an Amerindian society (Warao, Orinoco Delta, Venezuela).” PhD dissertation. Cambridge: University of Cambridge. Austin, John L. 1962. How to do things with words. Oxford: Clarendon. ———. 1949. “El Delta Amacuro. Población aborigen.” In “Bodas de plata de la Misión del Caroni: Veinticinco años de labor misionera e indigenista,” special issue, Venezuela Misionera 130–31: 45–66. ———. 1972. Mi batalla de Dios. Reflejos de la vida y afanes de un misionero. Vigo: Artes Gráficas Galicia. Bonhomme, Julien. 2009. “Dieu par décret: Les écritures d’un prophète africain.” Annales HSS 4: 887–920. Bonilla, Oiara. 2005. “O bom patrâo e o inimigo voraz: Predaçâo e comercio na cosmologia paumari.” Mana 11: 41–66. Bourdieu, Pierre. 1982. “Les rites comme actes d’institution.” Actes de la recherche en sciences sociales 43: 58–63. D’Aubeterre, Luis. 2007. “‘¡Es vergonzoso, esos indios que piden en los semáforos de la ciudad!’… La construcción discursiva de la Otredad del Warao en Ciudad Guayana.” Copérnic. Revista arbitrada de divulgación científica 6: 1522. Déléage, Pierre. 2010. “Rituels du livres en Amazonie.” Cahiers des Amériques Latines 63–64: 229–50. Dumont, Jean-Paul. 1978. The headman and I: Ambiguity and ambivalence in the fieldworking experience. Austin: University of Texas Press. Fausto, Carlos. 2008. “Donos demais: Maestria e dominio na Amazonia.” Mana 14: 329–66. Fraenkel, Béatrice. 2007. “Actes d’écriture: Quand écrire c’est faire.” Langage et société 121–22: 101–12. ———. 2008. “Comment tenir un registre.” Langage et société 124: 59–71. Garfinkel, Harold. 1967. Studies in ethnomethodology. Englewood Cliffs, NJ: Prentice-Hall. Goody, Jack. 1977. The domestication of the savage mind. Cambridge: Cambridge university Press. ———. 1986. The logic of writing and the organization of society. Cambridge: Cambridge university Press. Gordillo, Gastón. 2006. “The crucible of citizenship: iD-paper fetishim in the Argentinean Chaco.” American Ethnologist 33: 162–76. Gordon, Cesar. 2006. Economia selvagem: Ritual e mercadoria entre os indios Xikrin-Mebêngôkre. Sâo Paulo: Editora UNESP/ISA/NUTI. Gow, Peter. 1991. Of mixed blood: Kinship and history in Peruvian Amazonia. Oxford: Clarendon Press. ———. 2001. An Amazonian myth and its history. Oxford: Oxford University Press. Graeber, David. 2006. “Beyond power/knowledge: An exploration of the relation of power, ignorance, and stupidity.” Malinowki Memorial Lecture. Guss, David M. 1986. “Keeping it oral: A Yekuana ethnology.” American Ethnologist 13: 413–29. Guzmán-Gallegos, María A. 2009. “Identity cards, abducted footprints, and the book of San Gonzalo: The power of textual objects in Runa worldview.” In The occult life of things: Native Amazonian theories of materiality and personhood, edited by Fernando Santos-Granero, 214–34. Tucson, AZ: University of Arizona Press. Hawkins, Sean. 2002. Writing and colonialism in northern Ghana: The encounter between the LoDagaa and “the world on paper.” Toronto: University of Toronto Press. Heinen, H. Dieter. 1988. Oko Warao: Marshland people of the Orinoco Delta. Münster: Lit Verlag. Herzfeld, Michael. 1992. The social production of indifference: Exploring the symbolic roots of Western bureaucracy. Chicago: The University of Chicago Press. Hugh-Jones, Stephen. 1988. “The gun and the bow: Myths of white men and Indians.” L’Homme 106–7: 138–55. ———. 1992. “Yesterday’s luxuries, tomorrow’s necessities: Business and barter in northwest Amazonia.” In Barter, exchange, and value: An anthropological approach, edited by Caroline Humphrey and Stephen Hugh-Jones, 42–74. Cambridge: Cambridge University Press. ———. 2010. “Entre l’image et l’écrit: La politique tukano de patrimonialisation en Amazonie.” Cahiers des Amériques Latines 63–64: 195–227. Hull, Matthew S. 2003. “The file: Agency, authority, and autography in Islamabad bureaucracy.” Language and Communication 23: 287–314. ———. 2008. “Ruled by records: The expropriation of land and the misappropriation of lists in Islamabad.” American Ethnologist 35: 501–18. Kelly, José Antonio. 2011. State healthcare and Yanomami transformations: A symmetrical ethnography. Tucson: The University of Arizona Press. Lavandero, Julio. 2004. Documentos para la historia deltamacureña: Casas-Misión de Barima y Amacuro. Caracas: UCAB/Hermanos Menores Capuchinos. Mansutti, Alexander, and Catherine Alès. 2007. “‘La géométrie du pouvoir’: Peuples indigènes et révolution au Venezuela.” Journal de la Société des Américanistes 93: 173–93. Medina, Domingo A. 2003. “From keeping it oral to writing to mapping: The Kuyujani legacy and the De’kuana self-demarcation project.” In Histories and historicities in Amazonia, edited by Neil L. Whitehead, 3–32. Lincoln, NE: University of Nebraska Press. Miller, Joana. 2007. “As coisas: Os enfeites corporais e a noçâo de pessoa entre os Mamaindê (Nambiquara).” PhD dissertation, Museu Nacional/Universidade Federal do Rio de Janeiro. Muñecas, Rodrigo de Las. 1942a. “Araguaimujo en fiestas: Fiesta de la ‘carata.’ I.” Venezuela Misionera 44: 215–18. ———. 1942b. “Araguaimujo en fiestas: Fiesta de ‘la carata.’ III.” Venezuela Misionera 46: 265–68. Olsen, Dale A. 1974. “The function of naming in the curing songs of the Warao Indians of Venezuela.” Anuario Interamericano de Investigación Musical 10: 88–122. Overing, Joanna, and Alan Passes. 2000. “Introduction: Conviviality and the opening up of Amazonian anthropology.” In The anthropology of love and anger: The aesthetics of conviviality in native Amazonia, edited by Joanna Overing and Alan Passes, 1–30. London: Routledge. Papen, Uta. 2010. “Literacy mediators, scribes or brookers? The central role of others in accomplishing reading and writing.” Langage et société 133: 63–82. Piña-Cabral, Joäo. 1999. “Trafic humain à Macao: Les compatibilités équivoques de la communication interculturelle.” Ethnologie française 29: 225–36. Reed, Adam. 2006. “Documents unfolding.” In Documents: Artifacts of modern knowledge, edited by Annelise Riles, 158–77. Ann Arbor: University of Michigan Press. Riles, Annelise. 1998. “Infinity within the brackets.” American Ethnologist 25: 378–98. ———. 2006. “Introduction.” In Documents: Artifacts of modern knowledge, edited by Annelise Riles, 1–38. Ann Arbor: University of Michigan Press. Rodriguez, Juan Luis. 2011. “Rhetorical strategies and political gift-giving in the Orinoco Delta.” PhD dissertation. Carbondale, IL: Southern Illinois University. Scribner, Sylvia, and Michael Cole. 1981. The psychology of literacy. Cambridge, MA: Harvard University Press. Strathern, Marilyn. 1988. The gender of the gift: Problems with women and problems with society in Melanesia. Berkeley: University of California Press. Taylor, Anne-Christine. 1981. “God-wealth: The Achuar and the Missions.” In Cultural transformations and ethnicity in modern Ecuador, edited by Norman E. Whitten, Jr., 647–76. Urbana: University of Illinois Press. Turrado Moreno, Angel. 1945. Etnografia de los indios guaraunos. Caracas: Lit. y Tip. Vargas/Conferencia interamericana de agricultura. Venezuela, EE. UU. de. 1942. Memoria y cuenta que el Ministro de Relaciones Interiores presenta al Congreso Nacional en sus sesiones ordinarias del año de 1942. Caracas: Imprenta Nacional. Walker, Harry. 2012a. “Demonic trade: Debt, materiality, and agency in Amazonia.” Journal of the Royal Anthropological Institute 18: 140–59. ———. 2012b. “The pen and the dart: Dark arts of Amazonian lawfare.” Oral intervention at Amerindian modernity and political ontology—54th International Congress of Americanists, edited by Casey High and Harry Walker. Vienna. Weber, Max. 1978. Economy and society. Berkeley: University of California Press. Wilbert, Johannes. 1996. Mindful of famine: Religious climatology of the Warao Indians. Cambridge, MA: Harvard University Center for the Study of World Religions. Wilde, Guillermo. 2009. Religión y poder en las misiones de Guaraníes. Buenos Aires: Editorial Sb. Anxiété bureaucratique : relations asymétriques et rôle des documents dans le Delta de l’Orénoque, Venezuela Résumé : Cet article s’intéresse au rôle que jouent les documents dans les interactions entre Indiens Warao et acteurs nonindigènes. La signification de ces documents est loin d’être partagée, et leur opacité contribue à créer des doutes quant à l’identité des personnes impliquées dans l’interaction. Une telle incertitude fait partie intégrante de l’implication des Warao dans les relations asymétriques — inextricablement politiques, économiques et morales — qui les unissent aux autres Vénézuéliens ou aux étrangers. L’état actuel de ces relations doit être éclairé par leur histoire, et résulte de l’administration de la région par des missionnaires catholiques espagnols pendant une partie du XXe siècle, puis de la participation de la population indigène du Venezuela aux processus politiques nationaux. Aujourd’hui, les documents permettent aux Warao de revendiquer des droits administratifs et non plus seulement d’en appeler à la compassion d’autrui, et la nature performative des actes d’écriture est au cœur de leurs transactions bureaucratiques. Ceci permet également de rendre compte du mélange d’espoir et d’anxiété qui traverse les interactions des Warao avec des non-indigènes, auxquels ils empruntent leur propre technique pour mieux les influencer. Olivier Allard currently teaches anthropology at the Université de Picardie-Jules Verne, and is a research associate at Centre EREA / LESC (Université Paris Ouest Nanterre / CNRS). He started working with the Warao in Venezuela in 2007, and received his PhD from the University of Cambridge in 2011. ___________________ 1. The Warao here number about 35,000 individuals. I conducted fieldwork primarily in a midsize settlement in the vicinity of San Francisco de Guayo, a large mission-village, with additional research in other areas of the Delta. 2. Amerindian appropriations of writing are common, but should be contextualized: the Ye’kuana used to reject the writing down of their oral tradition (Guss 1986), but have now adopted it in the context of land claims (Medina 2003). 3. I use the term “Creole” to translate Venezuelan Spanish criollo, meaning a normative Spanish-speaking inhabitant of Venezuela. The Warao term hotarao is usually used as an equivalent, although it can be more encompassing in other contexts (where it means “nonnative”), while I have never heard it used to refer to other Amerindian populations, as used to be the case according to Barral (1949: 47). 4. The reason why those missionaries embraced their bureaucratic function and the nationalist goals of the Venezuelan government may be specific to the Spanish Capuchins of the first half of the twentieth century. The Salesians who worked among the Ecuadorian Achuar, for instance, conversely made attempts to disconnect Christianity from technological and economic matters (Taylor 1981: 652). 5. The titles granted to leaders (aidamo) were also already used among the Warao with a phonological adaptation: kobenahoro (gobernadoi), kapita (capitan), fiskari (fiscal), komisario (comisario). This is thought to be a legacy of missionary rule in colonial times (see Turrado Moreno 1945: 56; Wilbert 1996: 75–76). 6. This ceremony is strikingly similar to those described by Wilde (2009: 80–81) among seventeenth- and eighteenth-century Guarani who belonged to the so-called “Jesuit republic.” 7. Gordillo, building on a comparable situation, argues that the Argentinean Toba fetishized documents insofar as they saw the power of ID papers “as a quality that, even though originally granted by the state, has been incorporated by the substance of the object, in which it acquires a dynamic and force of its own” (2006: 164). During the “pacification” period, the Toba were protected from military violence by certificates of good conduct written by merchants, landowners, and missionaries. In the events described here, one might argue that the question is not whether documents were fetishized by the Warao, but rather how the missionaries played on the unequal legitimacy of (their own and others’) attempts at document fetishism. 8. Riles stresses that documents are “aesthetics objects with uses distinct from their quality as ‘texts’” (1998: 378). However, to perform “verdictive” or “exercitive” acts (judging what is and what should be, cf. Austin 1962: 152–56; Bonhomme 2009: 908), both material and linguistic conditions are necessary: a statement of appointment is effective only once a signature or a seal has been affixed to it, but also requires the use of assertive verbal forms. 9. See the revised version in this issue.—Ed. 10. The prerogatives of the Capuchin missionaries were progressively reduced in the new treaties signed between Venezuela and the Order in 1956 and 1967, which were not renewed in 1972 (Lavandero 2004: 21). The substitution of the Order by civil institutions was slow and rather chaotic—whether one considers schools, civil registry, or economic development programs—and the Capuchin missionaries continued to act, as one organization among others, with a tendency to focus more on its pastoral duties. 11. This also explains why Venezuelan Creoles consider the Warao to be much more natural beggars than other Amerindian groups (cf. D’Aubeterre 2007). Ales (2007) recalls how she published a book on behalf of her Yanomami informants in which they stressed that they were not to be pitied. 12. Misión Robinson for primary education, Misión Ribas for secondary education, and Misión Sucre for university-level education. 13. Such myths are very common in lowland South America, and take the form of a choice made by the original Indian, for instance between “the gun and the bow” (Hugh-Jones 1988). Among the Warao, the native character usually refuses the overabundant commodities and replenished store that he is offered (Allard 2010: 14). 14. Which is characteristically “disrespectful” behavior, see de Vienne (this volume). 15. This practice seemed irregular among the Warao, but it can be compared to the case of the Venezuelan Enepa, whose male population systematically received first names from their creole godfathers (Dumont 1978). 16. It could be said that there is something as aesthetically unpleasant in a partially filled out form, as in an international resolution full of brackets (Riles 1998), and that the pattern of the document exerts its agency on those who manipulate it (Reed 2006), compelling them to complete all entries. Such a logic appears even more explicitly in the case of vaccination campaigns, which are a modern version of mass-baptism, since both involve a bodily manipulation meant to ensure salvation, and a written administrative record: health officials explicitly told me they give names to the children they vaccinate (although, being given to very young children, such names are usually discarded and later replaced by the Warao), because they need to fill out records. 17. Data about indigenous forms of naming about the Warao is very lacking. Assuming that Turrado Moreno (1945: 265–66) can be trusted, it seems that they were similar to contemporary nicknames: referring to a particular event or physical resemblance, they were randomly given by someone and quickly adopted by other villagers—for instance Mesi (cat), Mojoko (white worm), Akaida (long legs), etc. Nowadays, comparable names are sometimes in Spanish—for instance Sapo (toad) or Negro (black)—although they often do not have any meaning at all.

Graeber: The divine kingship of the Shilluk This work is licensed under the Creative Commons | © David Graeber. Attribution-NonCommercial-NoDerivs 3.0 Unported. ISSN 2049-1115 (Online) The divine kingship of the Shilluk On violence, utopia, and the human condition, or, elements for an archaeology of sovereignty David Graeber, Goldsmiths, University of London Since Frazer’s time, Shilluk kingship has been a flashpoint of anthropological debates about the nature of sovereignty, and while such debates are now considered irrelevant to current debates on the subject, they need not be. This essay presents a detailed analysis of the history, myth, and ritual surrounding the Shilluk institution to propose a new set of distinctions: between “divine kingship” (by which humans can become god through arbitrary violence, reflexively defining their victims as “the people”) and “sacred kingship” (the popular domestication of such figures through ritual), and argues that kingship always represents the image of a temporary, imperfect solution to what is taken to be the fundamental dilemma of the human condition—one that can itself only be maintained through terror. “God kills us.” States, I once suggested, have a peculiar dual character. They are always “at the same time forms of institutionalized raiding or extortion, and utopian projects” (Graeber 2004: 65). Obviously they are also many other things. But those two elements always remain crucial to their nature. In this essay I’d like to put some flesh on this assertion by reexamining one of the most famous cases in the history of anthropology: the divine kingship of the Shilluk of the Nilotic Sudan.1 The Shilluk kingdom might seem an odd case since it clearly is not a state by any of the usual definitions of the term—the king lacked any sort of administration and had little systematic power. Nonetheless, I suspect this is one of the reasons generations of anthropologists have found the Shilluk case so compelling. There is an intuition, here, that some of the key mechanisms of political power are best observed when stripped to their bare essentials. I would also insist that this is not because the Shilluk political system is in any sense “primitive”; not because forms of sovereignty were only beginning to emerge like some half-formed idea. To the contrary, it seems obvious that anyone living so close to ancient centers of civilization like Egypt, Meroe, or Ethiopia was likely to be perfectly aware of what a state was. Rather, it was because those elements in Shilluk society who would have liked to create something along those lines had, by the time first Ottoman and then British colonial authorities arrived, achieved such limited success at convincing the bulk of the Shilluk population to go along with them. As a result, the Shilluk kingdom was a system of institutionalized raiding, and a utopian project, and very little else. The word “utopian” might seem odd here; but one might just as easily substitute “cosmological project.” Royal palaces, royal cities, or royal courts almost invariably become microcosms, images of totality. The central place is imagined as a model of perfection, but at the same time, as a model of the universe; the kingdom, ideally, should be another reproduction of the same pattern on a larger spatial scale. I emphasize the word “ideally.” Royal palaces and royal cities always fall slightly short of Heaven; kingdoms as a whole never live up to the ideals of the royal court. This is one reason the term “utopia” seems appropriate. These are ideals that by definition can never be realized; after all, if the cosmos, and the kingdom, really could be brought into conformity with the ideal, there would be no excuse for the predatory violence. Perhaps the most fascinating aspect about the Shilluk material is that these two elements are so clearly seen as linked. Sovereignty—that which makes one a sovereign—is seen as the ability to carry out arbitrary violence with impunity. Royal subjects are equal in that they are all, equally, potential victims; but the king too is a victim in suspense, and in myth as well as ritual, it is at the moments when the people gather together to destroy the king—or at least to express their hatred for him—that he is mysteriously transformed into an eternal, transcendental being. In a cosmological system where separation is seen as balanced antagonism, opposition literally as at least potential hostility, the king inhabits a kind of tiny paradise, set apart from birth, death, and sickness; set apart from ordinary society; representing exactly this sort of imperfect ideal. Yet his ability to do so rests on a delicate balance of relations of opposition and barely contained aggression—between humans and gods, between king and people, between fractions of the royal family itself—that will, inevitably, destroy him. All this will become clearer as I go on. Let me begin, though, with a very brief survey of theories of divine kingship and the place of the Shilluk in them. Then I will demonstrate how I think these pieces can be reassembled to create the elements for a genealogy of sovereignty. Theories of divine kingship The Shilluk first became famous, in Europe and America, through James Frazer’s The golden bough. They are so firmly identified with Frazer that most are unaware the Shilluk did not even appear in the The golden bough's first two editions (1890 and 1900). Originally, in fact, Frazer drew largely on Classical literature in making an argument that all religion was to some degree derived from fertility cults centered on the figure of a dying god, and that the first kings, who embodied that god, were ritually sacrificed. This idea made an enormous impression on anthropology students of the time (and even more, perhaps, on artists and intellectuals), many of whom were to fan out across the world looking for traces of such institutions in the present day. The most successful such student was a young doctor and amateur ethnologist named Charles Seligman, who discovered in the Shilluk kingdom an almost perfect example, in 1911 sending Frazer a description that he incorporated, almost verbatim, in the book’s third edition (Seligman 1911, Frazer 1916, Fraser 1990: 200-201). One reason the Shilluk seemed to fit the bill so nicely was that Frazer argued divine kingship was originally a variety of spirit possession. To find a king whose physical health was felt to be tied to the fertility and prosperity of the kingdom, or even, that was therefore said to be ritually killed when his powers begin to wane, was not difficult. There were endless examples in Africa and elsewhere. But for Frazer, divine kings were literally possessed by a god. Frazer also felt this notion would necessarily lead to a practical problem: how does one pass this divine spirit from one mortal vessel to another? Clearly it would demand some sort of ceremony. Yet death tends to be a random and unpredictable affair. Frazer concluded the only way to carry out the ritual in a predictable way was to execute the king, either after a fixed term, or at the very least, when his weakened condition meant death seemed to be approaching anyway. The Shilluk seemed to provide a genuine example. The Shilluk king, or reth, was indeed said to embody a divine being—a god or at least a demi-god—in the person of Nyikang, the legendary founder of the Shilluk nation. Every king was Nyikang. The reth was not supposed to die a natural death. He might fall in battle with the nation’s enemies. He might be killed in single combat after a rival prince demanded a duel, as they had a right to do, or be suffocated by his own wives or retainers if he was seen to be physically failing (a state which was indeed seen to lead to poor harvests or natural catastrophes). On his death, though, Seligman emphasized, Nyikang’s spirit left him and entered a wooden effigy. Once a new reth was elected, the candidate had to raise an army and fight a mock battle against the effigy’s army in which he was first defeated and captured, then, having been possessed by the spirit of Nyikang, which passed from effigy back into his body, emerged victorious again. Frazer made the Shilluk famous and their installation ritual has become one of the classic cases in anthropology—which in a way is rather odd, since the Shilluk are on of the few Nilotic peoples never to have been the subject of sustained anthropological fieldwork. In 1948, for instance, E. E. Evans-Pritchard, taking advantage of new ethnographic material, delivered his “Frazer lecture” on the subject. The lecture was essentially designed to put the death-blow to Frazer’s whole problematic. Evans-Prichard argued that there was no such thing as a divine king, that Shilluk kings were probably never ritually executed, and that the installation ritual was not really about transferring a soul, but about resolving the tension between the office of kingship (figured as Nyikang), that was set above everyone equally, and the particular individual who held it, with his very particular background, loyalties, and local support base: In my view kingship everywhere and at all times has been in some degree a sacred office. Rex est mixta persona cum sacerdote. This is because a king symbolises a whole society and must not be identified with any part of it. He must be in the society and yet stand outside it and this is only possible if his office is raised to a mystical plane. It is the kingship, and not the king who is divine (1948: 36). The intricacies of Shilluk royal ceremonial, according to Evans-Pritchard, arose from “a contradiction between dogma and social facts” (ibid: 38). The Shilluk were a people sufficiently well-organized to wish for a symbol of national unity, in this case, the king, but not enough to allow that symbolic figure to become the head of an actual government. Evans-Pritchard was always a bit coy about his theoretical influences, but it is hard not to detect here a distant echo of the Renaissance doctrine of the “King’s Two Bodies,” that is, the “body politic,” or eternal office of kingship, ultimately including the community of his subjects, and “body natural,” which is the physical person of the individual king. This intellectual tradition was later to be the subject of comprehensive study by the German historian Ernst Kantorowicz (1957), whose student Ralph Giesey (1967), in turn, explored the way that during English and French inauguration rituals, as well, the relationship between the two bodies was acted out through royal effigies. Later anthropologists (Arens 1979, 1984; Schnepel 1988, 1995) recognized the similarity with Shilluk ritual and went on to explore the parallels (and differences) much more explicitly. Evans-Pritchard’s essay opened the way to a whole series of debates, most famously, over his claim that ritual king-killing was simply a matter of ideology, not something that ever really happened. The “did Africans really kill their kings?” debate raged for years, ending, finally, with a general recognition that at least in some cases—the Shilluk being included among them—yes, they did. At the same time, Frazer’s ideas turned out to have not been nearly as dead as expected. No one has been more responsible for the Frazerian revival than the Belgian anthropologist Luc de Heusch—who, ironically, began his intellectual journey (1962) setting out from Evans-Pritchard’s point that in order to rule, a king must “stand outside” society. Essentially he asked: what are the mechanisms through which a king is made into an outsider? In any number of African kingdoms, at least, this meant that at their installations, kings were expected to make some kind of dramatic gesture that marked a fundamental break with “the domestic order” and domestic morality. Usually this consisted of performing acts—murder, cannibalism, incest, the desecration of corpses—that would, had anyone else performed them, have been considered the most outrageous crimes. Sometimes such “exploits” were acted out symbolically: pretending to lie next to one’s sister or stepping over one’s father’s body when taking the throne. At other times they were quite literal: kings actually would marry their sisters or massacre their close kin. Always, such acts marked the king as a kind of “sacred monster,” a figure effectively outside of morality (de Heusch 1972, 1982, 2000). Marshall Sahlins (1981, 1983, 2007) has taken all this much further, pointing out, for one thing, that in the vast majority of kings, in all times and places, not only try to mark themselves as exterior to society, but actually claim to come from someplace other than the places they govern. Or at least to derive from ancestors who do. There is a sense almost everywhere that “society,” however conceived, is not self-sufficient; that power, creative energy—life, even—ultimately comes from outside. On the other hand, raw power needs to be domesticated. In myth, this often leads to stories of wild, destructive young conquerors who arrive from faraway, only to be eventually tamed on marriage to “daughters of the land.” In rituals, it often leads to ceremonies in which the king is himself conquered by the people. De Heusch’s concern was different. He was mainly interested in how, in African installation rituals, kings are effectively “torn from the everyday kinship order to take on the heavy responsibility of guaranteeing the equilibrium of the universe” (1997: 231). Kings do not begin as outsiders, they are made to “stand outside society.” But in contrast to Evans-Pritchard, he insisted this was not just a political responsibility. They stand outside society not just so they can represent it to itself, but so that they can represent it before the powers of nature. This is why, as he repeatedly emphasized, it is possible to have exactly the same rituals and beliefs surrounding actual rulers, largely powerless kings like the Shilluk reth, and “kings” who do not even pretend to rule over anything at all, but were simply individuals with an “enhanced moral status.” Here, Frazer did indeed prove useful: especially because he began to map out a typology. In “The dying god” (Part III) Frazer described how kings can act as a kind of magical charm manufactured by the people, which de Heusch calls a “fetish body,” or “a living person whose mystical capacity is closely tied to the integrity of his physical being”2 And while Frazer might not have understood that such kings were seen as being created by the people, as de Heusch held, he was quite correct in holding that, having been so consecrated, their physical strength was tied to the prosperity of nature, and that’s why they could not be allowed to grow sickly, frail, and old. In a later volume, “The scapegoat” (Part VI), Frazer discovered a second, equally important, but very different aspect of divine kingship: the king who absorbs the nation’s sin and pollution, and is thus destroyed as a way of disposing of collective evil. The two are so different they would seem difficult to reconcile. Yet in a surprising number of cases (e.g., Quigley 2005) both seem to coexist. Recently, it has been the scapegoat aspect of divine kingship that has received particular attention—largely because so many students of the institution (e.g., Makarius 1970, Scubla 2002) have been influenced by the “scapegoat theory” of French historian and literary critic Rene Girard—a theory which argues that hidden psychological scapegoat mechanisms lie at the root of all forms of myth, ritual, and ultimately, social life itself. Girard’s is one of those arguments that seems on the face of it absurd—largely because it is; it is always absurd to argue that human social life can be reduced to one single mechanism, let alone a secret one—but somehow, despite that, contains at its core something that many serious scholars cannot help but find profoundly compelling. This seems to especially happen to when the argument sets out from the proposition that, despite appearances, all human society is really founded on some kind of fundamental violence. This is Girard’s argument. Since we learn to desire by observing what others desire; we all want the same things; hence we are all in competition. The only way humans can avoid being thus plunged into a Hobbesian war of all against all is to direct their mutual hostility outwards, onto some kind of external object. And this is what we regularly do, selecting some arbitrary victim, who is first reviled as the cause of all their troubles and expelled from the community, most often, by killing him. The most surprising element in Girard’s argument is that this invariably leads to a kind of reversal: once the victim is killed, the former scapegoat is suddenly come to seen to us, not an embodiment of evil, but an exalted being, even a god, because he is now the embodiment of our ability to create human society by the very act of killing him. This mechanism he argues is the origin of all, and continues to lie at the heart of, all society and culture. The argument is, in classic Freudian style, circular: since we cannot face the reality, we are always denying it; therefore, it cannot possibly be disproved. Still, applying this model to the problem of divine kingship has interesting effects. Kings become, effectively, scapegoats in waiting (Muller 1980). Hence de Heusch’s “exploits” are, for Girardians, actual crimes. They ensure that the king is, by definition, a criminal; hence it is always legitimate to execute him, should it come to that. His sacred pneuma, then, is anticipatory: the reflected glow of the role the king might ultimately play in embodying the unity of the people in finally destroying him. Over the course of all of these debates the idea that such kings embody gods was gradually abandoned. De Heusch rejected the expression “divine kingship” entirely. Kings actually taken to be living gods are extraordinarily rare: the Egyptian Pharaoh may well have been the only entirely unambiguous example (Frankfort 1978).3 Better to speak of “sacred kingship.” These are legion. But sacred kings are not necessarily temporal rulers. They might be; but many are utterly powerless. Different functions—the king as fetish, the king as scapegoat, king as military commander or secular leader—can either be combined in the same figure or distributed across many; in any one community, any given one of them may or may not exist (de Heusch 1997). De Heusch’s ultimate conclusion is that A. M. Hocart (1927, 1933, 1936) was right: kingship was originally a ritual institution. Only later did it become something we would think of as political—that is, concerned with making decisions and enforcing them through the threat of force. As with any such statement, though, the obvious question is: what does “originally” mean here? Five thousand years ago when states first emerged in Egypt and Mesopotamia? And if so, why is that important? Or is the idea, instead, that whenever states emerge, it is invariably from within ritual institutions? This seems highly unlikely to be true in every case. Or is he simply saying that it is possible to have kings with ritual responsibilities and no political power, but not the other way around? If so it would appear to be a circular argument, since then it would only be those political figures who have ritual responsibilities whom the analyst is willing to dignify with the name of “king.” It seems to me that de Heusch’s real accomplishment is to demonstrate that what we are used to thinking of as “government” (or maybe better, “governance”) is not a unitary phenomenon. Simonse (2005: 72) for instance observes that really, all most Africans ask of their sacred kings is what most Europeans demand of theirwelfare states: health, prosperity, a certain level of life security, protection from natural disasters.4 He might have added: however, most do not feel it necessary or desirable to grant them police powers in order to accomplish this. The question of governance, then, is not the same as the question of sovereignty. But what is sovereignty? Probably the most elegant definition is that recently proposed by Thomas Hansen and Finn Stepputat (2005, 2006): in its minimal sense, sovereignty is simply the recognition of the right to exercise violence with impunity. This is probably the reason why, as these same authors note, those arguing about the nature of sovereignty in the contemporary world—and particularly about the breakdown of states, the multiplication of new forms of semi-criminal sovereignty in the margins between them—rarely find the existing anthropological literature on sacred kingship particularly useful.5 This need not be so. Actually, the existing literature contains elements from which a relevant analysis could, quite easily, be constructed. It would have to begin with the notion of transcendence: the fact that in order to become the constitutive principle of society, a sovereign has to stand outside it. True, this is slightly different from what either Evans-Pritchard or de Heusch were proposing. Both are working essentially within the Durkheimian tradition that is mainly interested in the creation of a social order, how a group can only constitute itself as a group in relation to something that effectively stands outside it. The king is simply a particular example of those “sacred” objects through which profane society constitutes itself. Starting instead from the principle of sovereignty means beginning instead from the idea of moral order, and realities of violence. It then follows from the understanding that the various “exploits” or acts of transgression by which a king marks his break with ordinary morality are not normally seen to make him immoral, but a creature beyond morality. As such he can be treated as the constituent principle of a system of justice or morality—since, logically, no creature capable of creating a system of justice can itself be already bound by the system he creates. Let me appeal to one famous example here. European visitors to the court of King Mutesa of the Ganda kingdom would occasionally try to impress him by presenting him with some new state-of-the-art rifle; he would generally respond by testing the rifle out by randomly picking off one or two of his subjects on the street. Clearly this was a calculated political gesture; the Europeans were trying to make a point of their superior firepower, Mutesa responded by demonstrating his own absolute power within his own domains. But Ganda kings were notorious for arbitrary, even random violence against their own subjects. This however did not prevent Mutesa from also being accepted as supreme judge and guardian of the state’s system of justice. Instead, such random acts of violence confirmed in him in a status similar to that often (in Africa) attributed to God, who is seen simultaneously as an utterly random force throwing lightning and striking down mortals for no apparent reason, and as the very embodiment of justice and protector of the weak. This, I would argue, is the aspect of African kingship which can legitimately be labeled “divine.” Such creatures transcend all ordinary limitations. Whether they were said to embody a god is not the issue.6 The point is that they act like gods—or even God—and get away with it. For all that European and American observers ordinarily professed horror at behavior like Mutesa’s, this divine aspect is the one that is echoed in the modern nation-state. Walter Benjamin posed the dilemma quite nicely in his famous distinction between “law-making” and “law-maintaining” violence. Really it is exactly the same paradox, cast in the new language that became necessary once the power of kings (“sovereignty”) had been transferred, at least in principle, to an entity referred to as “the people”—even though the exact way in which “the people” were to exercise sovereignty was never clear. No constitutional order can constitute itself. We like to say that “no one is above the law” but if this were really true laws would not exist to begin with: even the writers of the United State constitution or founders of the French Republic were, after all, guilty of treason according to the legal regimes under which they had been born. The legitimacy of any legal order therefore ultimately rests on illegal acts—usually, acts of illegal violence. Whether one embraces the Left solution (that “the people” periodically rise up to exercise their sovereignty through revolutions) or the Right solution (that heads of state can exercise sovereignty in their ability to set the legal order aside) the paradox itself remains. In practical terms, it translates into a constant political dilemma. How does one distinguish “the people” from a mere unruly mob? How does one know if the hand suspending habeas corpus is that of a contemporary Abraham Lincoln, or of a contemporary Mussolini? What I am proposing here is that this paradox has always been with us. Obviously, any thug or bandit who finds he can regularly get away with raping, killing, and plundering at random will not, simply by that fact, come to be seen as a power capable of constituting a moral order or national identity.7 The overwhelming majority of those who find themselves in such a situation never think to make such claims—except perhaps among their immediate henchmen. The overwhelming majority of those who do try fail. Yet the potential is always there. Successful thugs do become sovereigns, even, creators of new legal and moral systems. And genuine “sovereignty” does always carry with it the potential for arbitrary violence. This is true even in contemporary welfare states: apparently this is the one aspect that, despite liberal hopes, can never be completely reformed away. It is precisely in this that sovereigns resemble gods and that kingship can properly be called “divine.” This is not to say that Evans-Pritchard was wrong to say that kings are also always sacred. Rather, I think this perspective allows us to see that the mechanics of sacred kingship—turning the king into a fetish or a scapegoat—often operate (whatever their immediate intentions) as a means of controlling the obvious dangers of rulers who feel they can act like arbitrary, petulant gods. Sahlins’ emphasis on the way Stranger Kings must be domesticated, encompassed and thus tamed by the people is a classic case in point. It is by such means that divine kings are rendered merely sacred. In the absence of a strong state apparatus, the situation of power is often fluid and tenuous: the same act that at one point marks a monarch as a transcendent force beyond morality can, if the balance of forces shift, be reinterpreted as simple criminality. Thus can divine kings be made into scapegoats. There is every reason to believe this applies to the Shilluk king (or reth) of the eighteenth and nineteenth centuries as well. Consider the following two stories, preserved in Westermann (and bearing in mind that while there is no way to know if these incidents ever actually happened, it doesn’t really matter, since the repetition of stories constitutes the very stuff of politics): Story 1: one day a man named ogam was fishing with a member of the royal family named Nyadwai. He caught a choice fish and the prince demanded he turn it over, but he refused. Later, when his fellow villagers suggested this was unwise, he pointed out there were dozens of princes, and belittled Nyadwai: ‘who would ever elect him king?’Some years later, he learned Nyadwai had indeed been elected king. Sure enough he was summoned to court but the king’s behavior appeared to make a point of rising above the matter. “The king gave him cattle; built him a village; he married a woman, and his village became large; he had many children.”Then one day, many years later, the King destroyed the village and killed them all (Westermann 1912: 141) Here, we have an example of a king trying to play god in every sense of the term. Such a king appears arbitrary, vindictive, all-powerful in an almost Biblical sense. If one examines it in the context of Shilluk institutions, however, it begins to look rather different. ordinarily, Shilluk kings did not even have the power to appoint or remove village chiefs. In the complete absence of any sort of administrative apparatus, their power was almost entirely personal: Nyadwai created and destroyed Ogam’s village using his own personal resources, his own herd of cattle, his own personal band of retainers. If he had tried to exterminate the lineage of a real village chief, not one he had himself created, he would likely have found himself in a very serious trouble. What’s more, a reth’s power in fact was almost entirely dependent on his physical presence: Story 2: There was once a cruel king, who killed many of his subjects, “he even killed women.” His subjects were terrified of him. Then one day, to demonstrate that his subjects were so afraid they would do anything he asked, he assembled the Shilluk chiefs and ordered them to wall him up inside a house with a young girl. Then he ordered them to let him out again. They didn’t. So he died (Westermann 1912: 175).8 The story might even serve as a story of the origin of ritual regicide, though it isn’t explicitly presented as such, since this was precisely the way kings were said to have originally been put to death. They were walled in a hut with a young maiden. (The custom was discontinued, it was said, when once the maiden died first, and the king complained so loudly about the stink that they agreed from then on to smother him: Seligman 1911: 222, 1932: 91-92, Westermann 1912: 136, Hofmayr 1925: 300). Stories like these help explain a peculiar confusion in the literature on Shilluk kingship. Nineteenth century travelers, and many twentieth century observers, insisted the reth was an absolute despot wielding complete and arbitrary power over his subjects. Others—most famously Evans-Pritchard (1948)—insisted that he was for most effective purposes a mere symbolic figurehead who “reigned but did not govern,” and had almost no systematic way to impose his will on ordinary Shilluk. Both were right. As divine king, reths were expected to make displays of absolute, arbitrary violence, but the means they had at their disposal were extremely limited, and most of all, they found themselves checked and stymied whenever they tried to transform those displays into the basis for any sort of systematic power. True, as elsewhere, these displays of arbitrariness were, however paradoxically, seen as closely tied to the reth’s ability to dispense justice: nineteenth century reths could spend days on end hearing legal cases, even if, under ordinary circumstances, they were lacking in the means to enforce decisions and appear to have acted primarily as mediators. Writing in the 1940s, at a time when displays of arbitrary violence on the part of a reth would certainly have been treated as criminal by colonial police, and when the royal office had become a focus for Shilluk national identity and resistance, Evans-Pritchard had every reason to downplay such stories of brutality.F9 Nonetheless they are crucial; not only for the reasons already mentioned, but also, because under ordinarily circumstances, the arbitrary violence of the king actually seems central in constituting that sense of national identity itself. To understand this, though, we must turn to another part of Sudan during a more recent period during which the police have largely ceased to function. The Shilluk as seen from Equatttoria Here let me turn to the work of Dutch anthropologist Simon Simonse on rainmakers among a belt of peoples (the Bari, Pari, Lulubo, Lotuho, Lokoya, among others) in the furthest southern Sudan. Rainmakers are important figures throughout the area but their status varies considerably. Some have (at one time or another) managed to make themselves into powerful rulers; others remain marginal figures. All of them are liable to be held accountable in the event that (as often happens in the southern Sudan) rain does not fall. In fact, Simonse, and his colleague, Japanese anthropologist Eisei Kurimoto, are perhaps unique among anthropologists in being in the vicinity when events of this kind actually happened.10 What Simonse describes (reviewing over two dozen case studies of historically documented king-killings) is a kind of tragic drama, in which the rainmaker and people come to gradually define themselves against one another. If rains are delayed, the people (led by the chief warrior age grade) will petition the rainmaker, make gifts, rebuild his residence or put back into effect taxes or customs that have fallen into abeyance so as to win back his favor. If the rain continues not to fall, things become tense. The rainmaker is increasingly assumed to be withholding the rains, and perhaps unleashing other natural disasters, out of spite. The rainmaker will attempt stalling techniques (blaming others, sacrificial rituals, false confessions); the young men’s age set will begin to rally more and more constituencies against the king to the point where finally, the king must either flee, or confront a community entirely united against him. The methods of killing kings, Simonse notes, tend to take on the gruesome forms they do—beatings to death, burials alive-because these are ways in which everyone could be said to have been equally responsible. It is the community as a whole that must kill the king. Indeed, it only becomes a unified community—”the people” properly speaking—in doing so: since the creation and dispatching of rainmakers is about the only form of collective action in which everyone participates. All this is, perhaps, what a Girardian would predict, except that, far from being solemn sacrificial rituals with willing victims that Girard describes, king-killing more often resembled lynch-mobs, and rainmakers fought back with every means at their disposal. often in fact we hear of one lonely armed rainmaker holding off an entire incensed population. During a famine between 1855 and 1859, one Bari king who had acquired a rifle used it on three separate occasions to disperse crowds assembled to kill him. A French traveler in the 1860s was later told: We asked Nyiggilo to give us rain. He made promises and demanded cattle as a payment. Despite his spells the rain did not come. So we got angry. Then Nyiggilo took his rifle and threatened to kill everybody. We had to leave him be. Last year the same thing happened for a third time: then we lost patience. We slit Nyiggilo’s stomach open and threw him into the river: he will no longer make fun of us (in Simonse 1992: 204). It is easy to see why rainmakers might wish to acquire a monopoly on firearms, or to develop a loyal personal entourage. In fact Simonse argues that, throughout the region, when state-like forms did emerge, it was typically when rainmakers, caught in an endless and very dangerous game of bluffing and brinksmanship with their constituents, successfully sought means to reinforce their position: by intermarrying with neighboring kings, allying themselves with foreign traders, establishing trade and craft monopolies, building up a permanent armed following, and so on (2001: 94-97). In such polities “the people,” insofar as such an entity could be said to have existed, was seen essentially as the king’s collective enemy. Simonse (1992: 193–195) records several striking instances of European explorers encountering kings in the region who urged them to open fire into crowds or to carry out raids against enemy villages, only to discover that the “enemies” in question were really their own subjects. In other words, kings often really would take on the role attributed to them in rain dramas: of spitefully unleashing arbitrary destruction on the people they were supposed to protect. Simonse compares the opposition between king and people with the segmentary opposition between lineages or clans described by Evans-Pritchard among the Nuer (Simonse 1992: 27-30), each side defining itself, coming into being really, through opposition to the other. This opposition is necessarily expressed by at least the potential for violence. It might seem strange to propose a segmentary opposition between one person and everybody else, but if one returns to Evans-Pritchard’s actual analysis (1940), it makes a certain degree of sense. Evans-Pritchard stressed that in a feud, when clan or lineage A sought to avenge itself on clan or lineage B, any member of lineage B was fair game. They were treated, for political purposes, as identical. In fact, this was Evans-Pritchard’s definition of a “political” group—one whose members were treated as interchangeable in relation to outsiders.11 If so, the arbitrary violence of divine kings—firing randomly into crowds, bringing down natural disasters—is the perfect concrete expression of what makes a people a people—an undifferentiated, therefore political group. All of these peoples—Bari, Pari, Lolubo, etc—became peoples only in relation to some particularly powerful rainmaker; and owing to the rise and fall of reputations, political boundaries were always in flux. Simonse’s analysis strikes me as important. True, in the end, he does appear to fall into a Girardian framework (probably unavoidably, considering his material), seeing scapegoat dramas as the primordial truth behind all politics. So he can say that ritual king-killing of the Shilluk variety is best seen a kind of compromise, an attempt to head off the constant, unstable drama between king and people by institutionalizing the practice,12 while the state, with its monopoly on force, is an attempt to eliminate the drama entirely (Simonse 2004). Myself, I would prefer to see the kind of violence he describes not as revealing of the essential nature ofsociety, but of the essential nature of a certain form of political power with cosmic pretensions—one by no means inevitable, but which is very much still with us. Three propositions The core of my argument in this essay boils down to three propositions, and it might be best to lay them out straightaway, before returning to the Shilluk material in more detail. The first proposition I have already outlined; the second is broadly inspired by the ongoing work of Marshall Sahlins on comparative cosmologies; the third might be considered my own extrapolation from Simonse: Divine kingship, insofar as the term can be made meaningful, refers not to the identification of rulers with supernatural beings (a surprisingly rare phenomenon),13 but to kings who make themselves the equivalent of gods—arbitrary, all-powerful beings beyond human morality—through the use of arbitrary violence.The institutions of sacred kingship, whatever their origins, have typically been used to head off or control the danger of such forms of power. A direct line can be traced from such divine kingship to contemporary forms of sovereignty. Sacred kingship can also be conceived as offering a kind of (tentative, imperfect) resolution for the elementary problematic of human existence proposed in creation narratives.It is in this sense that Clastres (1977) was right when he said that state authority must have emerged from prophets rather than chiefs, from the desire to find a “land without evil” and undo death; it is in this sense, too, that it can be said that Christ (the Redeemer) was a king, or kings could so easily model themselves on Christ, despite his obvious lack of martial qualities. Here, in embryo, can we observe what I have called the utopian element of the state. Violence, and more specifically, antagonism, plays a crucial role here. It is the peculiar quality of violence that it simplifies things, draws clear lines where otherwise one might see only complex and overlapping networks of human relationship. It is the particular quality of sovereign violence that it defines its subjects as a single people.This is, in the case of kingdoms, actually prior to the friend/enemy distinction proposed by Karl Schmitt. or, to be more specific, one’s ability to constitute oneself as a single people in a potential relation of war with other peoples, is premised on a prior but usually hidden state of war between the sovereign and the people. The Shilluk kingdom then seems to be especially revealing in all three of these areas, not, as I say, because it represents some primordial form of monarchy, but because, in Shilluk rulers’ attempts to build something like a state in the absence of any real administrative apparatus, these mechanisms become unusually transparent. I also suspect the reality behind divine kingship is also particularly easy to make out here because the particular nature of Nilotic cosmology: most of all, Nilotic conceptions of God, who manifests himself in mortal life almost exclusively through disaster. One consequence is a peculiar relation between the transcendent and utopian elements, where it is the hostility of the people that makes the king a transcendent being capable of offering a kind of resolution to the dilemmas of mortal life. Be this as it may, I will spend the rest of this essay examining how these three principles—divine kingship, sacred kingship, and sovereign violence—came together in the historical Shilluk kingdom, in its stories of mythic origin and in its royal ritual, before returning to make a final brief reflection on their wider implications. A brief outline of Shilluk history The Shilluk are something of an anomaly among Nilotic people. Most Nilotes are semi-nomadic pastoralists, for whom agriculture was very much a secondary occupation, famed for their fierce egalitarianism, whose social life revolves largely around their herds. The Shilluk were not entirely different—like Nuer and Dinka, they tended to see their lives as revolving around cattle—but in practice they have, for the last several centuries at least, become far more sedentary, as they were fortunate enough to find themselves along a particularly fertile stretch of the White Nile that allowed for intensive cultivation of durra, a local grain. The result was a very dense population—by the early 19th century estimated at around two hundred thousand—living in some hundred settlements arranged so densely along the Nile that foreigners often described the 200 miles of the heart of Shilluk territory as if it consisted of one continuous village. Many remarked it appeared to be the most densely settled part of Africa outside of Egypt itself (Mercer 1971, Wall 1976). “Fortunate” though might seem an ill-chosen word here, since owing to the density of population, a bad harvest could lead to devastating famine. Lacking significant trade-goods, the Shilluk soon became notorious raiders, attacking camps and villages for hundreds of miles in all directions and hauling off cattle and grain and other spoils. By the 17th century, the 300 mile stretch of the Nile north of the Shilluk country, unsuitable for agriculture, was already known as Shilluk “raiding country,” with small fleets of Shilluk canoes preying on caravans and cattle camps. Raids were normally organized by settlement chiefs. The Shilluk reth appears to have been just one player in this predatory economy, effectively one bandit chief among many, and not even necessarily the most important, since while he received the largest share of booty, his base was in the south, closer to the pastoral Dinka rather than the richer prey to the north (Mercer 1971:416). Nonetheless, the reth acquired a great deal of cattle and used it to maintain a personal entourage of Bath Reth or “king’s men” who were his principle retainers, warriors, and henchmen. It is unclear if there even was a single figure called the “reth” in the early 17th century, or whether the royal genealogies that have come down to us really justpatched together a series of particularly prominent warriors.14 The institutions of “divine kingship” that have made the Shilluk famous appear to have been created by the reths listed as number nine and ten on most royal genealogies: Tokot (c1670-1690), famous for his conquests among the Nuba and Dinka, but most of all, by his son Tugo (c1690-1710), who lived at a time when Shilluk successes had been reversed and the heartland itself was under attack by the Dinka. Tugo is said to have been the first to create a permanent royal capital, at Fashoda,15 and to create its shrines and famous rituals of installation (ogot 1964, Mercer 1971, Wall 1976, Schnepel 1990: 114; Frost 1974). Ogot was the first to suggest that Tugo effectively invented the sacred kingship, fastening on the figure of Nyikang—probably at that time just the mythic ancestor of some local chiefly line—and transforming him into a legendary hero around which to rally a Shilluk nation that was, effectively, created by his doing so. Most contemporary historians have now come around to his position. Actually, we are only beginning to understand the full significance of what happened in a larger region context. The Shilluk kingdom was just one among many, and there appears to have been an ongoing alliance, perhaps from quite early on, between the Shilluk and the powerful Funj Sultanate of Sinnar, not far to the north (Spaulding 2007). It is possible that at least some of these ideas on which Shilluk sacral kingship were originally pioneered by the Funj, originally refugees displaced by the Shilluk themselves, but who gradually created a rich synthesis of Nubian, Christian, and Islamic cultural elements.16 Much historical work still needs to be done. Nonetheless, the precise origin of these ideas is not what’s most important. What’s important is why they were adopted. Here, the one thing that’s most clear from reading the Shilluk’s own accounts is that what happened represented a kind of gender revolution. It is important to bear in mind here that in most Nilotic societies matters of war (hence politics) are organized through male age-sets. Presumably this must have once been true here as well, but over time, their Shilluk equivalents have been comparatively marginalized (Howell 1941: 56-66).17 Instead, political life came to be organized around the reth in Fashoda, and Fashoda, in turn, became a settlement composed almost entirely of women. We do know that at the time Shilluk divine kingship took shape, the status of women, or at least that of royal woman, was a much contested political issue. Tugo’s reign was in fact preceded by that of a female reth, one Queen Abudok, Tokot’s sister.18 According to Westermann’s account (1912: 149-50), Abudok came to power and ruled for some years, but eventually Shilluk chiefs took umbrage at being ruled by a woman, and demanded she step down. She concurred, naming Tugo, then a young man in her care, as her successor. Later, according to the story, she appeared in Fashoda with a bag of lily seeds, strewing them about as she announced that henceforth, the royal lineage would grow larger and larger and scatter across the country like those seeds, until it engulfed the country entirely. Abudok’s act is usually interpreted as a spiteful prophecy, but one could just as easily read it as a story about the foundation of Fashoda itself, and the creation of the system of divine kingship usually attributed to her former ward, Tugo. Was it really Abudok who designed these institutions, perhaps when she placed Tugo on the throne to begin with? We cannot know. But certainly the common wisdom, that these institutions were purely the brainchild of Tugo himself seems implausible. It is very difficult to imagine a king who decided on his own accord to deny himself the right to name his own successor, or to grant his own wives the right to have him executed. If nothing else, we can certainly say that the system that emerged was, effectively, a kind of political compromise between male princes, royal women, and commoner chiefs—one that ensured no woman ever again attempted to take the highest office, but otherwise, granted royal women an extraordinary degree of power. Let me outline just what an important role Shilluk royal women continued to play: Where most African kings lived surrounded by a hierarchy of male officials, these were entirely absent from Fashoda. The reth lived surrounded only by his wives, who could number as many as a hundred, each with her own dwelling. No other men were allowed to set foot in the settlement after nightfall (Riad 1959: 197). Since members of the royal clan could not marry each other (this would be incest) these wives were uniformly commoners. The king’s senior wife seems to have acted as his chief minister, and had the power to hold court, and decide legal cases, in the reth’s absence (Driberg 1932: 420). She was also responsible for recruiting and supervising secondary wives. In the absence of any administrative apparatus, royal women also appear to have become the key intermediaries between Fashoda and other communities. Royal wives who became pregnant returned in their sixth month to their natal villages where their children were born and raised. They were as the saying goes “planted out” and allied themselves with a localcommoner chief (Pumphreys 1941: 11) who became the patron of the young prince or princess. Those sons who were not eventually either elected to the throne or killed in internecine strife went on to found their own branches of the royal lineage, whose numbers, as Queen Abudok predicted, tended to continually increase over the course of Shilluk history as a result. Royal daughters remained in their mothers’ villages. They were referred to as “Little Queen” and “their council sought on all matters of importance” (Driberg 1932: 420). They are not supposed to marry or have children, but in historical times at least, they became notorious for taking lovers as they wished—then, if they became pregnant, demanding hefty payments in cattle from them to hush the matter up (Howell 1953b: 107-108.)19 Princesses might also be appointed as governors over local districts (Hofmayr 1925: 71; Jackson in Frost 1974: 133-134), particularly if their brothers became king. Royal wives who had borne three children, and royal widows, would retire to their natal villages to become bareth, or guardians of royal shrines (Seligman 1932: 77-78). It was through these shrines that the “cult of Nyikang” was disseminated.20 These women of course also became key political conduits between commoner chiefs and the royal court. While as noted above it was considered quite outrageous for a king to kill a woman, royal wives were expected to ultimately order to the death of the king. A reth as said to be put to death when his physical powers began to fade—purportedly, when his wives announced that he is no longer capable of satisfying them sexually (Seligman 1911: 222; Howell and Thomson 1946: 10). In some accounts (e.g., Westermann 1912: 136) the execution is carried out by the royal wives themselves.21 one may argue about the degree to which this whole scenario is simply an ideological facade, but it clearly happened sometimes: Hofmayr for instance writes of one king’s affection for his mother, “who had killed his father with a blow from a brass-ring” (1925: 127, in Frost 1974: 82). I should emphasize that Shilluk society as a whole was in no sense a matriarchy. Women held extraordinary power within the royal apparatus, but that apparatus was not in itself particularly powerful. The fact that the Queen could render judicial judgments, for instance, is less impressive when one knows royal judgmentswere not usually enforced. Governance of day-to-day affairs seems to have rested firmly in the hands of commoner male settlement chiefs, who were also in charge of electing a new king when the old one died. Village women also elected female chiefs who had jurisdiction over women’s affairs but these were much less important.22 Property was passed in the male line. The reth himself continued to exercise predatory and sometimes brutal power through his personal retainers, occasionally raiding his own people as a mode of intervening in local politics. Nonetheless, that (divine, arbitrary) power seems to have been increasingly contained within a ritual apparatus where royal women played the central political role. Insofar as royal power became more than a sporadic phenomenon; insofar as it came to embed itself in everyday life, it was, apparently, largely through the agency of the bareth and their network of royal shrines, spread throughout Shillukland. Here, though, the effects could hardly be overestimated. The figure of Nyikang, the mythic founder of the nation, came to dominate every aspect of ritual life, and to become the very ground of Shilluk social being. Where other Nilotic societies are famous for their theological speculation, with sacrifice—the primary ritual—always being directed to God and attendant cosmic spirits, here, everything came to be centered on the “cult of Nyikang.” This was true to such a degree that by the time Seligman was writing (1911, 1932), he found it difficult to establish what Shilluk ideas about God or lineage ancestors even were. To give some sense of the royal spirits’ pervasiveness: while Nuer and Dinka who fell ill typically attributed their condition to attack by “air spirits,” and sought cures from mediums possessed by such spirits, most Shilluk appear to have assumed they were being attacked by former kings—most often, Nyikang’s aggressive son Dak—and sought the aid of mediums possessed by Nyikang himself (Seligman and Seligman 1932: 101-102). While most ordinary Shilluk, as we shall see, assiduously avoided the affairs of living royalty, dead ones soon came to intervene in almost every aspect of their daily lives. The obvious question is how long it took for this to happen. Here, information is simply unavailable. All we know is that the figure of Nyikang gradually came to dominate every aspect of Shilluk life. The political situation in turn appears to have stabilized by 1700 and remained so for at least a century. By the 1820s however the Ottoman state began attempting to establish its authority in the region, and this coincided with a sharp increase in the demand for ivory on the world market. Arab merchants and political refugees began to establish themselves in the north of the country. Nyidok (1845-1863) refused to receive official Ottoman envoys, but he kept up the Shilluk tradition of guaranteeing the safety of foreigners. Before long there were thousands of the latter, living in a cluster of communities around Kaka in the far north. Reths responded by creating new trade monopolies, imposing systematic taxes, and trying to create a royal monopoly on firearms.23 They do not appear to have been entirely unsuccessful. Foreign visitors at the time certainlycame away under the impression they had been dealing with a bona fide monarch, with at least an embryonic administration. At the same time, some also reported northerners openly complaining it would be better to live without a reth entirely (Mercer 1974: 423-24). The situation ended catastrophically. As the ivory trade was replaced by the slave trade, northern Shilluk increasingly signed up as auxiliaries in Arab raids on the Dinka; by 1861, a foreign freebooter named Mohammed Kheir thus managed to sparked a civil war that allowed them to sack Fashoda and carry out devastating slave raids against the Shilluk heartland itself (Udall 1998: 474-82; Kapteijns and Spaulding 1982: 43-46). The sack of Fashoda was followed by some forty years of almost continual warfare. The north battled the south; foreign powers (first the Ottoman regime, then the Mahdist regime in Khartoum, then finally the British) intervened trying to establish client governments; several reths were executed as rebels against one side or the other; Shilluk herds were decimated and the carnage was such that the population fell by almost half. In 1899 British rule was established, Shilluk territory restricted and those outside it resettled, and the reth reduced to the usual tax-collector and administer of local justice under a system of indirect colonial rule. At the same time, the royal installation ritual, which had fallen into abeyance during the civil wars, was revived and probably reinvented, and royal institutions, along with the figure of Nyikang, became if anything even more important as symbols of national identity—as indeed, they remain to the present day. Today, the position of the reth remains, but, like the Shilluk themselves, just barely. The tiny Shilluk kingdom has been in recent decades unfortunate enough to be located precisely on the front-lines of the Sudanese civil war. ordinary Shilluk have been victims of massacres, famines, massive out-migration, and forced assimilation, to the extent that by the end of the war some were arguing there is a real danger of cultural or even physical extinction (e.g., Nyaba 2006). The peace settlement of 2005 has helped end the immediate existential crisis, but by no means brought the Shilluks’ troubles to an end (Johnson 2011). Mytho-history A word on Nilotic cosmologies In order to understand the famous Shilluk installation rituals we must first examine their mythic framework. This is somewhat difficult, since as almost all early observers point out, their Shilluk informants—much unlike their Nuer and Dinka equivalents—were not much given to cosmological speculation. Instead, everything was transposed onto the level of historical epic. Still, in either case, it would seem the same themes were working themselves, so it seems best to begin by looking at Nilotic cosmologies more generally. Nilotic societies normally treat God as a force profoundly distant and removed from the human world. Divinity itself is rendered little or no cult; at least not directly. Instead Divinity is usually seen to be “refracted” through the cosmos, immanent particularly in storms, totemic spirits, numinous objects, or anything inexplicable and extraordinary. In one sense, then, God is everywhere. In another, he is profoundly absent. Creation stories almost invariably begin with a traumatic separation. Here is one typical, Dinka version.24 Divinity (and the sky) and men (and the earth) were originally contiguous; the sky then lay just above the earth. They were connected by a rope... By means of this rope men could clamber at will to Divinity. At this time there was no death. Divinity granted one grain of millet a day to the first man and woman, and thus satisfied their needs. They were forbidden to grow or pound more.The first human beings, usually called Garang and Abuk, living on earth had to take care when they were doing their little planting or pounding, lest a hoe or pestle should strike Divinity, but one day the woman ‘because she was greedy’ (in this context any Dinka would view her ‘greed’ indulgently) decided to plant (or pound) more than the permitted grain of millet. In order to do so she took one of the long-handled hoes (or pestles) which the Dinka now use. In raising this pole to pound or cultivate, she struck Divinity who withdrew, offended, to his present great distance from the earth, and sent a small blue bird (the colour of the sky) called atoc to sever the rope which had previously given men access to the sky and to him. Since that time the country has been ‘spoilt,’ for men have to labour for the food they need, and are often hungry. They can no longer as before freely reach Divinity, and they suffer sickness and death, which thus accompany their abrupt separation from Divinity (Iienhardt 1961: 33-34). In some versions, human reproduction and death are introduced simultaneously: the woman needs to pound more grain specifically because she bears children and needs to feed her growing family. Always, the story begins with the rupture of an original unity. Once, heaven and earth were right next to each other, humans could move back and forth between them. Or: there was a rope, or tree, or vine, or some other means of passage between the two. As a result, people lived without misery, work, or death. God gave us what we needed. Then the connection was destroyed. Stories like this can be termed “Hesiodic” because, like Hesiod’s Prometheus story (or for that matter, the story of the Garden of Eden) they begin with blissful dependency—humans being supplied whatever they need from a benevolent creator—to an unhappy autonomy, in which humans eventually win for themselves everything they will need to grow and cook food, bear and raise children, and otherwise reproduce their own existence, but at a terrible cost. It does not take a lot of imagination to see these as first and foremost as metaphors of birth, the loss of the blissful dependency of the womb, which the cutting of the cord, in the Nilotic versions, simply makes unusually explicit. The problem is that once separation is introduced into the world, conjunction can only mean catastrophe. In the current state of things, when Divinity—as an absolute, universal principle—manifests itself in our lives, it can only take the form of floods, plagues, lightning, locusts, murrains. Natural disasters are, after all, indiscriminate; they effect everyone; thus, like the indiscriminate violence of divine kings, they can represent the principle of universality. But if God is the annihilation of difference, then sacrifice—in Nilotic society the archetypal ritual—is its recreation. The slaughter and division of an animal becomes a reenactment of the primal act of creation through separation; it becomes a way of expelling the divine element from some disastrous entanglement in human affairs and reestablishing everything in its proper sphere again.25 This is accomplished through violence: or to be more explicit, through killing, blood, heat, fire, and the division of once-living flesh. There is one way that Divinity enters the world that is not disastrous. This is rain. Rain—and water more generally—seen as a nurturant, essentially feminine principle, is often also treated as the only element through which humans can still experience some approximation of that primal unity. This is quite explicit in the southeastern societies studied by Simonse. The ancestors of rainmaking lines were often said to have emerged from rivers, only to be discovered by children minding cattle on the shore; in rituals, they recreated the vines that originally connected heaven and earth; they embody peace, coolness, fertilizing water (1994: 409-411). Hence during important rain-making rituals, communities must maintain a state of “peace” (edwar). Physical violence, drumming, shouting, drunkenness, dancing are all forbidden; even animals sacrificed in rain ceremonies had to be smothered, so no blood was spilled, and they had to be imagined to go to their deaths voluntarily, without resistance. The state was ended with a bloody sacrifice at the end of the agricultural season. Edwar though this was simply an exaggerated version of the normal mode of comportment with the community—within human, social space—since even ordinarily, hot, bloody, violent activity was exiled to the surrounding wilderness. This was true of hunting and war but it was also true of childbirth (the paradigm of traumatic separation): women in labor were expected to resort to the bush, and, like returning hunters or warriors, had to be purified from the blood spilled before returning to their communities (1994: 412-416). The legend of Nyikang The human condition, then, is one of irreparable loss and separation. We have gained the ability to grow our own food, but at the expense of hunger; we have gained sex and reproduction, but at the cost of death. We are being punished, but our punishment seems utterly disproportionate to our crimes. This is another element stressed by Lienhardt, and another way in which the Nilotic material resonates with the Abrahamic tradition. None of Lienhardt’s informants claimed to understand why wishing to have a little more food was such a terrible crime. It is our fate as humans to have no real understanding of our situation. If God is just, at the very least we do not understand in what way He is just; if it all makes sense, we cannot grasp quite how. It is possible that ultimately, there simply is no justice. When God is invoked, in Nilotic languages—including Shilluk—it is ordinarily as an exclamation, “Why, God?,” above all when a loved one falls sick, with the assumption that no answer will ever be forthcoming. Now, the Shilluk appear to be one of the few Nilotic peoples for whom such creation myths are not particularly important. The Shilluk past begins, instead,with an historical event: the exile of Nyikang from his original home. Still, one story is quite clearly a transposition of the other. Nyikang himself is the son of a king whose father descended from Heaven.26 His mother Nyakaya was a crocodile, or perhaps part crocodile: she continues to be revered as divinity inhabiting the Nile.27 He is sometimes referred to as “child of the river.” Originally Nyikang and his brother Duwat lived in a faraway land by a great lake or river in the south. They speak of it as the end of the earth, or some call it the head of the earth. . . . In that land death was not known. When a person became feeble through great age, he was thrown out in the cattle yard, or in the road near it, and the cows would trample him until he had been reduced to the size of an infant, and then he would grow to manhood again (Oyler 1918b: 107). Other versions downplay this element—probably because the story that follows turns on a dispute over royal succession, and it is difficult to understand how this would come up if no one ever died. In some the people are divided over who to elect. In others, Nyikang is passed over in favor of his half-brother Duwat, seizes some royal regalia, and flees with his son Cal and a number of followers. Duwat follows in pursuit. In the end the two confront each other on either side of a great river. In some versions (Hofmayr 1910:328) Duwat curses his brother to die, thus bringing death into the world. In others, he simply curses him never to return. Always, though, the confrontation ends when Duwat throws a digging stick at his brother and tells him he can use it to dig the graves of his followers. Nyikang accepts the stick, but defiantly, announces he will use it as an agricultural implement, to give life, and that his people will thus grow food and raise children to overcome the ravages of death (Hofmayr op cit, Oyler 1918b: 107-108, Westermann 1912: 167, Lienhardt 1979: 223). Obviously, this is another version of the creation story: the loss of a blissful deathless paradise where people were nonetheless permanently infantilized by their dependence on higher powers (in this version, arguing over succession to the kingship when the king in fact will never die.) Even the digging stick reappears. This is a story of loss, but—as in so many version of this myth—also a defiant declaration of independence. Nyikang’s followers create a kind of autonomy by acquiring the means to reproduce their own life. Turning the symbol of death into an instrument of production is thus a perfect symbol. Nyikang’s first sojourn is at a place called Turra, where he marries the daughter of the local ruler Dimo and has a son, the rambunctious and unruly Dak. Conflicts soon develop, and there are a series of magical battles between Nyikang and his father-in-law, which Nyikang always wins. Dak grows up to become a scourge of the community, attacking and pillaging at will. Finally, the entire community joins together to kill him. They decide they will sneak up on him while he’s relaxing outside playing his harp. According to Riad’s informant “they were very afraid that Nyikang would avenge his son’s death if only a few people murdered Dak, so they decided that all of them would spear him and his blood would be distributed upon all of them” (1959: 145). In other words, having been victims of arbitrary predatory violence, they adopt the same logic Simonse describes in the killing of sacred kings. “The people” as a whole must kill him. In this case, however, they do not succeed. Nyikang (or in some versions Dak) receives advance warning, and comes up with the idea of substituting an effigy made of a very light wood called ambatch, which he places in Dak’s stead. The people come and one by one spear what they take to be the sleeping Dak. The next day, when the real, live Dak appears at what is supposed to be his own funeral, everyone panics and runs away (Westermann 1912: 159; Oyler 1918b: 109; Hofmayr 1925: 16; Crazzolara 1951: 123-127). This is a crucial episode. While neither Nyikang or Dak are, at this point, kings (they are both later to become kings), the story is clearly a reference to the logic described by Simonse: that both king and people come into being through the arbitrary violence of the former, and the final, unified retaliation of the latter. At the same time it introduces the theme of effigies. Nyikang and Dak are, indeed, immortalized by effigies made of ambatch wood, kept in the famous shrine of Akurwa, north of Fashoda. These play a central role in the installation of a new reth and since Evans-Pritchard at least have been seen as representing the eternity of the royal office, as opposed to the ephemeral nature of any particular human embodiment. Here the first effigy is created literally as an attempt to cheat death. Even more, as we’ll see, it seems to reflect a common theme whereby the people’s anger and hostility—however paradoxically—becomes the immediate cause of the king’s transcendence of mortal status. To return to the story: Nyikang, Dak, and their small band of followers decide the time has come to move on and seek more amenable pastures. They have various adventures along the way. Here Dak serves as Nyikang’s advance guard and general, often getting himself in scrapes from which Nyikang then has to rescue him. The most famous is his battle with the Sun, in which Nyikang again confirms his aquatic character. Dak is the first to pick a fight with the Sun, and at first, he and his father’s followers are scorched by the Sun’s terrible heat, forcing Nyikang to revive many by sprinkling water over them. In the end Nyikang manages to best the enemy by using water-soaked reeds to slash—and thus “burn”—the legs of the Sun, who is thus forced to retreat (Westermann 1912: 161, 166; Oyler 1918b: 113-114, Hofmayr 1925: 18, 55; see Lienhardt 1954: 149, Schnepel 1988: 448). Finally, he enters Shilluk-land, settles his followers, brings over existing inhabitants, even—in many stories—discovering humans masquerading as animals and revealing their true nature, and turning them into Shilluk clans. The latter is actually a curious element in the story. Godfrey Lienhardt (1952) insisted that unlike Nuer or Dinka heroes, who as ancestors, created their people as the fruit of their loins, Nyikang creates the Shilluk as an “intellectual” project. He discovers, transforms, gives names, grants roles and privileges, establishes boundaries, gathers together a diverse group of unrelated people and animals and makes them equal parts of a single social order. This is true, though putting it this way rather downplays the fact that he does so through right of conquest: that is, that he appears amidst a population of strangers who have never done anything to hurt him and threatens to kill them if they do not do his will.28 It is not as if such behavior was considered acceptable behavior by ordinary people under ordinary circumstances. In most stories, the figure of Nyikang is saved from too close an association with unprovoked aggression by effectively being redoubled. He plays the largely intellectual role, solving problems, wielding magic, devising rules and status, while the sheer arbitrary violence is largely pushed off onto his son and alter ego, Dak. In the Shilluk heartland, especially, Nyikang is always described as “finding” people who fell from the skies or were living in the country or fishing in the river, and assigning them a place and a ritual task (to help build some house or shrine, to herd Nyikang’s sacred cattle, to supply the king with certain delicacies, etc.). Only in the case of people who transform themselves into animals—fish, turtles, fireflies, etc.—does he usually have to call in Dak, to net or spear or otherwise defeat them, whereon they ordinarily turn back into human beings and submit themselves. Submission is what renders people Shilluk (the actual word, Chollo, merely means subjects of the reth.)29 Though in a larger sense, intellectual understanding and physical conquest are conflated here; the stories of shape-shifters are paradigmatic: one can only tell what they really are by successfully defeating, even skewering them—that is, literally pinning them down. For all this, Nyikang’s conquest of Shilluk-land remains curiously unfinished. The myths specify that he managed to subdue the southern half of the country, up to about where the capital is now. After this things stalled, as the people, tired of war, begin to murmur and increasingly, openly protest Nyikang’s leadership. Finally, at a feast held at the village of Akurwa (what is later to become his temple in Fashoda), Nyikang chides his followers, instructs them on how to maintain his shrine and effigy, and vanishes a whirlwind of his own creation. Nyikang, all Shilluk insist, did not and could never die. He has become the wind, manifest in animals who behave in strange and uncharacteristic ways, birds that settle among crowds of people; he periodically comes, invisible, to inhabit one or another of his many shrines (Seligman 1911: 220-26, Seligman 1934, Westermann 1912: xlii, Oyler 1918a, Hofmayr 1925: 307, Howell and Thomson 1946: 23-24). Above all he remains immanent in his effigies, and the sacred person of the king. Yet in the story, his transcendence of the bonds of mortal existence follows his rejection by the people. Neither is this mere mumbling and discontent: some versions make clear there was at least the threat of actual rebellion. In one (Crazzolara 1951: 126), Nyikang is speared in the chest by an angry follower. He survives, but then assembles his people to announce his ascent. In every version, he is replaced by an effigy made of ambatch, and remains as the vehicle of the prayers of his people, as their intercessor before God. It is through Nyikang, for example, that the king appeals to God for rain (Schnepel 1991: 5859). Though even here the relationship of animosity does not disappear. Unlike more familiar gods, who by definition can do no wrong, the hero continues to be the object of periodic anger and recrimination: Their veneration of Nikawng does not blind their eyes to his faults. When a prayer has been offered to Nikawng, and the answer is not given, as had been hoped, the disappointed one curses Nikawng. That is true especially in the case of death. When death is approaching, they sacrifice to Nikawng and God, and pray that death may be averted. If the death occurs the bereaved ones curse Nikawng, because he did not exert himself in their behalf (Oyler 1918b: 285). This passage gains all the more power when one remembers that illness itself was often assumed to be caused by the attacks of royal spirits—most often, Dak—and that mediums possessed by the spirit of Nyikang were the most common curers. Yet in the end we must die, as Nyikang did not; his transcendence of death resulted from, and perpetuates, a relation of permanent at least potential antagonism. In fact, it was not just Nyikang. None of the first four kings of Shillukland died like normal human beings. Each vanished, their bodies never recovered; all but the last were then replaced by an effigy. Nyikang was replaced by his timid elder son Cal, who disappeared in circumstances unknown; then by the impetuous Dak, who also vanished in a fit of frustration over popular grumbling over his endless wars of conquest, and finally, by Dak’s son Nyidoro. Nyidoro marks a point of transition. He vanished, but only after death. Nyidoro was murdered by his younger brother odak, whereon his body magically disappeared. As a result there was some debate over where he merited a shrine and effigy at all, but in the end it was decided that he did. If Nyidoro was the first king to die, his killer and successor, odak was the first to be ritually killed. This, however, was a consequence of not of internal conflict (as in the case of his own usurpation), but external warfare: odak was defeated in a battle with the Dinka and the Fung. After witnessing the death of all of his sons except one, he threw Nyikang’s sacred spears in the river in a gesture of despair, crying “now all my sons are dead.” Needless to say this greatly hurt the feelings of the one son who remained alive. This young man, named Duwat, had been endlessly belittled by his father in the past, and this was the final straw. After promising his father he would degrade all those sons’ children to commoners, he snatched one of the spears from the river and single-handedly drove the enemy away (Hofmayr 1925: 66-68, 260-62). Apparently odak was discreetly finished off soon afterwards, and when Duwat became king, one of his first acts was to degrade the descendants of his brothers to a lower status than the royal clan. They became the Ororo, excluded from succession, but who nonetheless play a key role in royal ritual. The story began with a Duwat, and with this second Duwat, one might say the first round of the mythic cycle comes to an end. It begins with stories modeled on birth and ends with stories of death: first, the non-deaths of Nyikang and Dak, rejected by their subjects; then, establishing the two typical modes of putting an end to a particular holder of the royal office, that is, either through internal revolt (challenge by an ambitious prince) or being ritually put to death. Figure 1. mythic origins of the Ororo and the Royal line. Note: solid arrows refer to rulers who, rather than dying, vanished and were replaced by effigies; the broken arrow refers to rulers who died but whose body vanished and was not replaced by an effigy. The role of the Ororo is especially important. This is a class who represent a veritable institutionalization of this constitutive relation of hostility, and potential violence, on which the eternity of the kingdom is founded. Generally, princes who are not elected found their own lineage within the royal clan named after their royal ancestor, and his tomb becomes their lineage shrine. In theory, the king can degrade any of these branches to Ororo status by entering into their lineage shrine at night and performing certain secret rites, but the shrines are guarded and if they’re caught trying to enter, the attempt is considered to have failed. one reth (Fadiet) is remembered to have failed in an attempt to reduce the descendants of Nyadwai to Ororo status but it is not clear if any other king has ever been successful (Pumphreys 1941: 12-13, Hofmayr 1925: 66; Howell 1953: 202). Most sources suggested none have: another dramatic reflection on the limited power of Shilluk kings. Some (e.g., Crazzolara 1951: 139) suggest that one reason a king might wish to do so is that marriage is forbidden within the royal lineage; it is only by reducing a branch to Ororo status that a king can then take one of its daughters for his wife.30 Moreover, it is precisely this degraded nobility whose role it is to preside over the death of kings. Male members of the caste who accompany the king during ceremonies are sometimes referred to as the “royal executioners,” but here meaning not that they execute others on the king’s orders, but rather that it is they who are in charge of presiding over the execution of the king. A reth would always have a certain number of Ororo wives; it is they who are expected to announce when he is sick or failing in his sexual powers; according to some, it is they who actually suffocate the king (Seligman 1911: 222; cf. note 21). In other versions it is the male Ororo bodyguard, who also preside over his burial.31 All sources stress it is difficult to know anything for absolute certain about such matters, about which discreet people knew better than to much inquire, and doubtless practices varied, but it is critical that the king was constantly surrounded by those he had originally degraded, and who were eventually to kill him. At this point we have reached historical times, which begin with the long and prosperous reign of King Bwoc, immediately followed by Tokot, Queen Abudok, and the historical creation of the sacred kingship at the end of the 17th century. There is one last story worth telling here. This is the story of the mar. The mar was some kind of talisman or element of royal regalia that had originally belonged to Nyikang. By the early twentieth century no one quite remembered what it had been: a jewel of some kind, or perhaps a crystal, or a silver pot. According to some, it was a magical charm capable of assuring victory in war. According to others, it was a general token of prosperity and royal power (Hofmayr 1925: 72-75, Paul 1952). According to Westermann (1912: 143-144) the marwas a silver pot that, waved in front of one’s enemies, caused them to flee the field of battle. Tokot employed it in many successful wars against the Shilluk’s neighbors, many of whom he incorporated into Shillukland, but eventually—a familiar scenario now—his followers grew tired of fighting far from their wives and families, and began to protest and refuse his orders. In a fit of pique, he threw the mar into the Nile. Here the story fast-forwards about a half century to the reign of Atwot (c1825-1835), who is elected as a warrior king on the behest of a cluster of settlements plagued by Dinka raiders. He fights a battle with the invaders but is defeated, so, in a bold move, he decides to retrieve the talisman. Atwot consults with the descendants of Tokot’s wives at his lineage shrine, and, despite widespread skepticism, rows out with his companions to the spot where the mar was lost, sacrificing three cows along the way, and dives to the bottom of the river. He remains underwater so long his companions think he is lost, but after many hours, returns with the genuine article. Atwot proceeds to raise an army, conquers the Dinka and is victorious against all that stand in his path. However, before long, the same thing begins to happen: he is carried from conquest to conquest, but his warriors begin protesting the incessant wars, and finally Atwot too throws the pot back in the river in frustration. There have been no subsequent attempts to retrieve the mar.32 The story seems to be about why the Shilluk kingdom never became an empire. It is as if every time kings move beyond defending the home territory or conducting raids beyond its borders, every time they attempt to levy armies and begin outright schemes of conquest, they find themselves stymied by protests and passive resistance. They respond with passive aggression: vanishing in a huff, throwing precious heirlooms in the river. As we’ll soon see, the scene of the king sacrificing cows and then diving down into the river to find a lost object appears to be a reference to a stage in the inauguration ceremonies in which the candidate has to find a piece of wood that will be made into new body of Nyikang. Yet here, instead of an image of eternity, the river becomes an image of loss. According to one source (1952), the mar was “the luck of the Shilluk,” now forever lost. It seems likely the debate over the nature of the mar reflected a more profound debate about whether military good fortune was always luck for the Shilluk as a whole—a question on which royal and popular perspectives are likely often to have been sharply divided. And the fact that such arguments were said to be going on in the time of Tokot, in the generation immediately before the creation of the institutions of sacred kingship, once again underlines how much debate there was at that time about the very purposes of royal power. Return to Fashoda At this point we can return to those institutions themselves. First of all, a word about the role of violence. Godfrey Lienhardt (1952) insists Nyikang (and hence, the king) has to be seen only as a continuation of the Shilluk conception of God. God is ordinarily seen as neither good nor evil; anything extraordinary contains a spark of the divine; above all, God is the source of life, strength, and intelligence in the universe. Similarly, Nyikang is the source of Shilluk custom, but not, necessarily, of a system of ethics, and kings—who are referred to as “children of God”—were admired above all for their cleverness, and for the ruthless ingenuity with which they played the game of power.33 Royals regularly slaughtered their brothers and cousins as a preemptive measure, with the assumption that they were almost necessarily plotting against the king; assassination and betrayal was expected, and successful conspirators, admired. Lienhardt concludes that intelligence and success (the latter typically reflected in prosperity) were the main Shilluk social values: “kings, and all others inspired by juok [divinity], are sacred because they manifest divine energy and knowledge, and they do so by being strong, cunning, and successful, a well as appearing to be in closer touch with the superhuman than ordinary men” (Lienhardt 1952: 160; so too Schnepel 1988: 449). All this seems true—but the situation seems to have been rather more complicated. God was also spoken of as the source of justice, the last resort of the poor and unfortunate. The king of course dispensed justice as well. The apparent paradox is, as I’ve emphasized, typical of divine kingship: the king, like God, stands outside any moral order in order to be able to bring one into being. Still, while a prince who successfully lured potential rivals to a feast and then massacred them all might be admired for his cunning, this was hardly the way ordinary people were expected to behave. Nothing in the literature suggests that if a commoner, or even an ordinary member of the royal clan, decided to act in a similar fashion to head off later quarrels over their father’s cattle, this would be regarded as anything but a despicably criminal—by the king (if the matter was brought before him) or by anybody else. It was, rather, as if ruthlessness of this sort was to be limited to the royal sphere, and the royal sphere carefully contained and delimited from ordinary life in part for that very reason. Father Crazzolara, for instance, insists that this was precisely what the commoner chiefs (called Jago) who elected the king wanted: to ensure that everything surrounding kings and princes remained shrouded in mystery, so that it had no effect on ordinary life. “Disputes and intrigues among members of the royal family were known to exist and were shared by the great Jagos and their councilors, but seldom affected the people at large. . . . Strifes and murders in the higher social ranks were settled among the great men, in great secrecy, and could never imperil the unity of the country” (Crazzolara 1951: 129). Indeed, he observed, most ordinary Shilluk would never have dreamed of approaching the royal residence at Fashoda, and when the king did set out on a journey, “most people used to go into hiding or keep out of his path; girls especially do so” (ibid.: 139). At the same time, the organization of the kingship those chiefs upheld, with no fixed rule of succession, but rather, a year-long interregnum during which dozens of potential candidates were expected to jockey for position, plot and intrigue against each other, more or less guaranteed that only very clever, and very ruthless, men could have much chance of becoming reth. It also guaranteed that the violence on which the royal office was founded on always remained explicit, that reths were never too far removed from the simple bandit kings from which they were presumably descended. Everything is happening as if the reth’s subjects were resisting both the institutionalization of power, and the euphemization of power that seems to inevitably accompany it. Power remained predatory. Take for example the matter of tribute. The king’s immediate power was based in the Bang Reth, his personal retainers, a collection of men cut off from their own communities: orphans, criminals, madmen, prisoners taken in war. He provided them with cattle from his herds, along with ornaments and other booty; they minded his cattle, accompanied royal children, acted as spies, and accompanied him on raids against Arab or Dinka neighbors. They did not, however, have anything to do with the collection of tribute. According to one colonial source, there was no regular system for exacting tribute. Instead, the king would intervene in feuds between communities that had resisted his attempts at mediation: The Reths. . . were extremely rich in cattle. They acquired these largely in the following way. Whenever one settlement waged unjustified war upon another or refused repeatedly to obey his order, the Reth would raise as a “royal levy” the adjacent settlements, who would go and drive off the malefactors’ cattle and burn their villages. The strength of the levy would vary with the readily calculable strength of the opposition but a good margin of safety would be allowed to ensure that the levy would win. It is said that such levies were in fact seldom resisted, the victim being glad to save their skins at the cost of most of their cattle. The participants in the levy got a percentage of the cattle taken but the majority went to the Reth (Pumphreys 1941: 12; compare Evans-Pritchard 1948: 15-16). Significantly, it was precisely in the 1840s when Shilluk kings, emboldened by an alliance with foreign merchants, began trying to move beyond raiding and create a systematic apparatus for the extraction of tribute, that many ordinary Shilluk began to cast doubt on the very legitimacy of the kingship, and to throw in their lot with a different set of predatory freebooters (Mercer 1974: 423-24). As it turned out, the results were catastrophic—the Arab slave-traders with whom they aligned themselves turned out to be far more ruthless and destructive than anything they had previously encountered—but the pattern remains clear. As in the stories about the mar, popular resistance appeared at exactly the point where royal power tried to move beyond mere predatory raiding, and to formally institutionalize itself. The kings’ rather unsavory retainers lived at the margins of Fashoda. Its center was composed of his own compound, and the houses of his wives. All sorts of dark rumors surrounded the place. According to Seligman’s account, quoted near verbatim in The golden bough: During the day the king surrounded himself with his friends and bodyguards, and an aspirant to the throne could hardly hope to cut his way through them and strike home. It was otherwise at night. For then the guards were dismissed and the king was alone in his enclosure with his favourite wives, and there was no man near to defend him except a few herdsmen, whose huts stood a little way off. The hours of darkness were therefore the season of peril for the king. It is said that he used to pass them in constant watchfulness, prowling round his huts fully armed, peeping into the blackest shadows, or himself standing silent and alert, like a sentinel on duty, in some dark corner. When at last his rival appeared, the fight would take place in grim silence, broken only by the clash of spears and shields, for it was a point of honour with the king not to call the herdsmen to his assistance (Frazer 1913: 22, Fraser 1990: 200-201). This was to become one of Frazer’s more famous romantic images, but in the first edition in which the Shilluk material appear, in 1913, the passage was accompanied by a footnote explaining that “in the present day and perhaps for the whole of the historical period” succession by ritual combat “has been superseded by the ceremonial killing of the king” (Frazer 1913: 22n1). This would suggest we are not dealing with a victorian fantasy here—or not only—but with a Shilluk one, a legend about the ancient past.34 But even here things are confusing: Frazer is simply citing Seligman, but Seligman also contradicts himself by simultaneously insisting (i.e., 1911: 222; also Hofmayr 1925: 175) that even in his own day, reths did tend to sleep during the day and keep armed vigil at night, and that the drowsy behavior of the reth, the one time he did meet one, would appear to confirm this. In fact, such stories seem to be typical of the mysteries surrounding royalty. very few people knew what really went on at Fashoda, and everything concerning kings was tinged with confusion, fascination, and danger.35 All evidence suggests that, except perhaps during periods of civil unrest or when the reth had concrete evidence of some particular conspiracy, life in Fashoda was distinctly more relaxed. True, many observers do remark on the eerie quiet of the place, much in contrast with other Shilluk settlements. But this is for an entirely different reason. Fashoda was entirely lacking in children (e.g., Riad 1959: 197). As the reader will recall, not only was the settlement occupied almost entirely by women, the king’s wives were sent back to their natal villages in order to give birth, and the children were not raised in Fashoda. It is a place where there is sex, but no biological reproduction, no nursing, no child-rearing—but also, no old age, grave illness or natural death, since the king is not allowed to grow frail and pass away in the normal fashion, and his wives normally return to their parents’ settlements before they grow very old. All of this very much recalls the villages described by Simonse further to the south, where birth and killing—or anything involving the spilling of blood—were considered “hot,” violent, dangerous activities which should be kept entirely outside the confines of inhabited space. Even animal sacrifices had to be, like the Shilluk reth, smothered so that no blood was spilled. These restrictions were especially severe during the agricultural season, since they were the key to ensuring rain. Rain, in turn, was the temporary restoration of that happy conjunction of heaven and earth was severed in the beginning of time. It seems hardly coincidental, then, that almost all of the reths ritual responsibilities involved either presiding over ceremonies appealing to Nyikang to send the rains, or harvest rituals (Oyler 1918a: 285-286, Seligman and Seligman 1932: 80-82)—or even, that it was considered a matter of principle that the king and his wives did work at least a few symbolic fields, and followed the same agricultural processes as everybody else (Riad 1959: 196). I might add here that many of the more exotic-seeming practices of the capital seem to be adopted from ordinary Shilluk practice. As the reader will recall, all women, for example, were expected to leave their husbands and return to their natal villages in the sixth month of pregnancy (Seligman and Seligman 1932: 69)—though in the case of non-royals, they returned with their baby shortly after giving birth—old people deemed to be suffering unduly from incurable conditions were often “helped to die” (Hofmayr 1925: 299). According to Howell, even the effigies had a kind of demotic precedent, since if someone dies far from home her kin can hold a ceremony to pass her soul to a stick of ambatch, which is also the wood used to make effigies, so that it can be buried in her stead (1953: 159; see also Oyler 1918: 291). What I have described above, at any rate, were the things that an ordinary Shilluk was likely to actually know about Fashoda. The overall picture seems clear. Fashoda was a little image of heaven. However imperfect, it was the closest one could come, in these latter days, to a restoration of the primal unity that preceded the separation of the earth and heaven. It was a place whose inhabitants experience neither birth nor death, although they do enjoy the pleasures of the flesh, ease, treasure and abundance (there was rumored to be a storehouse of plundered wealth and certain clans were charged with periodically bringing the reth tasty morsels), and also engaged in agricultural production—if, like the original couple, Garang and Abuk, only just a little bit. It is, then, an undoing of the dilemma of the human condition. Obvious it was a partial, provisional one. The Shilluk reth was, as Bernhardt Schnepel aptly put it (1995), “temporarily immortal.” He was Nyikang, but he was also not Nyikang; Nyikang was God, but he was also not God. And even this limited degree of perfection could only be brought about by a complex play of balanced antagonism that would inevitably engulf him in the end. The intallation ritual: description All of this, I think, gives us the tools with which to interpret the famous Shilluk installation ceremonies. One must bear in mind here that this ritual is one of the few occasions in which an ordinary Shilluk was likely to actually see a reth: the others being while he is administering justice, and possibly, during raids or war. Almost every clan played some role in the proceedings, whether in the preparation or rebuilding of royal dwellings beforehand to bringing sacrificial animals, regalia, or presiding over certain stages of the rituals themselves. It was in this sense the only real “national” ritual. The sense of popular participation was made all the more lively since, the rituals being so endlessly complicated and there usually having been such a long a time since they had last been performed, each step would tend to be accompanied by lively debate by all concerned as to what the correct procedure was. When a king dies, he is not said to have died but to have “vanished,” or to have “gone across the river”—much as was said of Nyikang. Normally, Nyikang is both immanent in the person of the king, and also, in an effigy kept in a temple in the settlement of Akurwa, north of Fashoda. This effigy too is destroyed after a king’s death. The reth’s body is conveyed to a sealed hut and left there for about a year, or at least until it is certain that nothing remains but bones; at that point, the Ororo will convey the skeleton to its permanent tomb in the reth’s natal village, and conduct a public a funeral dance. It is only afterwards that a new reth can be installed. This interim period, while the king’s body lies decomposing and Nyikang’s effigy is gone, is considered a period of interregnum. It is always represented as a time of chaos and disorder, a “year of fear.” According to Howell and Thompson, who wrote the most detailed account of the rituals, messengers send out word that “There is no land—the Shilluk country has ceased to be” (1946: 18). others speak of the land as “spoiled” or “ruined,” the same language used in Dinka and Nuer songs to describe the state of the world since the separation of heaven and earth (Howell 1952: 159-160). At any rate it is clear that with the rupture in the center, the image of perfection on earth and thus guarantor of the kingdom, everything is thrown into disarray. During this time, all important matters are put on hold, other than, presumably, the frantic politicking surrounding the election of the new reth. There were usually at least a dozen potential candidates. Settlement chiefs lobbied for their favorites, princesses offered bribes, royals conspired and plotted and there was a real fear that everything would descend into civil war. As the chief of Debalo explained in 1975: It is the period when we fear each other. I fear you and you fear me. If we meet away from the village, we can kill each other and no-one will prevent us. So the meaning of wang yomo [year of fear] is that we are all afraid and keep to our own homes, because there is no king (Singer 1975, in Schnepel 1988: 443). This sounds very much like a Hobbesian war of all against all. Still, when the chief suggests that the chaos is the result of the mere absence of the king’s power to impose justice, one must bear in mind that this is a local official who grew up in a time of strong state authority, during which the reth was subordinated to, but also supported by, Sudanese police. In earlier centuries, as we’ve seen, the reth did not play this role. Rather, it would seem that the interregnum was the time when royal politics—ordinarily kept at a safe distance from ordinary people’s lives—really did spill over into society as a whole, and that, as a result, anyone became a potential enemy. Traditionally, the interregnum lasted roughly a year, and ended during the “cool months” after the harvest in January and February, when the new election would be held so that the reth could be installed. It was considered important the installation be completed in time to allow the new reth to preside over rainmaking ceremonies in April. Neither was the election itself, conducted by twenty major chiefs or Jago, presided over by the Chief of Debalo, definitive. As Schnepel (1988: 444) notes, the college of electors did not so much the select a king as identify the candidate the chiefs feel most likely to be able to successfully endure the series of tests and crises that make up the ritual. Every step, in fact, was a kind of an ordeal and, thus, another judgment. Candidates often feared assassination at critical points of the ceremony; it was said if they were so much as injured in the course of them, they would be declared unfit and disqualified. (For commoner participants, the rituals were also tinged with fear, but at the same time, enormously entertaining. The effigies of Nyikang and Dak, according to most sources, were seen as particularly amusing.) Let me lay out the events, in abbreviated form, in roughly their order of occurrence.36 Once the electoral college, presided over by chiefs of the Northern and Southern halves of the country, had reached a decision, word was sent to the prince, who could be expected to be lingering nearby: The method of summoning the reth was interesting. . . . The chief of Gol Nyikang37 sent his son by night to get him. Whether or not there was a mock fight between the selected candidate and the messengers I do not know, but the traditional form of the words announcing the choice was told to me. It is an interesting example of Shilluk “understatement” when talking of the reth—”you are our Dinka slave, we want to kill you” which means “You are our chosen reth, we want to install you in Fashoda” (Thomson 1945: 154). (At this point it is possible to finally proceed with the final burial of the old king and the initiation of his shrine—this, unlike the election, which is primarily an affair of commoners, is presided over strictly by royals.)38 The candidate-elect is now summoned, shaved and washed by Ororo women, and placed in seclusion. Immediately thereafter, select detachments of men from the Northern and Southern halves of the country begin, set out on expeditions to acquire materials needed in the ritual, and particularly, with which to remake the effigy of Nyikang.39 This effigy is so important, and so famous, that it is fitting to offer a full description. Actually Nyikang’s effigy is one of three such: in addition to his, there is also an effigy of his rambunctious son Dak, and finally, one of his older, but timid, son Cal. Nyikang and Dak’s effigies almost always appear together; the effigy of Cal is far less important, only appearing at the very last day of the ceremony. The body of Nyikang’s effigy consisted of a five and half-foot trunk of ambatch wood, adorned with cloth and bamboo, and topped with a crown of ostrich feathers. Dak is similar in composition but his body is much smaller; however, unlike Nyikang, his effigy is normally carried atop an eight-foot-tall bamboo pole. (The effigy of Cal consists primarily of rope.) ordinarily, all three are kept in Nyikang’s most famous shrine, in the village of Akurwa—said to be the very place where Nyikang vanished into the whirlwind. Their traditional keepers are a clan called Kwa Nyikwom (“Children of the Stool”), the inhabitants of the place: These effigies are not merely symbols. They may “become active” at any time, and when active they are Nyikang and Dak. The effigy of Nyikang is rarely taken on a journey in normal times, though it is often brought out to dance during religious festivals at Akurwa itself. The effigy of Dak makes periodical excursions through the country. Both effigies have an important part to play in the ceremonies of installation. The soul of Nyikang is manifest in the effigy for the occasion, and he must march from Akurwa to Fashoda to test the qualities of the new successor and to install him in the capital (Howell and Thomson 1946: 40). Before this can be done, however, the effigy of Nyikang—destroyed after the death of the former reth—has to be entirely recreated, and that of Dak, refurbished. All the expeditions that set out of the country to gather materials are organized like war parties, and some of them—such as those sent into the “raiding country” to acquire ivory, silver and cloth, originally were expected to acquire them by ambushing villages or caravans. In more recent times they have been obliged instead to buy them in markets to the north of Shillukland. However, whether they were sent outside the country to hunt ostriches or antelopes, or to gather rope or bamboo, all these parties are clearly seen as seizing goods by force, and they made little distinction between Shilluk and foreigners, since along the way “they are given, or take, what they want from Shilluk as they pass” (ibid.: 38).40 All of these expeditions also seem to be under the broad aegis of Dak, whose effigy remains in the temple during the whole of the interregnum, except when leading occasional expeditions outside.41 The “raiding country” to the north of Shillukland is seen as his particular domain. The most important of these expeditions by far is the one dispatched to find the new body of Nyikang. It is led by the effigy of Dak, accompanied by his keepers from among the Children of the Stool, along with some men from the settlement of Mwuomo in the far north of the Shilluk country, who act as divers. After sacrificing a cow so that its blood runs into the river, they set out from Akurwa in canoes to an island in the midst of the “raiding country” called “the island of Nyikang.” A drum is beaten; Dak scours the waters of the Nile; when a white bird appears to indicate the right spot, ornaments are cast into the water as an offering, along with a sacrificial ram, and a diver descends to search for an ambatch trunk of roughly the right size to make the new body of Nyikang (Howell 1953: 194). If he finds one, the body is wrapped in a white cloth and carried back to Akurwa, where both Nyikang and Dak are outfitted with their newly acquired cloth, feathers, and bamboo. But luck was not guaranteed. Riad’s informants emphasized that Nyikang himself has specifically instructed his descendants to observe this custom as an “ordeal,” to test the reth-elect, since, although the latter does not participate in the ceremony, Nyikang will not appear if he disapproves of the electors’ choice. In fact, they emphasized that if the trunk could not be found, the entire ceremony had to be conducted again, starting from Akurwa, and that after ten failures, the reth-elect would be killed and another candidate selected (1959: 189-190)—though, as with most dire warnings of the dangers of the ceremony, no one could remember a specific occasion when anything like this had actually occurred. Once Nyikang has been brought to life again in the form of an effigy, he and Dak march to the northern border of the country and begin to assemble an army, drawn from the men of the Northern half. It is said that they retrace the steps of his original conquest of the country. The effigies are carried, and surrounded, by the Children of the Stool, many armed with whips to frighten away those who come too close, followed by a retinue carrying his drums, pots, shields, spears, and bed. No one is allowed to carry weapons in the effigies’ presence, so when they stay overnight at village shrines, their hosts, who would ordinarily be carrying spears, carry millet stalks instead. During this time Nyikang would usually retire, and Dak come out to dance with, and bless, the assembled crowds. Everyone comes out to see the show, and to ask for cattle, sheep, spears, etc. But they also hide their chickens: It is usual for gifts of a sheep or a goat to be presented or exacted by Nyikang, and it was noticeable how all small stock or fowl were either shut up or driven away from the vicinity of Nyikang, for Nyikang has the right to anything he fancies. As Nyikang proceeds with Dak his son beside him, the escort chants the songs of Nyikang and Dak recounting their exploits of conquest. From time to time Nyikang turns round and dances back as if to threaten those following. When he does this, Dak rushes ahead, carried in a charging position, his body held horizontally pointed like a spear. . . (Howell and Thomson 1946: 41-42). Occasionally, though things could also get out of hand: It is accepted custom among the Shilluk that Nyikang and his followers may seize cattle, sheep or goats which cross their path (most Shilluk are wise enough to keep them out of the way) or to demand them as offerings together with other smaller gifts from the occupants of thevillages through which they pass. This licensed plundering, which is often abused beyond the bounds of piety by Nyikang’s retinue, is treated by the Shilluk with admirable tolerance... At one point on the march at Moro, however, their demands were thought to be excessive and were resisted, a demonstration which nearly ended in armed conflict and which delayed the party for a while (Howell 1953: 195). At the same time, the whole procedure is considered something of a farce. Howell remarks that “the effigies are treated by the Shilluk with a mixture of hilarity and dread: mixed emotions that are always apparent” (ibid.: 192). At any rate, it is clear enough what’s happening. The effigies, assembled from pieces drawn from outside the country, descend on Shillukland like an alien, predatory force. on one level what they are doing is all in good fun; on another, they represent forces that are quite real, and the consequences are potentially serious. Nyikang and Dak proceed from settlement to settlement, gathering their forces, retracing, as noted, their original path of conquest. often members of new communities will at first oppose them, then, energized, rally to their side. Finally, they approach Fashoda. The king has all this time been in seclusion in the capital, but on hearing of Nyikang’s passage through Golbainy, the capital of the Northern Half of the country, he flees at night to take refuge in Debalo, the capital of the Southern Half. During that night all fires are put out in both villages. The chief of Debalo challenges the reth-elect, asking his business. He replies “I am the man sent by God to rule the land of the Shilluk” (Hofmayr 1925: 145). Unimpressed, the chief has his men try to block his party from entering, leading to mock battles where, after being repelled three times, the reth-elect finally enters. At this point the fires are relit, using fire-sticks. According to Riad, three are lit in front of the king’s hut, one from the royal family, one from the Ororo, and one from the people. “These fires, one of the symbols of royalty, are never put out as long as the king lives, and are transported to Fashoda when the king moves to the capital” (Riad 1959: 190).42 Once in Debalo, the reth-elect gathers his own followers. At some times he is surrounded by men seeking forgiveness for sexual misdemeanors: he grants this in exchange for gifts of sheep and goats. At others he is himself treated “like a small boy,” belittled and humiliated by the chief, made to sleep in a rude hut and to herd sheep or cattle. He is formally betrothed to an eight- to ten-year-old girl, called the nyakwer or “girl of the ceremonies,” who will be his almost constant companion from them on. Gradually, the southern chiefs all arrive with their warriors, to match Nyikang’s army of the north. Both sides prepare for a ritual battle which is always fought along the banks of a river that represents the official border between the two divisions of the country. The candidate marches up surrounded by the Ororo, who are his bodyguards but at the same time, the symbols of his mortality. He proceeds north towards Fashoda sitting backwards on an ox, which is led by its tail, and alongside a heifer, also walking backwards. Nyikang dispatches messengers to mock him. Before crossing the river, he and the girl step over a sheep, then a black bull before crossing the river, thus consecrating them for sacrifice. It is said in earlier days he used to step over an old man who was then trampled by the people after him, usually, to death. The two forces proceed to do battle, each side unleashing a volley of millet stalks in lieu of spears. Nyikang’s followers, however, are also armed with whips, reputed to be so powerful that a direct blow could cause madness. As a result, the southern forces are put to rout, and at the height of the battle, the bearers of Nyikang and Dak sweep forward and surround the reth-elect, carrying him off as prisoner to Fashoda, together with the “girl of the ceremonies.” On their arrival, the heifer is ritually sacrificed. Once in the capital, however, the two figures begin to fuse. Nyikang’s sacred stool is taken from his shrine; a white canopy is arranged around it, and the effigies and their captives are brought inside. First Nyikang is first placed on the throne, then removed and replaced with the reth-elect. He begins to tremble, and exhibit signs of possession—the soul of Nyikang, it is said, has left the effigy and entered the king. He’s doused with cold water. At this point the effigies retreat to their shrine, and the reth is revealed to the assembled people, as his wives (newly transferred from the harem of the previous king) warm water for a ritual bath “while the Ret sat like a graven image on the chair” (Munro 1918: 151), himself now an effigy, and later was led out before the assembled people. In one case, at least, observers remarked he seemed visibly in trance. After the sacrifice of an ox, he was led to a temporary “camp” just opposite the shrine, where he was bathed in great secrecy, with water alternately warm and cool, to express the desire that he “rule with an even temper” (Howell and Thomson 1946: 64) and avoid extremes. This bath was part of a broader process of communion with the spirit of Nyikang of which was considered arcane knowledge about which outsiders should know little, but according to some, the reth spent many hours of contemplation as the soul passed fully into him. The transfer of Nyikang’s soul marked the new reth's last public appearance for at least three days. Afterwards king remained in seclusion, guarded only by some Ororo and a few of his own retainers. Once again he is treated like a boy, expected to tend a small herd of cattle, and accompanied only by his betrothed child bride. At some point, though, adult sexuality intervenes. An Ororo woman (or in some versions, there are three of them) lures the king away to the shrines on the mound of Aturwic in Fashoda and seduces him;43 while he is thus distracted, Nyikang steals out from another of the shrines and kidnaps the “girl of the ceremonies.” On the king’s return, he discovers her gone and, pretending outrage, begins searching everywhere. On finally realizing what’s happened, he confronts the chief of Kwa Nyikwom (who is acting as Nyikang’s spokesman), explaining that the girl had been properly betrothed by a payment in cattle, and Nyikang had no right to her. The chief however insists that the herds used—which are, after all, the old reth’s herds—are really Nyikang’s. Finally it comes down to another contest of arms. Both sides marshal their forces in Fashoda. This time, Nyikang is accompanied not only by the ferocious Dak but his hapless son Cal. A smaller mock battle follows, but this time, the Northerners’ whips prove ineffective. The reth sweeps in and recaptures the girl from Nyikang; finally, the effigies have to fight their way back into their own shrines, and negotiate their effective surrender. The girl remains with the king, who has, in his victory, demonstrated that he and not the effigy is the true embodiment of Nyikang. At this point the effigies disappear, and do not return for the remainder of the ceremonies. At this point, too, the drama is also effectively over. The new reth spends the next day on his throne at Aturwic, holding court amidst an assembly of the nation’s chiefs. Each places his spear head down in the ground and delivers a speech urging the new ruler to respect elders and tradition, protect the weak, preserve the nation, and similar sage advice. Drums salute their words; the king is invested in two silver bracelets that serve as marks of office; an ox is speared. Finally the king is given a tour of the capital. Everything is back in place. The newly installed reth sends cattle for sacrifice to each of the shrines of Nyikang scattered throughout the country Some weeks later he is ready to preside over his first major ritual, a series of sacrifices calling on Nyikang to call on God to send the rain. Once the first rains fall, the effigies leave Fashoda and return to their shrine in Akurwa, and do not return until the new king dies. Since the drama began with the people’s representatives announcing, “euphemistically,” that they wish to kill the candidate-elect, it might be best to end it by noting that even here, in the reth's most benevolent function, there were similar, darker possibilities. While one would imagine a newly inaugurated reth would have nothing but enthusiasm for his role as rainmaker, this was not always assumed to be the case. The king is the only authorized person to refuse or permit sacrifices at the important ritual ceremonies. The act of sacrificing animals to appease Juok, the highest spirit, and Nyikang, the demi-god, cannot be correctly undertaken without the king’s sanction. Without sacrifices the people’s wishes cannot be granted. It follows that the king is the real power in religious matters, and sometimes he withholds his beneficial powers if he feels the disloyalty of his subjects or their hatred towards him (Riad 1959: 205, citing Hofmayr 1925: 152n1). In other words, while the reth, unlike Simonse’s rainmaking kings, was not personally responsible for bring down rain through magical means, his role was, at least potentially, not so different. A drought might well be blamed on royal spite—and presumably, begin to spur a political crisis, even if it was unlikely to end with an actual lynch mob. The installation ritual: analysis To some degree, the symbolic structure of the ritual is quite transparent. There is a constant juxtaposition of North and South, the former the division of Nyikang, the latter, of the king. The North is identified with the eternal, universal “kingship”; the South, with the particular, mortal king. Hence as Evans-Pritchard put it, in the ritual, “the kingship captures the king” (1948: 27). Having been defeated as ahuman, the reth-elect becomes Nyikang, and is thus able to defeat the effigy and banish it back to its shrine. Another obvious element is the opposition of fire and water. At the same time as the image of Nyikang emerges from the river far to the north, new fires are lit in Debalo, the capital of the South, that will burn for the rest of the king’s reign and be put out when he dies. Water here is eternity. It doesn’t even “represent” eternity, it is eternity; the Nile will always be there, and always the same. With the rains, it is the permanent source of fecundity and life. It is utterly appropriate therefore that Nyikang, whose mother was a crocodile and is called “child of the river,” should emerge from its waters.44 Fire on the other hand is, like blood, the stuff of worldly transformation. In this case, the fires correspond to the mortal life of the individual king; they will exist exactly as long as he lives. It is thus equally appropriate that when the synthesis of Nyikang and reth, between the eternal principle and mortal office-holder, occurs, it should be accompanying by putting a fire to water. The “bath” during which the king becomes fully one with the demigod also unites the two elemental principles. Fire meets water as mortal man meets god. All these elements are, as I say, relatively straightforward. Other elements are less so. The most puzzling is in the role of Nyikang’s son Dak. Existing analyses—even those that have a great deal to say about the effigies (Evans-Pritchard 1946, Arens 1984, Schnepel 1988)—focus almost exclusively on Nyikang, who is always assumed to represent the timeless nature of the royal office. They rarely have anything to say about Dak. But in many ways Dak seems even more important than Nyikang: if nothing else, because (just as in the legends he is the first to transcend death through the means of an effigy) his is the only effigy that was genuinely eternal. When the king dies, Nyikang returns to his mother in the river. Dak remains. Dak’s effigy then presides over the re-creation of Nyikang’s. What is one to make of this? It might help here to return to the overall cosmological framework. The reader will recall that the Shilluk Creator is rarely invoked directly, but largely approached through Nyikang. The all-powerful being who exists in the minds of the Shilluk as a remote and amoral deity is called Juok. Juok is the Shilluk conception of God and is present to a greater and lesser degree in all things. Juok is the explanation of the unknown, the reassuring justification of all the supernatural phenomena, good and bad, of which life is made up. The principal medium through whom Juok is approached is Nyikang. The distinction between them is not clear. Nyikang is Juok, but Juok is not Nyikang. . . . Further the soul of Nyikang is reincarnate in every Shilluk reth, and thus exists both in the past and the present. Nyikang is the reth, but the reth is not Nyikang. The paradox of the unity yet separation is not easy to define. The Shilluk themselves would find it difficult to explain. Juok, Nyikang, and the reth represent the line through whom divinity runs. . . . The reth is clearly himself the medium through whichboth Nyikang and, more vaguely, Juok are approached, and is the human intercessor with God (Howell and Thomson 1956: 8). After many years of contemplation and debate scholars of Nilotic religions have learned to read such paradoxical phrases (“God is the sky, but the sky is not God”) as statements about refraction and encompassment. Nyikang is an aspect of God, but God is in no way limited to that aspect.45 We are presented, as in a rain-making ceremony, with a very straightforward model of a linear hierarchy: The reth intercedes for the people and asks Nyikang to intercede with God to bring the rains. If the rain comes, it temporarily joins everything together. However, as we’ve seen, at every point there is potential antagonism. The people may hate the reth or wish to kill him; they may curse Nyikang; the reth may withhold the rains out of resentment of the people; the king and Nyikang raise armies and do battle with each other. Only God seems to stand outside this, but only because God is so distant: in Nuer and Dinka cosmologies, where Divinity is a more immediate concern, we learn that the human condition was first created because of God’s (apparently unjustified) anger against humans, and there are even stories of defiant humans trying to make war on God and on the rain (Lienhardt 1961: 43-44). Antagonism here appears to be the very principle of separation. Insofar as the reth is not Nyikang, it is first of all because the two sometimes stand in a relation of mutual hostility. This too is fairly straightforward. Certainly, there are ambiguities—for instance, about how and whether the people themselves could be said to partake of divinity, since divinity is, after all, said to be present in everything—but these are the ambiguities typical of any such hierarchical system of encompassment. Things get a little more complicated when one examines prayers offered directly to God. Here is one in Westermann, pronounced during a cattle sacrifice to cure someone who is sick: There is no one above thee, thou God. Thou becamest the grandfather of Nyikango; it is thou (Nyikango) who walkest with God; thou becamest the grandfather (of man), and thy son Dak. If famine comes, is it not given by thee? So as this cow stands here, is it not thus: if she dies, does her blood not go to thee? Thou God, and thou who becamest Nyikango,and thy son Dak! But the soul (of man), is it not thine own? (1912: 171, also in Lienhardt 1952: 156).46 Here we have the same sort of hierarchical participation (God became Nyikang) but king is gone and Dak appears in his place: Dak’s presence might not be entirely surprising here because it was most often Dak’s attacks that make people to sick to begin with. If so Dak, however much subordinated, also represents the active principle that sets everything off. This often seems to be his function. Certainly, Dak is nothing if not active. This is especially obvious when Dak is paired with Nyikang, which he normally is. Nyikang’s effigy is larger and heavier; it is clearly meant to embody the gravitas and dignity of authority. His image thus tends to stay near the center of things. In ordinary times, the effigy remains in the temple at Akurwa even when Dak’s effigy leaves it to tour the country; when the two do travel together, it is always Dak who moves about, interacts, while Nyikang takes on a more “statesmanlike” reserve (Schnepel 1988:437). True, one could argue this is simply a consequence of Dak’s subordinate status: Nyikang is the authoritative center, Dak his worldly representative, his errand-boy. But even here there are ambiguities. Most strikingly, while Dak is smaller than Nyikang, he towers above him, always being carried atop an eight foot pole. Nyikang, in contrast, stays close to the ground; in fact his effigy is often held parallel to the ground, while Dak’s is ordinarily vertical. Similar ambiguities appear in stories about the two hero’s lives. Sometimes, especially in his youth, it is Dak who is always getting himself in trouble and Nyikang with his magical power who has to step in to save him. But later, during the conquest of Shillukland, it is more likely to be the other way around: Nyikang is foiled by some problem, and Dak proves more ingenious, or more resourceful with a spear, and manages to solve it. There is also the peculiar feature of Cal, Nyikang’s feckless older son who never accomplishes anything and whose image appears only when the effigies’ forces lose. Dak and Cal seem to represent opposites: pure aggression versus absolute passivity, with Nyikang again defining the center. Yet in what way is Nyikang superior if one is more like the useless Cal? What I would suggest is that this is not just a dilemma of interpretation for the outside analyst; it reflects a fundamental dilemma about the nature of political power that Shilluk tend to find as confusing as anybody else. Rituals can be interpreted as ways of puzzling out such problems, even as, simultaneously, they are also ways of making concrete political change in the world. Critical here is the role of the interregnum, the “year of fear.” Wherever there are kings, interregna tend to be seen as periods of chaos, violence, times when the very cosmological order is thrown into disarray. But as Bernhardt Schnepel (1988: 450) justly points out, this is the reason most monarchies try to keep them as brief as possible. There is no particular reason those organizing the Shilluk installation ceremonies could not have declared, say, a three to five-day period of chaos and terror—in fact, by the 1970s, that’s exactly what they did decide to do, abandoning the year-long interregnum entirely (ibid.: 443). If for centuries before they didn’t, it indicates, if nothing else, that this year of fear was fundamentally important. Its importance, I think, is the key to understanding the importance of Dak as well. During the interregnum, royal politics, ordinarily bottled up in the figure of the reth, overflows into society at large. The result is constant peril. During this period Nyikang is gone, and Dak alone remains. The return to normalcy begins with the stage of “preparations,” conducted under Dak’s general aegis, and often, under his direct supervision. Expeditions set out to appropriate the materials with which to reconstruct the royal office, starting with the effigies. They uproot plants, they hunt and kill animals, they ambush and plunder camps and caravans. Nor do they limit their depredations to foreigners. They “take what they like” from Shilluk communities as well. Dak’s expeditions, then, represent indiscriminate predatory violence directed at every aspect of creation: vegetable, animal, every sort of human being. As I have pointed out, “indiscriminate” in this context also means “universal.” ordinarily, when one is in the presence of a power that can rain destruction equally on anyone and everyone, that is what Shilluk refer to as Juok, or God.47 This is not to say that Dak is God (or, to be more precise, it is. God is Dak, but Dak is not God.) Dak is the human capacity to act like God, to mimic his capricious, predatory destructiveness. In the stories, this is how he first appears: raining death and disaster arbitrarily. From his own perspective “taking what he likes.” From the perspective of his victims, playing God. During the interregnum then, it is not just royal politics that spills over into society at large; it is divine power itself—the violent, arbitrary divine power that is, as Shilluk institutions ensured no one could never forget, the real essence and origin of royalty. Of course God (Juok) is not simply a force of destruction—he is also, originally, the creator of everything—and it is probably worth noting that this is also the only point in the ceremonies where anyone really makes or fashions anything. Still, this is not what’s emphasized. What is emphasized is appropriation, which is perhaps the most distinctly human form of activity. Through a combination of appropriation and creation, Dak’s people thus fashion Nyikang. Once they have done so, and Nyikang returns, he (unlike Dak) limits his depredations to his own Shilluk nation, retracing his original journey of conquest. But there seems to be a calculated ambiguity here. Do the Shilluk become Shilluk—Nyikang’s subjects—because they collectively construct Nyikang (the classic fetish king, created by his people) or because he then goes on to conquer them (the classic divine king, raining disaster or the threat of disaster equally on all)? The interregnum, then, is a time when divine power suffuses everything. This is what makes the creation of society possible. It is also what makes the creation of society necessary, since it results in an undifferentiated state of chaos and at least potential violence of all against all. Social order—like cosmic order—comes of separation, and the resultant creation of a relatively balanced, stable set of antagonisms. That one is, in fact, dealing with divine power here is confirmed by stories about the nature of the election itself. The electoral college is made up primarily of commoners, with a few royal representatives, but many insisted that “in former times” a delegation from the Nuba kingdom, the ancient allies of Nyikang, performed a ritual, a “fire ordeal,” involving throwing either sticks or pebbles in a fire, that ensured that the new reth was chosen directly by God (Westermann 1912: 122; Hofmayr 1925: 451). Even in current times, the election is taken to represent God’s choice: this is what allows the reth to tell the chief of Debalo that he is the man “sent by God to rule the land of the Shilluk” (Lienhardt 1952: 157).48 The people and God are here interchangeable. With Nyikang’s return, God leaves the picture, and Dak is again reduced to his father’s deputy. Divinity begins to be properly bottled up. Nyikang may continue Dak’s predatory ways, looting and pillaging as he reenacts his conquests, but it has all become something of a burlesque. Over the course of the ceremonies, Nyikang’s spirit, having been coaxed out of the river, is transferred first into the effigy, then, just as reluctantly, into the body of the reth-elect. In doing so, Nyikang is also moving forward in history: from his birth from the river in mythic times, to his heroic exploits in the beginning of Shilluk history, to his current incarnation in the body of a contemporary king. If we look at what is happening in the South, surrounding the candidate, however, we see a very different kind of drama. I have already mentioned the contrast between the water symbolism surrounding Nyikang and the fire symbolism surrounding the king. This is also a juxtaposition of mortality and eternity. Nyikang might be constructed, but he is constructed of eternal materials. (There will always be a river, just as there will always be ostriches and bamboo.) He then moves from the generic—and thus timeless—to the increasingly particular, and hence historic. But he will never actually die, just disappear and begin the cycle all over again. The king on the other hand is from the start surrounded by reminders of his own mortality. If the fires are the most obvious of these reminders, the most important are surely the Ororo. The Ororo preside over every aspect of the king’s mortality. As degraded nobility, their very existence is a reminder that royal status is not eternal: that kings have children, that most of them will not be kings; that eventually, royal status itself will pass away. In royal ritual, Ororo have a jurisdiction over everything that pertains to sexuality and death. They are the men who carry out the sacrifices for the king by spearing and roasting animals, they are the women that wash, shave, and seduce the king; they will provide his highest-ranking wives; they protect but eventually kill him; they officiate over the decomposition and burial of his corpse. Throughout the ceremonies, the reth-elect is surrounded by Ororo. When he is defeated and seized by Nyikang, he is plucked from amidst his own mortality. This is not to say that the reth is ever more than “temporarily” immortal. Even after his capture, the Ororo soon return. This theme plays itself out throughout the ceremony. If the drama in the North is about the gradual containment of arbitrary, divine power, the drama in the South is about human vulnerability. The reth-to-be is mocked, treated as a child, forced to ride backwards on an ox. His followers never wield arbitrary power over humans. Unlike Dak and Nyikang, they do not loot or plunder or hold passers-by for ransom. They do, however, constantly offer animals up for sacrifice. Just about every significant action of the king is marked by his stepping over (thus, consecrating) some animal that is later sacrificed.49 In one sense this is the exact opposite of what Nyikang and Dak are doing. Sacrificial meat is redistributed,50 so instead of stealing live beasts, he is distributing the flesh of dead ones. This is especially significant since, when presiding over sacrifices meant to resolve feuds, Shilluk kings have been known to state quite explicitly that the flesh and blood of the animal he sacrifices should be considered as his own (Oyler 1920: 298). Since in ordinary Shilluk sacrifices the life and blood of the creature (unlike the flesh) are said to “go up to God”—and to Nyikang—it would seem the king is here playing the part of humanity as a whole, placing himself in a willfully subordinate position to the cosmic powers that will ultimately take hold of him. In a larger sense, sacrifice—in all Nilotic religions the paradigmatic ritual—is about the re-establishment of boundaries.51 Divinity has entered into the world; the ordinary divisions of the cosmos (for instance, between humans, animals, and gods) have become confused; the result is illness or catastrophe. So while sacrifice is, here as everywhere, a way of entering into communication with the Divine, it is ultimately a way of putting Divinity back in its proper place. If the interregnum, the reign of Dak, is a time of indiscriminate violence against every aspect of creation, sacrifice is about restoring discriminations: respect (thek) to use the Nuer/Dinkaphrase,52 separation, appropriate distance. In this sense, the entire installation ceremony is a kind of sacrifice, or at least does the same thing that a sacrifice is ordinarily meant to do. It restores a world of separations. Of course, if the ritual is a kind of sacrifice, it is reasonable to ask: who is the victim? The reth-elect? A case could be made. The ceremony begins with the people informing the candidate that they wish to kill him. During his time in Debalo, he is treated very much like an ox being prepared for sacrifice: sacrificial oxen, too, are secluded, manhandled and mocked—even while those who mock them also confess their sins (Lienhardt 1961: 292-95).53 Then in the end the ox’s death becomes the token of a newly created community, its unity brought into concrete being in the sharing of the animal’s flesh. Here one could almost see the humiliated princely candidate in a messianic role, giving of himself to man and god, sacrificing himself in the name of Shilluk unity. But if so, there’s an obvious objection. He doesn’t seem to be sacrificing very much. To the contrary: the ceremonies end with the new king happily installed in Fashoda, accepting the allegiance of his subjects, inspecting the buildings, reassembling a harem, perhaps, if so inclined, plotting bloody revenge on those who had formerly insulted him. Still, all this is temporary. The king is, ultimately, destined to die a ritual death. So is the king a sacrificial victim on temporary reprieve? I would say in a certain sense he is. Every act of sacrifice does, after all, contain its utopian moment. Here, it is as if the king is suspended inside that utopian moment indefinitely—or at least, so long as his strength holds out. Let me explain what I mean by this. Normally, this utopian moment in sacrifice is experienced first and foremost in the feast, after the animal is dead, when the entire community is brought together for the collective enjoyment of its flesh. Often this is a community that has been created, patched together from previously unrelated or even hostile factions, by the ceremony itself. Even if that is not the case, they must put aside any prior differences. According to Lienhardt, for Dinka, such moments of communal harmony are the closest one can come to the direct experience of God—or, to be more specific, God in his aspect of benevolent universality: In Divinity the Dinka image their experience of the ways in which human beings everywhere resemble each other, and in a sense form a single community with one original ancestor created by one Creator. . . . When, therefore, a prophet like Arianhdit shows that he is able to make peace between normally exclusive and hostile communities, to persuade them to observe between them the peaceful conventions which they had previously observed only internally, and to unite people of different origins in a single community, he proves that he is a ‘man of Divinity’. . . . A man is recognized as a powerful ‘man of Divinity’ because he creates for people the experience of peace between men and of the uniting of forces which are normally opposed to each other, of which Divinity is understood to be the grounds (1961: 157). It is in this sense that God “also represents truth, justice, honesty, uprightness,” and so on (1961: 158). It is not because God, as a conscious entity, is just. In fact, Dinka—like most Nilotic peoples—seem haunted by the strong suspicion that he isn’t. It is because truth, justice, etc, are the necessary grounds for “order and peace in human relations,” and therefore, truth, justice, etc, are God. The point of sacrificial ritual then is to move from one manifestation of the divine to the other: from God as confusion and disaster, to God as unity and peace. Normally it is the feast which seems to act as the primary experience of God, but often the divine element takes even more concrete form in the undigested grass extracted from the cow’s stomach. It seems significant that the one shilluk sacrifice for which we have any sort of description—other than those meant to bring the rain—is aimed at creating peace between two parties to a feud (Oyler 1920b). The reth here plays the part of the Dinka prophet. After he emphasizes that the ox’s flesh and blood are really his own, the animal is speared, and the chyme, the half-digested grass in question, used to anoint the former feuding parties. “That was done to show their united condition” (ibid.: 299).54 Nuer insist that chyme is, like the blood and more generally the “life,” the part of the sacrificed animal that belongs to God (Evans-Pritchard 1956: 212; Evens 1989: 338). Generally speaking, in Nilotic ritual, chyme55 is treated as the stuff of pure potential: it is grass in the process of becoming flesh; undifferentiated substance in the way of creative transformation. As such it is itself the pure embodiment of life. It seems to me that this is the utopian moment in which the reth is suspended. Not only is he, as reth, the ground for “order and peace in human relations,” of unity and hence of justice, he is the person actually responsible for mediating and resolving disputes. This then is the social equivalent of rain, and chyme, like falling water, is the very physical substance of the divine in its most benevolent aspect. All this is stated almost explicitly in the peace sacrifice: the king is the ox, he is God, he is peace, he is the unity of all his subjects. This too, is how the reth can be both sacrificial victim in suspense, and living in a kind of small version of paradise. The installation ritual begins with a nightmare vision of a world infused with divine power, in which no separations exist, and all human relations are therefore tinged with potential violence. It is the worst kind of unity of God and world. It ends with the restoration of the best kind. In this sense, it is the transformation of divine king into sacred king. Dak, in his untrammeled form, embodies the former. The proceedings seem to be based on the assumption that the primordial truth of power—that it is arbitrary violence—has to be acknowledged so it can then be contained. one might argue the two main forms of sacred kingship identified by Luc de Heusch are the two principle strategies for doing this. Each plays itself out in a different division of the country. In the North, divine power is reduced to a fetish—literally, an effigy—which is constructed by, and hence to some degree therefore manageable by ordinary humans. In the South we see the making of a classic scapegoat king. Ultimately the two become one: the king not only becomes Nyikang, he also, at least momentarily, becomes an effigy. Ordered, hierarchical relations (God-Nyikang-king-people) are restored. The new king is (as Dak was originally) in a sense all of them at once, even as he is also the means to keep them apart, suspended in a kind of balanced antagonism. As such he is a victim himself suspended, temporarily, in miniature version of the original unity of Heaven of Earth, in a strange village with sex but without childbirth, a place of ease and pleasure, devoid of hunger, sickness and death. The paradise however is temporary, and the solution always provisional, incomplete. Arbitrary violence can never be entirely eliminated. Heaven and earth cannot really be brought together, except during momentary thundershowers. And even the simulation of paradise is bought at a terrible price. Some words by way of conclusion I have framed my argument in cosmological terms because I believe one cannot understand political institutions without understanding the people that create them, what they believe the world to be like, how they imagine the human situation within it, and what they believe it is possible or legitimate to want from it. While every cosmology is in a certain sense unique, anyone coming at this material from a background in Judaism, Christianity, or Islam is unlikely to feel on entirely unfamiliar ground here—certainly, much more familiar ground then when dealing with Polynesian or Amazonian cosmologies. There is a reason early anthropologists often saw Nilotic peoples as the closest living cousins of the Biblical patriarchs: not only are Semitic and Nilotic languages distantly connected, in each case we are dealing with semi-nomadic pastoralists, monotheists with a lineage-based form of organization. While many of the earliest ethnographers, such as Seligman himself, used such similarities to make explicit racist arguments that all that they considered the real achievements African civilization were a result of migrations from the Middle East, the problem here was simply the assumption that the similarities were somehow created by “higher civilizations” from outside, rather than being, in a sense already there because the ancestors of African and Middle Eastern pastoralists were simply more similar in their existential concerns—their way of framing the basic dilemmas of human existence—the reasons for suffering, the justice of God—than adherents of Abrahamic religions were necessarily willing to acknowledge. Though to some degree, too, they deal with issues that are universal. It would have to be so, or it would not be possible to make cross-cultural generalizations about “divine kingship,” “sacred kingship,” or “scapegoats” to begin with. This essay is really founded on two such generalizations. The first is that it is one of the misfortunes of humanity that we share a tendency to see the successful prosecution of arbitrary violence as in some sense divine, or at least, to identify it with some kind of transcendental power. It is not entirely clear why this should be. Perhaps it has something to do with the utterly disproportionate quality of violence, the enormous gap between action and effect. It takes decades to bring forth and shape a human being; a few seconds to bring all that to nothing by driving a spear into him. It takes very little effort to drop a bomb; unimaginable effort to have to learn to get about without legs for the rest of one’s life because one no longer has any as the result of one. Even more, acts of arbitrary violence are acts that for the victims and their friends and families must necessarily have enormous significance, but have no intrinsic meaning. Meaning after all implies intentionality. But the definition of “arbitrary” is that there is no particular reason why one person was shot or blown up and another wasn’t. Such acts are therefore by definition meaningless; this is just what allows arbitrary acts of weather to be referred to as “acts of God.” Meaning abhors a vacuum. Particularly when we are dealing with actions or events of enormous significance, it is hard to resist the tendency to ascribe some kind of transcendental meaning, or at least to assume that one exists. It is in this absolute absence of meaning that we encounter the Divine. Of course this is only a potential. As I remarked earlier, it is not as if any bandit who finds himself in a position to wreak havoc with impunity is necessarily going to be treated as a god (except perhaps in his immediate presence). But it is also clear that the apparatus of sacred kingship is a very effective way of managing those who do. Here I introduce my second cross-cultural generalization. The sacred, everywhere, is seen as something that is or should be set apart. As much as an object becomes the embodiment of a transcendental principle or abstraction, so much is it to be kept apart from the muck and mire of ordinary human life, and surrounded, therefore, with restrictions. These are the kind of principles of separation that Nuer and Dinka, at least, refer to with the word thek, usually translated “respect.” violent men almost invariably insist on tokens of respect, but tokens of respect taken to the cosmological level—”not to touch the earth,” “not to see the sun”—tend to become severe limits on one’s freedom to act violently. If nothing else the violence can, as in the Shilluk case, be bottled up, limited to a specific royal sphere which is under ordinary circumstances scrupulously set apart from ordinary daily affairs. We will never know the exact circumstances under which Shilluk royal institutions came into being, but the broad outlines seem fairly clear. The ancestors of the Shilluk were likely in most essentials barely distinguishable from their Nuer or Dinka neighbors—fiercely egalitarian pastoralists who settled along an unusually fertile stretch of the Nile. There they became more sedentary, more populous, but also began regularly raiding their neighbors for cattle, wealth and food. To some degree this appears to have been born of necessity; to some degree, it no doubt became a matter of glory and adventure. An incipient class of war chieftains emerged who assembled wealth in the form of cattle, women, and retainers. They became the ancestors of Shilluk royalty. However, the royal clan itself only appears to have developed, at least in the form it eventually took, after a prolonged struggle over the nature of the emerging political order, the role of women, and the power and jurisdiction of commoner chiefs. The compromise that eventually emerged appears to have been brilliantly successful in creating and maintaining a sense of a unified nation, capable of defending itself and usually dominating the surrounding territories, without ever giving the royals with their fractious politics much chance to play havoc with local affairs. Above all, ordinary Shilluk appear to have resisted the emergence of anything resembling an administrative system. Communications between Fashoda and other settlements were maintained not by officials, but principally by relations with and between royal sisters, wives and daughters. Any attempt at creating systematic tribute relations, at home or abroad, appears to have been met with such immediate and widespread protest the very legitimacy of the kingship was soon called into question. As a result the royal treasury, such as it was, consisted almost entirely of wealth that had been stolen—seized in raids against foreigners, or against Shilluk communities that resisted attempts to mediate disputes. The playful raiding during installation rituals was simply a reminder of what everyone already knew: that predatory violence was and would always remain the essence of sovereignty. Above all, there seemed to be an at least implicit understanding that such matters ought not be in any way obfuscated—that the euphemization of power was essential to any project of its permanent institutionalization, and this was precisely what most people did not wish to see. My use of the term “utopia” is somewhat unconventional in this context. By “utopia,” I mean any place that represents an unattainable ideal, particularly, an impossible resolution of the basic dilemmas of human existence—however those might be conceived. Utopia is the place where contradictions are resolved.56 Part of my inspiration here is Pierre Clastres’ argument (1977) that among the Amazonian societies he knew, states could never have developed out of existing political institutions. Those political institutions, he insisted, appeared to be designed to prevent arbitrary coercive authority from ever developing. If states ever could emerge in this environment (and it seems apparent now that, in certain periods of history, they did) it could only be through figures like the Tupi-Gurani prophets, who called on their followers to abandon their existing customs and communities to embark on a quest for a “land without evil,” an imaginary utopia where all would become as gods free of birth and death, the earth would yield its bounty without labor, and all social restrictions could therefore be set aside (H. Clastres 1995). The state can only arise from such absolutist claims, and above all, from an explicit break with the word of kinship. Luc de Heusch’s original insight on African kingship (1962) as having to mark an explicit break with the domestic order anticipates such arguments. Obviously he was to take it in what might seem a very different direction. But how different is it really? Certainly, Shilluk kings do share certain qualities with Nuer and Dinka prophets, even if unlike them they don’t predict the coming of a new world where all human dilemmas will be resolved.57 Certainly, the organization of the royal capital did represent a kind of partial unraveling of the dilemmas of the human condition. But we can also consider de Heusch’s idea of the “body-fetish.” The reader will recall that the basic idea here is that rituals of installation turn the king’s own physical person into the equivalent of a magical charm; he is the kingdom, its milk and its grain, and any danger to the king’s bodily integrity is thus a threat to the safety and prosperity kingdom as a whole. If he grows old and sickly, defeats, crop failures and natural disasters are likely to result. Hence the principle, so common in Africa, that kings ought not to die a natural death. For this reason the king “must keep himself in a state of ritual purity,” as Evans-Pritchard stressed, and also, “a state of physical perfection” (1948: 20). All sources agree on this latter point, and it is a common feature of sacred kingship. A legitimate candidate to the throne must not only be strong and healthy, he must have no scars, blemishes, missing teeth, asymmetrical features, undescended testicles, deformities, and so forth. What’s more his bodily integrity must be fastidiously maintained, particularly at ritual moments: we are told that if during the installation ceremonies the reth is injured in any way, “even if the king is only punched and blood appears” (Singer in Schnepel 1988: 444), he was immediately disqualified for office. For this reason some sources insist kings could not even fight in war, but were rather borne along as a kind of standard while others were fighting; historical narratives suggest this was not always the case, but certainly, if the king were seriously injured, this could not be allowed to stand, and he would be discreetly dispatched. The very idea of physical perfection is revealing. What does it mean to say someone is physically perfect? Presumably that they correspond to some idealized model of what a human is supposed to be like. But how do we even know what humans are supposed to be like? There is only one way: by observing actual humans. But actual humans are never physically perfect; in fact, when compared with the model of generic human we have in our heads, most seem at least slightly wrong, too short, too fat, too thin, misshapen. This is partly because when moving from tokens to types, we wipe out change and process: real humans grow, age, and so on; generic humans are, first of all, caught forever at some idealized moment of their lives. But it is also an effect of the process of generalization itself: in moving from tokens to types, we always seem to generate something which we find more proper or appealing than the tokens—or at least the overwhelming majority. In this sense the kind is indeed an abstraction or transcendental principle: the ideal-typical human, though here I am using the phrase not in Weber’s sense, but rather, from the understanding that, like Leonardo da vinci, when we try to imagine the typical, we usually instead end up generating the ideal.58 Insofar as the reth is the embodiment of the nation, and of humanity as a whole before the divine powers; insofar as he is the generic human, he must be the perfect human. Insofar as he is an image of humanity removed from time and process, he must be preserved from any harmful transformation until the point where, when this becomes impossible, he must be simply destroyed and put away. In the sense the king’s body is less a fetish than itself a kind of micro-utopia, an impossible ideal. There is always, I think, a certain utopian element in the sacred. That which is sacred is not only set apart from the mundane world, it is set apart particularly from the world of time and process, of birth, growth, decay, and also, simple bodily functions, ways in which the body is continuous with the world. I have explored this phenomenon in detail elsewhere (Graeber 1997). What is most striking in the case of sacred kingship, this is reflected above all in an urge to deny the king’s mortality; and this denial is almost invariably effected by killing people. To put it in the terms of my three proposals from the first section: there is a kind of circle here. on the one hand, insofar the dangers of divine kingship are contained by the rituals of sacred kingship, the utopian element, the degree to which kings represent a magical resolution of the dilemmas of human existence, themselves become a kind of sacred trap. But ultimately the utopia itself can only be maintained through terror. What we now think of as “the people” or “the nation” emerges as a side-effect. Rulers of early states—Egyptian and Mesoamerican pyramid-builders being only the most famous example—had a notorious tendency to develop obsessions with their own mortality. In a way this is not hard to understand; like Gilgamesh, having conquered every other enemy they could imagine, they were left to confront the one that they could never ultimately defeat. Killing others, in turn, does seem one of the few ways to achieve some sort of immortality. That is to say, most kings are aware that there are rulers remembered for reigns of peace, justice and prosperity, but they are rarely the ones remembered forever. If history will accord them permanent significance, it will most likely either be for one or two things: vast building projects (which often themselves entail the death of thousands) or wars of conquest. There is an almost literal vampirism here: ten thousand young Assyrians or Frenchmen must be wiped from existence, their own future histories aborted, so the name of Assurbanipal or Napoleon can live forever. Shilluk refused to allow their reths to engage in this sort of behavior, but in the institutions of Frazerian sacred kingship we encounter the same relation in a far more subtle way. The connection is so subtle, in fact, that it has gone largely unnoticed. But it comes especially clearly into focus if one compares the Shilluk kingdom with its most notoriously brutal cousin: the kingdom of Buganda located on the shores of Lake Victoria a few hundred miles to the south. In many ways, the similarities between the two are quite remarkable. Ganda legends too trace the kingship back to a cosmic dilemma about the origins of death; here too, the first king did not die but mysteriously vanished in the face of popular discontent; here too, the next three kings vanished as well; here too, there were elaborate installation rituals with mock battles, the lighting of ritual fires, and a chaotic yearlong interregnum. Yet in other ways the Ganda kingship is an exact inversion. Much of the difference turned on the status of women. In Buganda, women did almost all subsistence labor, while having no autonomous organizations of their own; men formed a largely parasitical stratum, the young ones organized into militarized bands, older ones into an endlessly elaborate administrative apparatus that seemed to function largely to keep the younger ones under control, or distracted in endless wars of conquest. The result was, by any definition, a bona fide state. It was also one of those rare cases when bureaucratization did not in any sense lead to any significant euphemization. While the king was not identified with any divine being, he remained very much a divine king in our sense of the term: a dispatcher of arbitrary violence, and higher justice, both at the same time. However, where the Shilluk king was surrounded by executioners whose role was eventually to kill him, the Ganda king was surrounded by executioners whose role was to kill everybody else. Thousands might be slaughtered during royal funerals, installations, or when the king periodically decided there were too many young men on the roads surrounding the capital, and it was time to round a few hundred up and hold a mass execution. Kings might be killed in rebellions, but none were ritually put to death. As Gillian Feeley-Harnik (1985: 277) aptly put it, regicide, here, seems to have been replaced by civicide.59 When David Livingstone asked why the king killed so many people, he was told that if he didn’t, everyone would assume that he was dead. Ray remarks that the capital was, as so often in such states, “a microcosm of the kingdom, laid out so that it reflected the administrative order of Buganda as a whole” (1991: 203); the king was the linchpin of the social cosmos, distributor of titles and spoils, and hence, the ultimate arbiter of all forms of value. His was a secular court, with few of the formal trappings of sacral kingship. Still, the person of the king is always sacred, and the very fact that this was a regime based almost solely on force meant that the ritual surrounding the person of the king took on a unique ferocity. The kabaka, as he was called, did not leave the palace except carried by bearers and the punishment for gazing directly at him was death. The rules of courtly etiquette, such as the prohibition against sneezing or coughing in the king’s presence. . . were considered as important as the laws of the state, for behavior towards the king’s person was regarded as an expression of one’s allegiance to the throne he represented. Thus Mutesa sometimes condemned his wives to death because they coughed while he was eating (ibid.: 172). Foreign observers like Speke and Livingstone wrote in horror of even wellborn princesses being dragged off to execution for the slightest physical infraction of courtly etiquette.60 This might seem about as far as one can get from the Shilluk court, where women were sacrosanct and it was the king who was eventually executed. But in fact it was a precise inversion. The constant element is the illusion of physical perfection at the center, which brings with it the need to suppress whatever are taken to be the most significant signs of bodily weakness, illness, or lack of physical control—and above all, the fact that this illusion was ultimately enforced by threat of death, The difference is simply that the direction of the violence is here reversed. It is, perhaps, a simple result of a different balance of political forces. In the war between sovereign and people, the reth was at a constant disadvantage. The kabaka in contrast had definitively won. His ability to rain arbitrary destruction was unlimited not just in principle, but largely in practice, and the bodies of royal women were simply the most dramatic means of its display. Granted, the situation was not ultimately viable. Such victories can never be sustained. Even in the 19th century, it was assumed that every kabaka, driven mad by power, would eventually go too far, and be destroyed—if not be real flesh-and-blood rebels, then at least by the angry ghosts of his victims. By the end of the century the entire system was overthrown and mass executions abolished. What I really want to draw attention to here though is first of all, the intimate connection between the otherworldly perfection of royal courts and their violence—to the fact that such utopias do, always, rest on what we euphemistically call “force.” The second point is that the violence always cuts both ways. This is the truth that is being acknowledged in the Shilluk stories that show how Dak’s effigy—which represents what I have called the principle of divine kingship, the human capacity to become a god through violence—was created when the people as a whole set out to kill Dak, or how Nyikang vanished and became a god when everybody hated him. What I would suggest is that this has remained the hidden logic of sovereignty. What we call “the social peace” is really just a truce in a constitutive war between sovereign power and “the people,” or “nation”—both of whom come into existence, as political entities, in their struggle against each other. Furthermore, this elemental war is prior to wars between nations. To call this a “war” is to fly in the face of almost all existing political theory, which—whether it be a matter of Karl Schmidt’s argument that the first gesture of sovereignty is declaring the division of friend and foe, to Max Weber’s monopoly of legitimate use of force within a territory, to the assertion in African Political Systems that states are entities that resolve conflict internally through law, and outside, through war—assume there is a fundamental distinction between inside and outside, and particularly between violence inside and violence outside—that, in fact, this is constitutive of the very nature of politics. As a result, just about everyone (with the possible exception of anthropologists) who wishes to discuss the nature of “war” starts with examples of armed conflicts between two clearly defined political and territorial entities, usually assumed to be nation-states or something almost exactly like them, involving a clash of armies that ends either with conquest, or some sort of negotiated peace.61 In fact, even the most cursory glance at history shows that only a tiny percentage of armed conflicts have taken such a form. In reality there is almost never a clear line between what we’d now call “war” and what we’d now call “banditry,” “terrorism,” “raids,” “massacres,” “duels,” “insurrections,” or “police actions.” Yet somehow in order to be able to talk about war in the abstract we have to imagine an idealized situation that only rarely actually occurs. True, during the heyday of European colonialism, from roughly 1648 to 1950, European states did attempt to set up a clear system of rules to order wars between nation-states, and in this period one finds a fair number of wars that do fit this abstract model, and in which all parties are playing by the rules; but these rules were only considered to apply within Europe, a tiny corner of the globe. Outside it, the same European powers became notorious for their willingness to toss aside the most apparently solemn agreements with native governments and their willingness to engage in every sort of indiscriminate violence. Even the signatories to the Geneva Conventions were careful to include a proviso that they did not apply to colonial wars. Most of the world was, effectively, a free-fire zone. Since 1950, the rest of the globe has come to be included in the system of nation-states, but as a result, since that time, no wars have been formally declared, and despite hundreds of military conflicts, there have been only a handful that have involved the clash of regular armies between nation-states. Obviously, the conceptual apparatus—the way we imagine war—is important. But it seems to me it is mainly important in occluding that more fundamental truth that the Nilotic material brings so clearly into focus. As those European travelers discovered, when asked by Nilotic kings to conduct raids or rain random gunfire on “enemy villages” that actually turned out to be inhabited by the king’s own subjects, there is no fundamental difference in the relation between a sovereign and his people, and a sovereign and his enemies. Inside and outside are both constituted through at least the possibility of indiscriminate violence. What differentiates the two—at least, when the differences are clear enough to bear noticing—is that the insiders share a commitment to a certain shared notion of utopia. Their war with the sovereign becomes the ground of their being, and thus, paradoxically, the ground of a certain notion of perfection—even peace. Any more realistic exploration of the nature of sovereignty, I believe, should proceed from examination of the nature of this basic constitutive war. Unlike wars between states, the war between sovereign and people is a war that the sovereign can never, truly, win. Yet states seem to have an obsession with creating such permanent, unwinnable wars: as the United States has passed over the last half century from the War on Poverty to the War on Crime to the War on Drugs (the first to be internationalized) and now, to the War on Terror. The scale changes but the essential logic remains the same. This is the logic of the assertion of sovereignty. Of course, no war is (as Clausewitz falsely claimed) simply a contest of untrammeled force. Any sustained conflict, especially one between state and people, will have elaborately developed rules of engagement. Still, behind those rules of engagement always lies at least the threat—and usually, periodically, the practice—of random, arbitrary, indiscriminate destruction. It is only in this sense that the state is, as Thomas Hobbes so famously put it, a “mortal god.” I don’t think there is anything inevitable about all this. The will to sovereignty is not, as reactionaries always want us to believe, something inherent in the nature of human desire—as if the desire for autonomy was always also necessarily the desire to dominate and destroy. Neither however does the historical emergence of forms of sovereignty mark some kind of remarkable intellectual or organizational breakthrough. Actually, taken simply as an idea, sovereignty, like monotheism, is an extraordinarily simple concept that almost anyone could have thought of. The problem is it is not simply an idea: it is better seen, I think, as proclivity, a tendency of interpretation immanent in certain sorts of social and material circumstances, but one which nonetheless can be, and often is, resisted. As Luc de Heusch makes clear, it is not even essential to the nature of government. only by putting sovereignty in its place, it seems to me, can we can begin to look realistically at the full range of human possibilities. Bibliography Anonymous. 1956. “The installation of a new Shilluk king.” Sudan Notes and Records 37: 99–101. Arens, W. 1979. “The divine kingship of the Shilluk: a contemporary evaluation,” Ethnos 44: 167–181. ——— 1983. “A note on Evans-Pritchard and the prophets,” Anthropos 78: 1–16. ———. 1984. “The demise of kings and the meaning of kingship: royal funerary ceremony in the contemporary southern Sudan and Renaissance France.” Anthropos 79: 355–67. Beidelman, Thomas O. 1966a. “The ox and Nuer sacrifice: some Freudian hypotheses about Nuer symbolism.” Man (n.s.) 1: 453–467. ———. 1966b. “Swazi royal ritual.” Africa 36 (4): 373–405. ———. 1981. “The Nuer concept of thek and the meaning of sin: explanation, translation, and social structure.” History of Religions 21 (2): 126–155. Beltrame, Giovanni. 1881. Il fume bianco e i Dénka. Verona: Mazziana. Burton, John W. 1980. “Sacrifice: a polythetic class of Atuot religious thought.” Journal of Religion in Africa 11 (2): 93–105. ———. 1982. “Nilotic women: a diachronic perspective.” Journal of Modern African Studies 20 (3): 467–91. Clastres, Hélène. 1995. The land-without-evil: Tupi-Guarani prophetism, translated y Jacqueline Grenez Brovender. Urbana: University of Illinois Press. Clastres, Pierre. 1977. Society against the state: the leader as servant and the humane uses of power among the Indians of the Americas. Translated by Robert Hurley. New York: Urizen Books. Crazzolara, J. P. 1950. The Lwoo. Part I: Lwoo migrations. Verona: Missioni Africane. ———. 1951. The Lwoo. Part II: Lwoo traditions. Verona: Missioni Africane. Driberg, J. H. 1932. “The status of women among the Nilotics and Nilo-Hamites.” Africa 5 (4): 404–421. Evans-Pritchard, Edward E. 1940. The Nuer: a description of the modes of livelihood and political institutions of a Nilotic people. Oxford: Clarendon press. ———. 1948. The divine kingship of the Shilluk of the Nilotic Sudan. The 1948 Frazer Lecture. Cambridge: Cambridge University Press. ———. 1951. “Shilluk king-killing.” Man 51: 116. ———. 1954. “The Meaning of Sacrifice Among the Nuer.” Journal of the Royal Anthropological Institute 84: 21–33. ———. 1956. Nuer Religion. Oxford: Clarendon Press. Evens, T. M. S. 1989. “The Nuer incest prohibition and the nature of kinship: alterological reckoning.” Cultural Anthropology 4 (4): 323–346. Feeley-Harnik, Gillian. 1984. “Issues in divine kingship.” Annual Review of Anthropology 14: 273–313. Frankfort, Henri. 1978. Kingship and the gods: a study of ancient near eastern religion as the integration of society and nature. Chicago: University of Chicago Press. Fraser, Robert. 1990. The making of The golden bough: the origins and growth of an argument. New York: St. Martin’s Press. Frazer, James. 1913a. The dying god. Part III of The golden bough. London: Macmillan. ———. 1913b. The scapegoat. Part VI of The golden bough. London: Macmillan. Frost, John. 1974. A history of the Shilluk of the southern Sudan. Ph.D. dissertation, University of California, Santa Barbara. Giesey, R. E. 1967. The royal funerary ceremony in renaissance France. Geneva: Libraire E. Droz. Girard, René. 1977. Violence and the sacred. Baltimore: Johns Hopkins University Press. Graeber, David. 1996. “Beads and money; notes toward a theory of wealth and power.” American Ethnologist 23 (1): 4–24. ———. 1997. “Manners, deference and private property.” Comparative Studies in Society and History 39 (4): 694–728. ———. 2004. Fragments of an anarchist anthropology. Chicago: University of Chicago Press. ———. 2005. “Fetishism and social creativity, or fetishes are gods in process of construction.” Anthropological Theory 5 (4): 407–438. Gray, Robert. 1961. A history of the southern Sudan 1839–1889. Oxford: Oxford University Press. Hansen, Thomas Blom and Finn Stepputat. 2005. Sovereign bodies; citizens, migrants, and states in the postcolonial world. Princeton: Princeton University Press. ———. 2006. “Sovereignty revisited.” Annual Review of Anthropology 35; 295–315. Heusch, Luc de. 1962. “Aspects de la sacralité du pouvoir en Afrique.” In Le Pouvoir et le sacré, edited by L., de Heusch, P., Derchain, A., Finet, 139–158. Bruxelles: Université Libre, Institut de sociologie (Annales du Centre d’études des religions, 1). ———. 1972. Le Roi ivre ou l’origine de l’état. Paris: Gallimard. ———. 1982. Rois nés d’un coeur de vache. Paris: Gallimard. ———. 1997. “The symbolic mechanisms of kingship; rediscovering Frazer.” Journal of the Royal Anthropological Institute 3 (2): 313–332. ———. 2000. Le roi de Kongo et les monstres sacrés. Paris: Gallimard. ———. 2005. “Forms of sacralized power in Africa.” In The Character of Kingship, edited by Declan, Quigley, 25–37. Oxford: Berg Hocart, A. M. 1927. Kingship. London: Oxford University Press. ———. 1933. Progress of Man; a short survey of his evolution, his customs and his works. London: Metheun. ———. 1936. Kings and councilors: an essay in the comparative anatomy of human society. Chicago: University of Chicago Press. ———. 1952. The life-giving myth, and other essays. London: Methuen. ———. 1954. Social origins. London: Watts. Hofmayr, Wilhelm. 1911. “Religion der Shilluk.” Anthropos 6: 120–131. ———. 1925. Die Shiiluk. Geschichte, religion, und leben einze Niloten-Stammes. Mölding: Verlag der Administration des Anthropos. Howell, P. P. 1941. “The Shilluk settlement.” Sudan Notes and Records 24: 47–67. ———. 1952a. “The death and burial of Reth Dak wad Fadiet of the Shilluk,” Sudan Notes and Records 33: 156–65. ———. 1952b. “Observations on the Shilluk of the Upper Nile. The laws of homicide and the legal functions of the Reth.” Africa, 22: 97–119. ———. 1953a. “The election and installation of Reth Kur wad Fafiti of the Shilluk,” Sudan Notes and Records 34: 189–203. ———. 1953b. “Observations on the Shilluk of the Upper Nile. Customary law: marriage and the violation of rights in women.” Africa 23: 94–109. Howell, P. P. and W. P. G., Thomson 1946. “The death of a Reth of the Shilluk and the installation of his successor.” Sudan Notes and Records 27: 5–85. ———. 1952. “The death of Reth Dak wad Fadiet and the installation of his successor: a preliminary note.” Man 52: 102–4. Johnson, Douglas H. 2011. The root causes of Sudan’s civil wars. James Currey: Woodbridge and Rochester NY. Kapteijns, Lidwien, and Jay, Spaulding. 1982. “Precolonial trade between states in the Eastern Sudan, ca 1700 - ca 1900.” African Economic History 11: 29–62. Kantorowicz, E. H. 1957. The king’s two bodies: a study in medieval political theology. Princeton: Princeton University Press. Kunijwok, G. A. W. 1982. Government and community in a modern state: a case study of the Shilluk and their neighbors. Ph.D. thesis, Oxford: Bodleian library. Kurimoto, Eisei. 1992. “An ethnography of ‘bitterness’: cucumber and sacrifice reconsidered.” Journal of Religion in Africa 22 (1): 47–65. Lienhardt, R. Godfrey. 1952. “The Shilluk of the Upper Nile.” In African Worlds, edited by Daryl, Forde, 138–63. Oxford: Oxford University Press. ———. 1953. “Nilotic kings and their mothers’ kin.” Africa 25 (1): 29–42. ———. 1961. Divinity and experience: the religion of the Dinka. Oxford: Clarendon Press. ———. 1979. “Getting your own back: themes in Nilotic myth,” in Studies in Social Anthropology, edited by J. H. M., Beattie and R. G., Lienhardt, 213–37. Oxford: Clarendon Press. Mercer, Patricia. 1971. “Shilluk trade and politics from the mid-seventeenth century to 1861.” Journal of African History 12 (3): 407–426. Mogenson, Hanne. 2002. “The resilience of juok: confronting suffering in Eastern Uganda.” Africa 72 (2): 420–436 Muller, Jean-Claude. 1980. Le Roi bouc émissaire: pouvoir et rituel chez les Rukuba du Nigeria Central. Montreal: Serge Fleury. Munro, P. 1918. “Installation of the Ret of the Chol (King of the Shilluks).” In Pagan Tribes of the Nilotic Sudan, edited by C. G., and B. Z., Seligman, 541–547. London: Routledge. Nyaba, Peter Adwok. 2006. “The Chollo Predicament: the Threat of Physical Extermination and Cultural Extinction of a People.” http://www.dur.ac.uk/justin.willis/nyaba.htm, accessed 4/8/2007. Ogot, B. A. 1961. “The concept of Juok.” African Studies 20: 134–140. ———. 1964. “Kingship and statelessness among the Nilotes.” In The Historian in Tropical Africa, edited by Jan, Vansina, R., Mauny, L.-V., Thomas, 294–299. London: Oxford University Press. Oyler, D. S. 1918a. “Nikawng’s place in the Shilluk religion.” Sudan Notes and Records 1: 283–301. ———. 1918b. “Nikawng and the Shilluk migration.” Sudan Notes and Records 1: 107–15. ———. 1919. “The Shilluk’s belief in the Evil Eye.” Sudan Notes and Records 2: 122–137. ———. 1920a. “The Shilluk’s belief in the Good Medicine Man.” Sudan Notes and Records 3: 110–116. ———. 1920b. “The Shilluk peace ceremony.” Sudan Notes and Records 3: 296–299. ———. 1926. “Shilluk notes.” Sudan Notes and Records 9: 57–68. Paul, A. 1952. “The mar of the Shilluk.” Sudan Notes and Records 33: 165–166. Pumphreys, M. E. C. 1936. “Shilluk ‘royal’ language conventions.” Sudan Notes and Records, 20: 319–321. ———. 1941. “The Shilluk tribe.” Sudan Notes and Records 24: 1–46. Quigley, Declan. 1993. The interpretation of caste. Oxford: At the Clarendon Press. ———. 2000. “Scapegoats: the killing of kings and ordinary people.” Journal of the Royal Anthropological Institute 6 (2): 237–54. ———. 2005. “Introduction: the character of kingship.” In The Character of Kinghip, edited by Declan, Quigley, 1–23. Oxford: Berg. Ray, Benjamin C. 1991. Myth, ritual, and kingship in Buganda. New York: Oxford University Press. Riad, Mohamed. 1953. “Some observations of a fieldtrip among the Shilluk.” Wiener Volkerkundliche Miteilungen, 3: 70–78. ———. 1959. “The divine kingship of the Shilluk and its origin.” Archiv für Völkerkunde 14: 141–284. Richards, Audrey. 1968. “Keeping the king divine.” Proceedings of the Royal Anthropological Institute, 23–35. Sahlins, Marshall. 1981. “The stranger-king or Dumézil among the Fijians.” Journal of Pacific History 16: 107–132. ———. 1983. “Raw women, cooked men, and other “great things” of the Fijian Islands.” In The Ethnography of Cannibalism, edited by Paul, Brown and Donald, Tuzin, 72–93. Washington D.C.: Society for Psychological Anthropology. ———. 2008. “The stranger-king: or, elementary forms of the politics of life.” Indonesia and the Malay World 36 (105): 177–99. Schnepel, Burkhard. 1986. Five approaches to the theory of divine kingship and the kingship of the Shiluk of the Southern Sudan. Ph.D. thesis, Oxford: Bodelian Library. ———. 1988. “Shilluk royal ceremonies of death and installation,” Anthropos 83: 433–52. ———. 1990. “Shilluk kingship: power struggles and the question of succession,” Anthropos 85: 105–24. ———. 1991. “Continuity despite and through death: regicide and royal shrines among the Shilluk of Southern Sudan.” Africa 61 (1): 40–70. ———. 1995. Twinned beings: kings and effigies in Southern Sudan, East India and renaissance France. Göteborg: Institute for Advanced Studies in Social Anthropology. Scubla, Lucien. 2002. “Hocart and the royal road to anthropological understanding.” Social Anthropology 10 (3): 359–76. ———. 2003. “Roi sacré, victime sacrificielle et victime émissaire.” Revue du MAUSS 22: 197–221. Seligman, Charles Gabriel. 1911. “The cult of Nyakang and the divine kings of the Shilluk,” Fourth Report of the Welcome Tropical Research Laboratories, Volume B, General Science: 216–38. ———. 1931. “The religion of the Pagan Tribes of the White Nile.” Africa 4 (1): 1–21. ———. 1934. Egypt and negro Africa: a study in divine kingship. The 1933 Frazer Lecture. London: Routledge. Seligman, Charles Gabriel and Brenda Z. Seligman. 1932. “The Shilluk.” In Pagan Tribes of the Nilotic Sudan. London: Routledge, pp. 37–105. Simonse, Simon. 1992. Kings of disaster: dualism, centralism and the scapegoat king in the Southeastern Sudan. Leiden: Brill. ———. 2005. “Tragedy, ritual and power in Nilotic regicide: the regicidal dramas of the Eastern Nilotes of Sudan in contemporary perspective.” In The Character of Kinghip, edited by Declan, Quigley, 67–100. Oxford: Berg. Singer, André. 1975. The Shilluk of Southern Sudan. BBC, Disappearing World Series. Directed by Chris Curling. Anthropological consultants Paul Howell, Walter Kuniy Wok, and André Singer. 52 minutes Spaulding, Jay. 2007. The heroic age in Sinnar. Asmara: Red Sea Press. Thomson, W. P. G. 1948. “Further notes on the death of a reth of the Shilluk (1945).” Sudan Notes and Records 29: 151–60. Udal, John O. 1998. The Nile in darkness: conquest and exploration, 1504–1862. Norwich: Michael Russell. Wall, L. Lewis 1975. “Anuak politics, ecology, and the origins of Shilluk kingship.” Ethnology 15: 151–162. Westermann, Dietrich. 1911. Shilluk people: their language and folklore. Philadelphia: Board of Foreign Missions of the United Presbyterian Church of North America. Wilson, Monica. 1959. Divine kings and “the breath of men.” The 1959 Frazer Lecture. Cambridge: Cambridge University Press. Wrigley, Christopher. 1996. Kingship and state: the Buganda dynasty. Cambridge: Cambridge University Press. Young, M. 1964. “The divine kingship of the Jukun: a re-evaluation of some theories.” Africa 36: 135–52. La royauté divine des Shilluk : Violence, utopie, et condition humaine, ou éléments pour une archéologie de la souveraineté. Résumé : Depuis Frazer, la royauté Shilluk a été au cœur des débats anthropologiques sur la nature de la souveraineté. Or, ces débats ne devraient pas aujourd’hui être considérés comme non pertinents pour les discussions actuelles sur le sujet. Cet essai présente une analyse détaillée de l’histoire, du mythe et du rituel qui entourent l’institution Shilluk et propose de distinguer entre « royauté divine » d’une part (un moyen pour des humains de devenir dieux par l’usage de la violence arbitraire, et de défnir leurs victimes comme « le peuple »), et la « royauté sacrée » de l’autre (une domestication populaire de telles figures par le rituel). il est proposé que la royauté représente toujours une image temporaire et une solution imparfaite à ce qui est le dilemme fondamental de la condition humaine — et qui ne peut se maintenir que par la terreur. David Graeber’s original research project focused on relations between former nobles and former slaves in a rural community in Madagascar, focusing on magic as a tool of politics, the nature of power, character, and the meaning of history (Lost people: magic and the legacy of slavery in Madagascar, Indiana University Press, 2007). He has also worked extensively on value theory (Towards an anthropological theory of value: the false coin of our own dreams, Palgrave, 2001), and has recently completed a major research project on social movements dedicated to principles of direct democracy, direct action, and has written widely on the relation (real and potential) of anthropology and anarchism. He has recently published a book on the history of debt (Debt: the first 5,000 years, Melville House Publishing, 2011). ___________________ 1. I should note that “Shilluk” is an Arabization of the native term, Collo or Chollo. Most of the king’s current subjects now use Chollo when writing in English. I have kept to the historical usage largely to avoid confusion. 2. I am summarizing, not assessing, theories at this point so I will not enlarge on the fact that de Heusch seems to me to be working with a fundamentally mistaken idea of the nature of African fetishes, which are rarely embodiments of fertility but ordinarily embodiments of destructive forces (Graeber 2005). I think he is quite right and profoundly insightful when he argues that kings are often created by the same mechanisms as fetishes, as I have myself argued for Merina sovereigns (1996), mistaken when he goes on to claim that the key innovation here is that unlike fetishes the power of kings does not have to be constantly ritually maintained, since there are any number of counter-examples (e.g., Richards 1968). 3. Though part of the problem of course is that it is not entirely clear what to “be” a god would even mean. 4. Simonse’s comment has a particularly piquant irony when one considers the current popularity of the notion of “biopower”—the idea that modern states claim unique powers over life itself because they see themselves not just ruling over subjects, or citizens, but as administering the health and well-being of a biological population. Probably the question we should be asking is how it ever happened that there were governments that did not have such concerns. 5. I am simplifying their argument. Sovereign power for Hansen and Stepputat is marked not only by impunity but by a resultant transcendence—the “crucial marks of sovereign power” are “indivisibility, self-reference, and transcendence” (2005: 8), as well as a certain “excessive” quality. In many ways their argument, especially when it draws on that of Georges Bataille with his reflections on autonomy and violence, comes close to the one that I will be developing. But it is also exactly in this area that it deviates the most sharply, since Bataille’s position is ultimately profoundly reactionary, reading authoritarian political institutions back into the very nature of human desire. My position is more hopeful. 6. The Ganda kingship, for example, was almost entirely secular. Not only are we not dealing with a “divine king,” in the sense of one identified with supernatural beings, we are not even dealing with a particularly sacred one—except insofar as any king is, simply by virtue of hierarchical position, by definition sacred. 7. Benjamin himself suggested that popular fascination with the “great criminal” who “makes his own law” derives from precisely this recognition. 8. Though we should probably make note of the denouement: they elected a new king, who promptly accused them of murder and killed them all. 9. In a broader sense, he was doubtless aware that the colonial perception of Africa as a place of arbitrary violence and savagery had done much greater violence to Africans—that is, justified much worse atrocities—than any African king had ever done. This is the reason most contemporary Africanists tend to avoid these stories. But it seems to me there’s nothing to be gained by covering things up: especially since the actual arbitrary violence performed by most African kings was in fact, negligible or even completely imaginary (what mattered were the stories), and even those who even came close to living up to Euro-American stereotypes, like Shaka or Mutesa, killed far few of their own subjects than most European kings during the period before they became figureheads. 10. If nothing else one can say the question “do they really kill their kings” can be said to be definitively settled: though, at the same time, it is also clear that it is the least powerful of these figures who are the most likely to fall victim. 11. So today: an American citizen might be so little regarded by his own government that she is kicked out of hospitals while seriously ill or left to starve on the street; if, however, she then goes on to be killed by the agent of a foreign government, an American has been killed and it will be considered cause for war. 12. It’s also important to note here that, as Schnepel emphasizes (1991: 58), the Shilluk king was not himself a rainmaker: rather, he interceded on the part of his subjects with Nyikang, who was responsible for the rains. 13. As I mentioned earlier, the Egyptian Pharaoh may be the only example. Another is the Nepali king. But the latter case makes clear that identification with a deity is not is in itself, necessarily, an indicator of divine kingship in my sense of the term. The Nepali king is identified with vishnu, but this identification either originated or only came to be emphasized in the 19th century when the king lost most of his power to the Prime Ministers; it was, in fact, the token of what I’ve been calling sacred kingship, in which the king became too “set apart” from the world to actually govern. 14. Frost (1974: 187-188) suggests the institution might ultimately derive from military leaders referred to as bany, who at least among the neighboring Dinka, also have rain-making responsibilities. 15. The name is an Arabization of its real name, Pachod. It is, incidentally, not the same as the “Fashoda” of the famous “Fashoda crisis” that almost brought war between Britain and France in 1898, since “Fashoda” in this case is—however confusingly—an Arabization of the name of a rather desultory mercantile town called Kodok outside Shilluk territory to the north. 16. According to Spaulding (2006) the Funj, who were not Nilotic, both practiced ritual king-killing, and a similar marriage pattern, whereby royal wives moved back in such a way as to become the conduits between the capital and villages. 17. Among the eastern Nilotic societies considered by Simonse, the chief warrior age set was also responsible for representing the people against, and ultimately, if necessary, killing the king. Among the Shilluk this role seems to have been passed to royal women. 18. Actually, it is not entirely clear when Abudok ruled. Some genealogies leave her out entirely. Hofmayr places her before Tokot, and this has become the generally accepted version. Westermann (1912: 149-50) is ambiguous but seems to agree; however, his version also seems to make her the founder of Fashoda, which should place her closer to the time of Tugo, and elsewhere, in his list of kings (on page 135) he places Abudok after Tokot. Crazzolara (1950: 136n4) insists that she ruled after Tokot, as regent while Tugo was still a child. 19. They, not the fathers, remained in control of the offspring of such unions. Colonial sources (Seligman 1911: 218, Howell 1953b: 107-108) insisted that in the past, princesses who bore children would be executed along with the child’s father, 20. Another key medium for the spread of the cult of Nyikang appears to have been mediums loosely attached to the shrines, who had usually had no previous attachment to the court. According to Oyler (1918a: 288) these too were mainly women. 21. Seligman and Seligman (1932: 91) say there were two versions of how this happens: in one, the wives strangle the king themselves, in the other, they lay a white cloth across his face and knees as he lies asleep in the afternoon to indicate their judgment to the male Ororo who actually kill him. 22. Oyler says they acted as “magistrates” but their jurisdiction was limited to disputes between women (1926: 65-66). 23. Already in the 1840s foreign sources begin speaking of an annual tribute in cattle and grain, sometimes estimated at 10% (Frost 1974: 176). This seems however to have only been an early- to mid-19th century phenomenon. 24. One anomalous element has been eliminated: in this version the cord ran parallel to the earth; in most, it is arranged vertically. 25. So too, incidentally, with Vedic sacrifice, which reproduces the original creation of the world through the division of the body of a primordial being, or Greek sacrifice, which constantly recreated the divisions between gods, animals, and mortals, and so. All these religious traditions appear to be historically related. 26. In other versions, he traces back to a white or grey cow, created by God in the Nile. 27. Seligman (1931: 87-88) describes her as the embodiment of the totality of riverine creatures and phenomena, and notes that the priestesses who maintain royal shrines also maintain her cult. Offerings to her are left on the bank of the Nile. She is also the goddess of birth. When river creatures act in unusual ways, they are assumed to be acting as her vehicle; when land ones do the same they are assumed to be vehicles of Nyikang. 28. I will return to this point later. Of course, one could argue that this sort of behavior was considered legitimate in dealing with strangers: Shilluk were notorious raiders, and were in the 18th and 19th centuries apparently not above acts of treachery when dealing with Arabs or other foreigners in the “raiding country”—for instance, offering to ferry caravans across the Nile and then attacking, robbing or even massacring them. (At the same time foreigners who entered Shillukland itself were treated with scrupulous courtesy and guaranteed the safety of their persons and property.) Still, as we will see, ordinary Shilluk tended to rankle most of all at attempts to turn predatory violence into systematic power which is exactly what Nyikang was doing here. 29. Westermann (1912: 127-134) summarizes the origins of seventy-four different clans. If one discounts the three royal lineages included, and the six for whom no origin is given, we find that forty nine were descended from “servants” of Nyikang, six from “servants” of Dak, six of Odak, one of Tokot, and, most surprisingly, three from servants of Queen Abudok, the last royal figure to play this role—another testimony to her onetime importance. 30. However Seligman and Seligman (1932: 48) says kings would only take Ororo wives if they were “unusually attractive” since no child of an Ororo could ever become king. 31. In some versions, the Ororo men are responsible for killing the king “by surprise” if he is wounded in battle or grievously ill (Hofmayr 1925: 178-180), the women kill him otherwise. 32. This sort of behavior was occasionally noted even in colonial times. According to Howell and Thomson (1946: 76), there used to be ceremonial drums kept in Fashoda for royal funerals with special guardians, until reth Fafiti, annoyed that his predecessor had not used them to honor the previous reth, threw them in the Nile. 33. Similarly Schnepel: the ingenious application of violence was valued in itself—or, at least, valued insofar as it was seen to contribute to the “vitality” of the Shilluk nation as a whole. 34. Curiously, Evans-Pritchard (1948) ended up arguing exactly the opposite: that stories of ritual king-killing were the myth, and that in most cases one was really dealing with assassinations or rebellions. Mohammed Riad (1956: 171-177) however went through all existing historical information and could only find two examples of important rebellions in all Shilluk history, only one of which could really be called successful. of twenty six historical kings, he noted, fifteen “surely met their death in the ceremonial way” (ibid.: 176). Of the others, two were killed in war, three executed by the government in Khartoum, and six died of unknown causes. On the other hand he includes the four known cases of murder by rival princes as ceremonial deaths, which does rather muddy the picture. At least it makes clear this did happen, but only rarely. 35. On both sides: Hofmayr writes “at night he [the Reth] is awake and walks heavily armed around the village. His hand is full of spears and rifles. Whoever comes close to him is doomed” (1925: 175, in Schnepel 1991: 50). 36. Schnepel (1988) provides the best published blow-by-blow summary. What follows is drawn from my own reading of the standard primary sources: Munro 1918; Oyler 1918a; Hofmayr 1925; Howell and Thomson 1946, 1952; Thomson 1948; Howell 1952a, 1953a; Anonymous 1956; but also Riad 1959, who adds some telling details. All these seem to be derived from three ceremonies: the installations of Fafiti (1917), Anei (1944) and Dak (1946). 37. The name given the northern half of the country during the ritual, the south being Ghol Dhiang. It is interesting of course that the northern half should be named after Nyikang since this is the portion of the country Nyikang is said not to have conquered, but it is also where his effigy normally resides. 38. There is some confusion over when this ceremony takes place. Schnepel (1988) follows Howell and Thomson (1946) in placing it immediately after the election, but Riad (1959: 182) suggests the latter were describing an exceptional circumstance and that the funeral normally occurred well after the new reth’s installation. 39. Wendy James, one of the peer reviewers notes, not without justice, that this essay rather downplays the larger significance of the division of Shillukland into a Northern and Southern half, which has political as well as ritual implications. Invaders and powerful foreign influences (the Turks, the Mahdi, Arab traders, the Funj) invariably enter from the north, and this historical pattern is endlessly reproduced in the ritual. Fashoda of course stands precisely at the border between the two halves. 40. The Ororo who carry the king’s skeleton to its final resting place have a similar right to “seize small gifts and ransom from those unfortunate enough to cross their path” (Howell 1952: 160) and even those villages preparing gear for the ritual can do the same from anyone passing by at the time (Anonymous 1956: 99). But as we’ll see it is the effigies of Nyikang and Dak especially who are famous for this sort of thing. 41. For example, two months before the ceremonies begin, the effigy of Dak presides over an expedition to Fanyikang to obtain certain sacred ropes (Howell and Thomson 1946: 38). 42. Actually, Riad claims these fires are traditionally lit at the same time as the water ordeal—but in order to make the claim, he has to also argue that in former times, the king used to move back and forth between Fashoda and Debalo during his seclusion. Whether or not this is the case, the parallel he or more likely his informants are trying to draw here—between water in the North, and fire in the south—seems significant. Seligman (1934: 9) adds one of the three fires is transported to Fashoda as the “life token” of the king. 43. According to certain other versions he now commits incest with a half-sister, a very outrageous act. This is incidentally the closest the reth comes to committing one of de Heusch’s “exploits” and most sources do not even mention it. 44. All this is actually quite explicit: “as soon as the king dies, the spirit of Nyikang goes to his mother Nyikaya in the river, and the people will have to go to the river and bring him, and they will have to beg him to accept” (Singer in Schnepel 1988: 449). 45. Though in this case made even more confusing by reversing the order in the second example. If this is not simply a mistake on the author’s part, it could be taken as a telling sign of the reversibility of some of these hierarchies. 46. Actually Westermann claims this is the only prayer offered directly to God but Hofmayr (1925: 197-201) and Oyler (1918a: 283) both produce other ones (viz. Seligman 1934: 5). 47. God seems particularly immanent in violence or destruction. The above-cited prayer says “spear-thrusts are of Juok”, and one of the few ways that God was regularly invoked in common speech is when people call out “why, God?” when someone falls seriously ill. Among related Nilotic speakers in Uganda, “anything to do with killing must have juok in it” (Mogenson 2002: 424). On the other hand, in formal speech, God, so absent from the everyday life of ordinary Shilluk, pervades every aspect of royal existence. When speaking with members of the royal clan, one can never speak of their going someplace, or getting up, or staying someplace, or entering a house, instead they are “taken by God” to that place, “lifted by God,” “nursed by God,” “stuffed in the house by God,” and so on (Pumphreys 1937). 48. The presence of foreigners here—even if legendary—seems to be a reminder of the universality of the divine principle. Note too the opposition between this “fire ordeal” in which the candidate is chosen by God, and the “water ordeal” in which he is confirmed by Nyikang. 49. It happens so often that most such examples I actually purged from my account, above, to avoid monotony. 50. This is not to say that Nyikang’s passage does not include some acts of sacrifice, since otherwise there could be no feasts; only that this is not a particularly important aspect of what he does. With the king it is clearly otherwise. 51. In the absence of any detailed publish material on Shilluk sacrifice, I am drawing here on Evans-Pritchard (1954, 1956) on the Nuer, but even more Lienhardt’s work on the Dinka (1961) and Beidelman’s (1966a, 1981) reinterpretations of this material. 52. The Shilluk cognate appears to be pak, usually translated “praise,” which also refers to specialized formal language used within and between clans (see Crazzolara 1951: 140142). As usual, though, there isn’t enough material on Shilluk custom to make a sustained comparison. 53. Admittedly, I am relying here on Lienhardt’s detailed description and analysis of Dinka sacrifice, supplemented by Evans-Pritchard’s Nuer ethnography, but this is as I say because no parallel Shilluk account exists. For what it’s worth Evans-Pritchard (1954: 28) felt it appropriate to use Shilluk statements to throw light on Nuer practices. 54. “The thought was that the animal eats a bit here and there, but in the stomach it all becomes one mass. Even so the individuals of the two factions were to become one.” (Oyler 1920b: 299). 55. And also chyle, which is the further digested grass in the animal’s second stomach. This is the stuff even more closely identified with life but I thought I would spare the reader all the niceties of bovine digestive anatomy. 56. Or better put, the place where existential dilemmas are reduced to contradictions, so that they can be resolved. 57. Specially, kings were like prophets seen as being possessed by divine spirits (Shilluk prophets, when they appeared, were often possessed by Nyikang), mediated disputes on a national level that local authorities could not deal with, and relied on a following of young men who were themselves cut off from the ordinary domestic order because, having no access to cattle, they could not ordinarily expect to marry. 58. This is of course what “ideal” actually means: it is the idea lying behind some category. 59. Probably literally: Christopher Wrigley, the grand old man of Ganda studies, makes a plausible case that what we are dealing with here is a very old and probably fairly typical institution of sacred kinship suddenly transformed, a few generations before, into a state (1996: 246). A bureaucracy was superimposed with disastrous results. 60. No doubt some of this was simply to impress foreigners with the absoluteness of royal power; but such customs aren’t improvised whole cloth. 61. Or, sometimes they skip from description of monkeys, other sorts of animal behavior, or speculations about early hominids to wars between fully constituted nation-states. But generally there is nothing in between.

Bonhomme: The dangers of anonymity This work is licensed under the Creative Commons | © Julien Bonhomme. Attribution-NonCommercial-NoDerivs 3.0 Unported. ISSN 2049-1115 (Online) The dangers of anonymity: Witchcraft, rumor, and modernity in Africa Julien Bonhomme, École Normale Supérieure This article deals with a series of rumors that spread across West and Central Africa during the last two decades. These rumors of penis snatchers, of killer mobile phone numbers, and of deadly alms constitute a transnational genre that is characteristic of Africa’s occult modernity. While the literature on the modernity of witchcraft has been criticized for its macrosociological orientation, the article strives to counterbalance this bias by drawing on microsociology in order to explore the interactional repertoires in which these new forms of the occult are grounded. It shows that they exploit anxieties born out of mundane situations: shaking hands with strangers, receiving unidentified phone calls, or accepting anonymous gifts. New forms of the occult thus focus on the dangers of anonymity and point to the risk of being forced into opaque interactions with unknown others. They draw on two different situations of anonymity, which can be connected to two distinctive repertoires of modernity. Face-to-face encounters with strangers are typical of—but not exclusive to—urban modernity, while mediated interactions with distant and often invisible agents are part and parcel of technological modernity. Therefore, insofar as modernity has extended the scope of human sociality in unprecedented ways, it has extended as well the scope of the occult. This article casts new light on witchcraft and the occult in contemporary Africa, and suggests new ways of tying together micro and macro levels of analysis, by grounding the wide-ranging dynamics of modernity in the minutiae of human interaction. Keywords: modernity, witchcraft, rumor, cities, anonymity, Africa Consider the following three incidents: 1 In June 2010, in Port-Gentil, the second largest city of Gabon, Élie, a bricklayer in his thirties, is looking for his apprentice. Since he cannot find him, he approaches a young boy, shakes hands with him and offers him the job. Junior declines, but asks his younger brother to help the bricklayer. As Élie is about to leave with his new assistant, he is grabbed by Prince, a friend of Junior, and taken to a nearby house. There, Junior bluntly accuses Élie of the theft of his genitals: just after shaking hands with him, he felt an electric shock in his groin and suddenly realized his penis had shrunk. Élie denies the accusation, the discussion becomes heated, and a crowd gathers around the house. The unfortunate bricklayer is tied naked to an electric post and brutalized for hours by a small group of locals. His ex-wife, who happens to be passing, tries to intervene, but in vain. She calls the police and they finally rescue Élie, who is by now in a poor condition. A few weeks later, the two main culprits, Junior and Prince, are given a fine and a three-month suspended sentence. In January 2010, in Thiaroye, a densely populated suburb of Dakar in Senegal, Mustafa, a retired agronomist, goes to the market near the railway station with the intention of giving alms to the poor in the form of goat’s meat. He spots two women at their stalls, greets them and offers them chunks of meat wrapped in black plastic bags. The women decline even after being pressed to accept this “sacrifice” in charity. Disappointed, Mustafa decides to buy some fruit before going home. But he soon realizes that a hostile crowd is now gathering to his side of the market, after the women have raised an alarm, suspecting that the alms he has offered them will kill them. Mustafa catches sight of a policeman near the railway track and asks for his protection. The policeman tries to escort him to the police station, but a crowd numbering in the hundreds rapidly overwhelms the two men. While stones are being thrown at them, they have to take refuge in a community building in the neighborhood. Young men manage to break into the premises and start to assault Mustafa. Police reinforcements come to his rescue and take him to the police station. Mustafa cannot go home until sunset, when the crowd waiting for him in front of the building finally disperses. Though very much shaken, Mustafa decides not to file a complaint. In July 2004, in Calabar, the capital city of Cross River State in Nigeria, Vincent, a sixteen-year-old boy, hears his sister’s mobile phone ringing; she is busy and cannot answer the phone herself. When he looks at the screen, he sees “a number and not a name.” After some hesitation, he answers the call. When he puts the phone to his ear, he hears a strange noise. He says hello, then hears a man coughing and suddenly loses consciousness. His sister Rachel rushes and picks up the phone, only to collapse, too, after saying hello. When the brother and sister regain consciousness, they find they are surrounded by family members and neighbors, who are praying for their recovery. A few hours later, a newspaper journalist comes to investigate the incident. He decides to call back the alleged “killer number,” stored on Rachel’s phone. The following conversation ensues: Journalist. Hello, who am I speaking to? (A female voice answers.) Woman. How can you call my number and ask me, “who am I speaking to?” Journalist. This number is said to have been used to call two persons who collapsed. Then the woman hangs up. The journalist calls back several times, but his calls are left unanswered. When he tries again from a phone center, the woman finally picks up his call. Journalist. I expect you to put up a defense on whether or not your number sends people to sleep. Is it true that people collapse after receiving a call from your line? Woman. I cover you with the blood of Jesus! I say, I cover you with the blood of Jesus! Journalist. Thank you for that prayer. But when will I collapse? (The woman hangs up on him.) What are we to make of these three strange episodes? Is it even appropriate to group these seemingly unrelated events together? In a review article on recent academic literature on African witchcraft and sorcery, Terence Ranger (2007) calls into question the validity of the category of “the occult,” prompting further debate on the topic in the columns of Africa (ter Haar and Ellis 2009; Meyer 2009). Ranger criticizes anthropologists such as Jean and John Comaroff (1999) for an “over-generalization of the occult” (Ranger 2007: 274). Instead of lumping together incommensurable phenomena, he suggests, “we need to disaggregate and to historicize” (ibid.: 277). In this view, the misadventures of Élie, Mustafa, and Vincent should rather be regarded as three independent events involving three distinct forms of the occult in three different African countries, each one deserving to be studied in its own particular context. And yet they have much in common, as I will demonstrate. Our recognition of the limitations of the term “occult” should not lead us to “retreat into the study of the particular,” as Birgit Meyer (2009: 413) convincingly argues in her comments on Ranger’s essay. From this perspective, the objective of this article is to put forward an analytical framework in which the peculiar stories of Élie, Mustafa, and Vincent can be fully compared and may even illuminate each other. Rumors and the modernity of witchcraft The mishaps of Élie, Mustafa, and Vincent are not isolated incidents. All three are closely linked to rumors that were already circulating. Stories of penis snatchers, of deadly alms, and of killer mobile phone numbers were circulating wildly in Gabon, Senegal, and Nigeria at the time of these incidents. Far from being parochial gossip, these diverse rumors are closely interrelated, large-scale phenomena. Of the three, penis snatching is perhaps the most archetypal or paradigmatic. It is the most widespread of the rumors (both in space and time) and the one to which other rumors are constantly compared. The origins of penis snatching can be traced to early-1970s Nigeria (Bonhomme 2009). During the two following decades, it was mostly confined to Nigeria and Cameroon. However, in the 1990s, the rumor spread to most of West and Central Africa, from Mauritania to Sudan and the Democratic Republic of Congo. The rumor has been sporadically active ever since. The scenario is always broadly similar: strangers are accused of stealing (or sometimes only shrinking) the genitals of other people during a public encounter, often just a simple handshake. Though sporadic, penis snatching has left its mark on the popular imagination in Africa. Stories of penis snatching are covered extensively in the news and public accusations often lead to instant justice and mob violence. For instance, about twenty people accused of penis snatching were lynched in April 2001 in Nigeria, and another dozen in January 2002 in Ghana. Rumors of killer mobile phone numbers also originated in Nigeria as early as July 2004 (Bonhomme 2011). It was said that receiving calls from certain phone numbers could instantly kill those who took the call, with blood pouring out of the mouth, nose, and ears. This new rumor was immediately compared to penis-snatching stories, which were still in circulation at the time (Agbu 2004). From Nigeria, the rumor quickly spread to Cameroon and Gabon within the same month. It mysteriously reappeared in India two years later and, from there, spread to several countries in South Asia and eventually came back to Africa.2 The rumor was present in Ghana and Angola in 2008 and again swept through Africa in 2010, from Kenya to Mali. Compared to the other two, the rumor about deadly alms is the most recent, the most transient and the least widespread. There has been only one short wave of this rumor, at least until now. It appeared in Senegal in January 2010, in Dakar or Saint-Louis, and quickly spread throughout the country and to Gambia and the Malian border before disappearing within a week or two. During this short time, the rumor was nevertheless hot news and triggered many incidents involving public accusations and violence, even though nobody was actually killed. According to the rumor, people had died mysteriously after a stranger driving an SUV had given them alms. Just like killer phone numbers in Nigeria, deadly alms were systematically compared to penis snatching, which has repeatedly broken out in Senegal since the mid-2000s. Many similar rumors concerning, for instance, ritual murders and the trafficking of body parts, could be compared to penis snatching, killer phone numbers, and deadly alms. All of them, which have been circulating in Africa since the 1990s (at least), belong, I suggest, to the same transnational genre.3 The way the rumors circulate gives the genre its cohesion. They are “stories that travel” (des histoires qui tournent), according to a Senegalese friend. New stories constantly appear, drawing on older stories, while rumors that were believed to have disappeared sporadically reappear. Between different versions of the same rumor or even between different rumors, the plot and details might vary but a family resemblance remains. These rumors are not mere tales. They are told as the personal accounts of witnesses and victims. They also cause real incidents, such as those involving Élie, Mustafa, and Vincent (the first two told from the perspective of the persons accused of occult wrongdoings, the third from the perspective of the person subjected to such wrongdoings).4 People collapse after receiving unidentified phone calls or shaking hands with strangers, while others are accused and lynched. Like many rumors of this kind, they are self-fulfilling prophecies. The frantic spread of the rumor in a particular locality creates a climate of fear, or even panic, which makes phone calls, handshakes, or alms suddenly appear suspicious to many. And this suspicion inevitably triggers incidents, which by a sort of snowball effect, impart new vigor to the rumor’s circulation. Rumors are circulated by word-of-mouth at home, at work, at the market, on public transport, or in churches and mosques. In Nigeria for instance, some churches warned their followers against phone calls from “satanic numbers.” Mobile phones and text messages were also used to spread rumors, including, ironically, stories about the dangers of receiving phone calls. But these rumors are not an exclusively oral genre. They are also given extensive media coverage, in newspapers, on the radio, on television, as well as in digital media. Even though the media did not invent them, these rumors would not have circulated so widely without the media. There is indeed a correlation between the time when the rumors appeared and the way in which they circulate. In many African countries, notably in Francophone ones, the 1990s saw the introduction of the freedom of the press and thus coincided with the rapid emergence of a popular press, which often draws on rumors and “factoids.” This genre of press cashes in on stories about witchcraft and the occult (Bastian 2001). Penis snatching, killer phone numbers, and deadly alms thus come to represent witchcraft in the media age. The media are not only playing an active role in the transnational circulation of rumors, they also contribute to the social perception of the phenomenon—this is why media reports are essential materials if we are to study these rumors to their fullest extent.5 Neither purely oral nor purely written, this genre of rumor therefore implies a constant back and forth between word-of-mouth and the mass media, a hybrid genre of communication often called radio-trottoir (“pavement radio”) or radio-cancan (“radio gossip”) in Francophone Africa (Ellis 1989). This genre of rumor maintains complex and ambiguous relations with traditional interpretations of misfortune in terms of witchcraft. In their lively discussions about the rumors, local people usually did not agree about how to categorize these phenomena: are they entirely new or do they fit into the old categories of witchcraft? For instance, in Senegal, penis snatching has been regarded either as an unusual form of cannibal witchcraft (dëmm in Wolof), as a spell of impotence (xala) cast by a marabout-sorcerer, as the malevolent trick of a spirit (jinne), as a somewhat special kind of theft, as a psychosomatic syndrome, as a fraud to extort money, or even as a simple hoax.6 In the same manner, deadly alms have usually been thought of as a new form of sorcery (liggéey). Likewise, in Nigeria, killer phone numbers have been said to proceed from a combination of old magic and new technology: “juju [magic power or artifact] has been merged with the technology of GSM to produce instant effectiveness” (Vanguard, August 20, 2004). In Cameroon, it was said to be “tele-witchcraft” (télé-sorcellerie) and an article posted on an Internet forum was titled: “Africa ‘digitalizes’ witchcraft” (L’Afrique ‘digitalise’ la sorcellerie). Attitudes toward these phenomena are also diverse and range from absolute conviction to corrosive disbelief, such as this Gabonese woman who declared about penis snatching: “Now even impotent men are going to claim their penis has been stolen.” However most people remain baffled about what to think, such as this Gabonese fireman: “Sometimes we are led to believe it, sometimes we wonder how this could really happen. I still don’t have a personal opinion on this issue. The media say there is some evidence. However, they haven’t shown us a man without a penis yet. Anyway, I believe magic is behind all this.” These transnational rumors bear witness to the dynamics of the reconfiguration of the occult in contemporary Africa. They bespeak Africa’s occult modernity. Scholars like Jean and John Comaroff (1993, 1999) and Peter Geschiere (1997) have quite rightly stressed the irreducible modernity of witchcraft. In the 1990s and 2000s, the twin expressions “modernity of witchcraft” and “witchcraft of modernity” have thus become the master tropes of academic discourses on the occult in Africa (see Ciekawy and Geschiere 1998; Bernault and Tonda 2000; Moore and Sanders 2001; Meyer and Pels 2003; West and Sanders 2003, to cite just a few collective works). Max Weber’s Eurocentric conception of modernity as an unequivocal process of “disenchantment of the world” has proved to be too simplistic and the emphasis has instead been placed on the existence of “alternative modernities,” following Arjun Appadurai (1996). In this view, witchcraft, magic, and the occult are no longer regarded solely as a traditional heritage, but rather as inherently modern phenomena, which must be set back in the context of global flows of symbols and images—what Appadurai aptly calls “mediascapes.” Discourses about witchcraft are shaped by globalized imaginaries and spread through modern means of communication. They thrive in the most modern sectors of society, such as State politics (Rowlands and Warnier 1988) or the capitalist economy (Comaroff and Comaroff 1999). Therefore, witchcraft is not only part of modernity; it is about modernity. African popular discourses about witchcraft are but “a metacommentary on the deeply ambivalent project of modernity” (Sanders 1999: 128). And witches are “modernity’s prototypical malcontents,” inasmuch as they embody the contradictions of the experience of modernity (Comaroff and Comaroff 1993: xxix). The diagnosis is undoubtedly relevant and rumors of penis snatching, killer phone numbers, and deadly alms offerings could easily fit into such an explanatory scheme. Yet the “witchcraft and modernity” paradigm has been justifiably criticized for its analytic uncertainties (Sanders 2003). The argument about the link between witchcraft and modernity remains “suggestive, not demonstrative,” as Sally Falk Moore (1999: 305) puts it in her comments on the Comaroffs’ lecture on occult economies. It amounts indeed to turning “general context into particular explanation” (ibid.: 306). Studies of the witchcraft of modernity often rest upon a somewhat loose articulation between the local (the phenomenon under study) and the global (which serves as an explanatory framework). Hence the abstraction of their conclusions: new forms of witchcraft are said to express popular discontent with—or at least ambivalence toward—modernity, globalization, “millennial capitalism,” or the “culture of neo-liberalism.” Even though this interpretation may be true, the danger is that we may move away from fine-grained ethnographic analysis toward a metanarrative based on macrosociological abstractions with a blend of cultural relativism (Englund and Leach 2000). The Comaroffs themselves are fully aware of these analytic pitfalls: at the end of their lecture, they observe that tropes such as globalization (to which we could add modernity, capitalism, or individualism), like all catchwords and clichés, [they] are cheapened by overuse and underspecification, by confusing an expansive metaphor for an explanatory term. As a result, much of what is currently being written about them in the social sciences is Anthropology Lite, fact-free ethnography whose realities are more virtual than its virtues are real. (1999: 294) This paper critically engages with the literature on witchcraft and modernity and argues that the macrosociological concept of modernity will remain devoid of any relevance, unless it is more finely particularized and broken down into interactional repertoires open to ethnographic scrutiny. This matter is first and foremost an issue of scale. What could be the most relevant unit of analysis to “make sense of the enchantments of modernity” (ibid.: 279)? Penis snatching and other occult rumors, like the sale of body parts mentioned by the Comaroffs, are “at once profoundly parochial and so obviously translocal” (ibid.: 282). They thus raise the issue of “doing ethnography on an awkward scale, neither unambiguously ‘local’ nor obviously ‘global’—but on a scale in between that, somehow, captures their mutual determinations” (ibid.). But how exactly are we to situate these phenomena in larger-scale perspectives without yielding to Anthropology Lite and fact-free ethnography? And precisely what kind of analytical scale could capture the dialectical interplay between the local and the global? In the scope of this article, I would like to propose an experiment in imaginative sociology, by putting forward a possible way of tying together micro and macrolevels of analysis, in order to cast new light on Africa’s occult modernity. I suggest counterbalancing the macrosociological bias of the “witchcraft and modernity” paradigm by drawing on microsociology, in order to explore more carefully the interactional repertoires in which new forms of the occult are grounded. How do the wide-ranging dynamics of modernity translate into the basic substance of everyday social interaction on a smaller scale? This resort to microsociology does not mean reducing global phenomena to narrow circumstances and retreating into parochial ethnography. On the contrary, local circumstances may presumably play a part in the appearance or reappearance of a rumor at a given time, in a given place, and in the occurrence of related incidents. Yet the rumors’ transnational scope proves that they are by no means reducible to these local circumstances. Narratives of penis snatching, for instance, have remained strikingly stable over forty years in about twenty different countries. Different incidents taking place in Dakar, Libreville, or Lagos proceed, I suggest, from the recurrence of the same local situations. In this view, the microsociological focus on interactional repertoires allows us to abstract the general from the particular: local situations, as a result, acquire a more general relevance in helping to understand phenomena on a wider scale. As we will see, rumors of penis snatching, of killer phone numbers, and of deadly alms all revolve around anonymous interactions that turn out to be fatal. These new forms of the occult focus on the dangers of anonymity, which is precisely an interactional repertoire typical, and even constitutive, of modernity. In what follows, I will briefly review each of the three rumors and concentrate on a series of incidents, before wrapping up my argument on the dangers of anonymity in a concluding section.7 Shaking hands with strangers Élie’s misadventure is representative of the circumstances in which accusations of penis snatching occur. First, penis snatching is an exclusively urban phenomenon. New forms of the occult are often associated with urban life. Witchcraft phenomena—violence by and against alleged witches—represent but one aspect of multifarious urban insecurity (Ashforth 2005). Cities are insecure places, partly because they are spaces of anonymity. Penis snatching is indeed closely connected to urban anonymity.8 Incidents occur in public settings, but never in the intimacy of the home. They happen in crowded spaces, such as streets, markets, minibus taxis, or around mosques. In Lagos, the first “cases” of penis snatching in 1975 took place around markets in densely populated neighborhoods. It is unsurprising that incidents related to penis snatching (as well as to deadly alms) often occur in markets. Markets are often considered to be dangerous places where witches roam (Bastian 1998). They are also places where all sorts of rumors circulate and where the slightest incident can degenerate into mob violence. Moreover, penis snatching always involves total strangers: the alleged thief and the victim are unrelated to each other. “My tormenter has no past history with me!” cries out a distraught Congolese man, who cannot understand why he was singled out in the anonymous crowd by a penis snatcher. In this regard, penis snatching can be compared to two other contemporary rumors. The first appeared in Cameroon in 1984 and spread to Gabon and Congo the following year, before reappearing in Cameroon and Gabon in 2007. Stories were told of people dying mysteriously after an old lady begged them for a cup of water. Only those who offered her water succumbed shortly after; those who bluntly refused her gesture of hospitality (or who gave her undrinkable salty water instead) remained safe. Each time the rumor triggered incidents: women were lynched because they were mistaken for the “Killer Granny” (Mamie tueuse) as she was sometimes nicknamed. A similar rumor circulated in Senegal in 1990: an old lady attacked people in the streets and savagely bit their necks (in contrast with the other rumors, in this case, the aggression is direct and not hidden behind innocent gestures). She was nicknamed Mère Mataté (from màttat, “to bite” in Wolof). Several women accused of being the “Biting Mammy” were assaulted, just because they were old, had long teeth, or were said to stare strangely at passersby. Penis snatchers, the Killer Granny, and the Biting Mammy are all anonymous figures and, consequently, incidents occur on the occasion of mundane interactions between complete strangers.9 Because they focus on public encounters between strangers, these rumors are intimately linked to urban sociality. Anonymity is indeed a defining feature of modern urban life, as Georg Simmel’s pioneering work “Metropolis and mental life” (1950b: 409–24) rightly stressed. In contrast to village life, large cities are characterized by countless possibilities for encounters with strangers. Urban public space is a “world of strangers,” where they have become the norm rather than the exception (Lofland 1973). In public settings, “traffic relationships” are the interactional repertoire par excellence (Goffman 1963). These relationships suppose only minimal mutual acknowledgement between strangers, who routinely deal with each other as “nonpersons.” In this view, accosting someone on the street represents a potentially hazardous or at least uncertain event. “The sheer numbers of people around the urbanite makes it impossible for him or her to feel secure about the motivations and interests of others. Strangers do not, as in the rural areas, come from outside. They live next door or even in the next room,” as Misty Bastian (2001: 75) observed in a paper about modern witchcraft in Nigeria. From this perspective, penis snatching appears as a dramatic illustration of the microhazards of anonymous encounters, to which African urbanites are constantly exposed. It is the witchcraft of traffic relationships and accosting people on the street. The very circumstances of the anonymous encounters that are misinterpreted as penis snatching confirm this interpretation. In 2004, a Nigerian civil servant passing through Kano was accused of penis snatching and lynched to death after asking a local for directions. In a similar fashion, in 2006, on a street in Bobo-Dioulasso (Burkina Faso), a preacher was accused of penis snatching after accosting a young man to preach him the gospel. Penis snatching always requires contact. It can be a fortuitous physical contact in a crowded setting: a benign collision with a stranger misunderstood as an occult aggression. In 2001, on a street in Libreville, a young man was inadvertently brushed against by a Nigerian man and suddenly felt an electric shock. He thought he heard the stranger say, “You got it?” Then he felt his penis shrinking. He raised the alarm. Passersby caught the man and beat him so badly that he “confessed.” But snatching can also come from simple eye contact. In 2006, in a neighborhood of Parakou (Benin), a young man stared at a girl. She immediately accused him of stealing her genitals (a few cases of genital—or breast—snatching in fact concern women).10 Locals lynched the young man. Though a shared gaze usually expresses an intention to communicate, a too insistent stare between strangers triggers uneasiness and can even be perceived as hostility. This is clearly what happened in Parakou: the girl misconstrued the young man’s enticing gaze as an occult aggression. The most typical circumstances in which penis snatching occurs are handshakes between strangers. “A man I did not know shrank my penis when he shook hands with me after he asked me what time it was,” complained one Gabonese victim of penis snatching. “I was passing by and I saw this man who held out his hand to greet me. I did not know him but, as a Cameroonian, I shook hands with him,” said a man involved in a similar incident. In the popular imagination, penis snatching is commonly viewed as the occult side of the anonymous handshake. It is not surprising, given the social importance of greetings in Africa (see, for instance, Irvine 1974; Akindele 1990). Greetings are access rituals that “ritually regularize the risks and opportunities face-to-face talk provides” (Goffman 1981: 19). Greetings between strangers are therefore more hazardous, especially in potentially hostile environments (Youssouf, Grimshaw, and Bird 1976). This is precisely what is at stake in the case of penis snatching. The anonymous handshake, whose initiator is always the alleged penis snatcher, is perceived by the other party as a threatening gesture. Writing about rural France, where handshakes can also trigger suspicions of witchcraft, Jeanne Favret-Saada points out that shaking hands is “such an ordinary gesture of recognition that one usually forgets what is involved” (1980: 114). The lesson clearly applies to penis snatching as well. Compared with accidental collisions or eye contact, handshakes represent the most paradoxical situation of penis snatching, since the aggression is disguised as its very opposite, namely solidarity and friendship. A symmetric interaction is turned into asymmetric predation. “‘Hello,’ a sign of friendship and fraternity between men, has now become a source of misfortune,” as one Gabonese journalist put it (L’Union, June 18, 1997). As a consequence, all anonymous handshakes suddenly become suspicious and are to be avoided. Public warnings are broadcast in newspapers, on the radio, on television, in churches, and in mosques.11 “Keep your hands in your pockets. Avoid any unexpected handshake. Your friends are maybe not the ones you think they are,” a Gabonese journalist advocates. (L’Union, June 18, 1997). “Inhabitants of Port-Gentil make sure they do not brush past strangers. As far as handshakes are concerned, from now on they are proscribed, except between close relatives or old friends,” adds another (Agence France Presse, October 17, 2005). In Sudan, another journalist gave the following advice: “I consider it my duty to warn anyone who wants to come to Sudan to refrain from shaking hands with a dark-skinned man. Since most Sudanese are dark-skinned, he had better avoid shaking hands with anyone he doesn’t know.”12 In Niger, even religious communitas is undermined by the rumor: “Now the faithful waver over whether they should shake hands after prayer, as recommended by the Sunna of the Prophet” (Agence de Presse Africaine, March 23, 2007). The rumor, as short-lived as it may be, thus raises a critical issue: how should people behave with strangers? All local comments about penis snatching indeed stress the opposition between acquaintances and strangers and dramatize the dangers of anonymity. They advocate maintaining distance and vigilance toward strangers and sometimes even pre-emptive aggression—a suggestion prone to cause new incidents. In 1990, in Lagos, “it was thought that inattention and a weak will facilitated the taking of the penis or breasts. Vigilance and anticipatory aggression were thought to be a good prophylaxis” (Ilechukwu 1992: 96). In a striking manner, this calls to mind the “slight aversion” that regulates urban life, according to Simmel: “Outer reserve is not only indifference but, more often than we are aware, it is a slight aversion, a mutual strangeness and repulsion, which will break into hatred and fight at the moment of a closer contact” (Simmel 1950b: 415–16). And yet distance and reserve are social norms typical of Western cities, but not of African ones. Urban life as a universal phenomenon inevitably generates anonymity; but this does not mean that anonymity is lived and perceived in the same manner everywhere, for instance in Paris and in Dakar. The observing of “civil indifference” in Western cities aims to keep strangers in a state of anonymity and thus reasserts the crucial social distinction between public and private spheres; but the same distinction does not hold in most African cities where contacts between strangers in public settings are not to be avoided. On the contrary, shaking hands with strangers is perfectly common. Conversely, distance and coldness are very negatively perceived as behaviors typical of white people. For instance, in Brazzaville, “it is impossible to stay in a bus or a cab without being talked to…. Urbanites feel uncomfortable remaining indifferent to the presence of others, be they total strangers” (Milandou 1997: 124). Rumors of penis snatching, therefore, appear all the more striking. They elicit, at least temporarily, avoidance behaviors that would otherwise be largely disapproved of. The rumor thus calls into question the everyday norms of urban life. Max Gluckman (1972: 2) once observed that situations of “moral crisis,” in which people’s behavior is torn between contradictory social norms or values, are a hotbed for witchcraft accusations. In my view, penis snatching brings about such a situation of moral crisis (rather than it being the consequence of a preexisting crisis). It puts urban sociality to the test (just as rumors about the Killer Granny put hospitality toward strangers to the test). Rumors indeed reveal the tensions and contradictions underlying uncertain social situations—as deadly alms and killer phone numbers will confirm. Exposed to the hazards of anonymity, urban sociality appears to be torn between contact and distance. The personal account of a Togolese man, interviewed for a radio program on penis snatching in 2005, perfectly illustrates this tension. The man said he felt confused because he did not know how to behave in the street anymore; he was torn between his resolution not to shake hands with strangers for fear of penis snatchers and the social obligation to greet others for fear of being overly rude. However the instruction to avoid strangers cannot long withstand the strong social preference for contact. As soon as the rumor vanishes life returns to normal and strangers start shaking hands again. If penis snatchers are always unrelated to their victims, accusations tend to target specific categories of strangers. In Nigeria, people accused of penis snatching are often Hausas.13 This is also the case in other countries, such as Chad, Senegal, Mali, or Ivory Coast. However, outside Nigeria accusations are prone to switch from Hausas to Nigerians in general. In Togo for instance, in 2005, the Nigerian ambassador was compelled to call a press conference to deny rumors that his fellow citizens were stealing the genitals of Togolese men. As the rumor moves away from Nigeria, accusations are extended from specific subgroups to the larger group that subsumes them. In Gabon, they target not only Nigerians, but also more broadly West African immigrants. “This trick most certainly comes from West Africa,” claims a Gabonese. “Let’s deal with foreigners who come to our country and mess up everything,” adds another man. Accusations are thus fueled by xenophobia and build on preexisting tensions and often a past history of communal violence. They usually lead to even greater violence and stronger collective mobilization of “us” against “them.” Groups rather than single individuals are targeted and lynchings turn into riots. For instance, when penis snatching hit Cotonou in 2001, accusations targeted Ibos, a Nigerian ethnic group present in large numbers in Benin where they often work as traders. Several Ibos were lynched and collective violence soon degenerated into rioting: the market where Ibos usually worked was looted and burned. As the accusation patterns show, the abstract figure of the unknown other easily leads to specific categories of strangers. Even though unrelated people cannot rely on biographical information to frame their interaction, they use perceptual clues (such as clothes, physical complexion, or language) to identify others and categorize them according to social stereotypes (such as social status, ethnicity, or nationality). In large cities, such categorizing modes of thought are by necessity the most important way of identifying others. “The city created a new kind of human being—the cosmopolitan—who was able, as his tribal ancestors were not, to relate to others in the new ways that city living made not only possible but necessary. The cosmopolitan did not lose the capacity for knowing others personally. But he gained the capacity for knowing others only categorically” (Lofland 1973: 177). This is why ethnicity has not lost its relevance in African cities: ethnic diversity often leads urbanites to categorize others spontaneously depending on their supposed ethnicity (Shack and Skinner 1979; Mitchell 1987). The importance of ethnic stereotypes inferred from perceptual clues is manifest in accusations of penis snatching. Hausas are often singled out by their long, flowing robes (boubou): the inadvertent brushing of the robe against a passerby is enough to raise the alarm against its wearer. In Mali and Ivory Coast, Fulanis “who look like Hausas” or “boubou-wearing Sahelians” have also been accused of penis snatching.14 This illustrates how ethnic or regional stereotypes are used to single out penis snatchers in a world of strangers. Penis snatchers are thus, following Simmel’s well-known conception of the Stranger (1950a: 402–8), those who are at the same time familiar and foreign, spatially near but socially far, such as Hausa traders in West Africa or West African immigrants in Gabon. The relational tension that shapes everyday interactions with those familiar strangers helps to explain why they are targeted. The rumor, therefore, bears witness to the “shifting ethnoscapes” (Appadurai 1996: 34) of African cosmopolitanism. It is intimately related to the two defining features of modern urban life. The figure of the stranger is associated both with urban anonymity and with ethnic diversity, the latter being a common feature of all major African cities since the colonial period. Even though urbanization may be a global social phenomenon, the forms of urban sociality, however, differ in Shanghai, São Paulo, New York, or Lagos. Rumors of penis snatching precisely illuminate one aspect of Africa’s own version of urban sociality: how African urbanites experience the tension between distance and proximity, between strangeness and familiarity in their daily lives. “The making of a new Africa lies in the city, for better or for worse,” announced Georges Balandier decades ago (1985: vii). African cities are indeed places of experimentation for new social relations (Simone 2004). The rumor reveals all the ambiguity of this burgeoning urban life and highlights its occult side. Anonymous gifts Rumors of deadly alms have much to do with penis snatching, even though they have been much more limited in space and time (at least until now). Both rumors are intimately related to anonymity, but they focus on distinct social occasions: on handshakes between strangers for penis snatching, on anonymous charity for deadly alms. When they appeared in Senegal in 2010, rumors about deadly alms (frequently called offrande de la mort in local French) were based on the following scenario: “An SUV driven by two men—one of whom was said to wear a turban—would ‘generously’ distribute meat, ten thousand CFA francs and percale as alms to passers-by. Unfortunately ‘all those who accepted this charity had a sudden attack and passed away’” (L’Observateur, January 26, 2010). The content of the “deadly package” (as it was sometimes called) varies in some narratives, but meat and money are by far the most frequently cited items. Though not always mentioned, the presence of percale (white cotton cloth) has an obvious symbolic meaning: it is traditionally used for Muslim shrouds and thus refers to death. Narratives most often enumerate the items given, but do not specify the almsgiver’s identity. About the “mysterious charity killer,” only two things can be hinted: he is rich and he is a stranger. His wealth is suggested by the fact that he drives an SUV and gives out ten thousand CFA franc banknotes, the highest denomination and a highly unusual sum for charity. This generosity is too extraordinary not to appear suspicious. In some incidents, conspicuous wealth plays a direct role in the accusation. A Senegalese man who lived and worked in Europe was back in his hometown of Rufisque where he planned to build some houses. As soon as he arrived on the building site driving a big black SUV he started distributing money around him. His gesture was motivated by the desire to ingratiate himself with the locals in order not to jeopardize his real estate project. Unfortunately, recipients misinterpreted this generosity: they accused him of being the deadly almsgiver and immediately assaulted him. Policemen eventually rescued the man and escorted him to the police station, but the crowd followed them and tried to break into the police station to seize the almsgiver. Wealth is a central aspect of the rumor and this theme recurs in all local discussions about it. It is obvious to everyone that the rich donor’s motivation is to get even richer by distributing deadly alms and that the targeted victims are always those who are likely to accept gifts from strangers out of necessity, be they beggars or, more broadly, the have-nots. The rumor and its local interpretations are shaped by a moral imagination that resorts to the occult to explain social inequalities: the rich are rich only because they selfishly sacrifice the poor by occult means. Even though nothing in the deadly alms scenario alludes to politics, people often suspect politicians of being behind deadly alms. They are thought to give deadly alms to get elected or appointed to a position of power. “They do it all the time,” a disillusioned journalist told me. In this view, deadly alms must be seen in the political context of Senegal in the late 2000s. The popular hope placed in Abdoulaye Wade at the time of his election in 2000 had given way to bitter disillusion because of the economic crisis and the persistent suspicions of corruption and racketeering in the higher reaches of government. The figure of the almsgiver who drives an SUV and distributes high denomination banknotes refers implicitly to the nouveaux riches of Wade’s era who live in ostentatious luxury, while the have-nots struggle to make ends meet. Such interpretations in terms of witchcraft of riches and power have also been made occasionally in relation to penis snatching and killer phone numbers; but deadly alms more clearly connect anxieties over strangers and anonymity to occult economies and occult politics, the two main themes that have been investigated so far by studies of the witchcraft of modernity. Indeed, the almsgiver is not only rich but he is also unrelated to his victims, just like the penis snatcher. He is said to drive around in his SUV and to stop in a locality only to circulate deadly alms before leaving hastily. Some (but not all) narratives hint at the almsgiver’s ethnicity: he is light-skinned and wears a turban. This ethnic stereotype could refer to Fulanis (who are present all across Senegal) or to Moors from Mauritania. In one incident, this stereotype played a role in the accusation. A peddler passing through Tambacounda, in the south of Senegal, was accused of distributing deadly alms in a market. He was lynched by an angry mob, rescued by a policeman and admitted to intensive care in the local hospital. The man attracted suspicion not only because he was carrying a black plastic bag (though he had no intention of giving its contents), but also because he was light-skinned and was wearing a turban. The peddler was a Fulani coming from the border of Mauritania and he was clearly perceived as an outsider in Tambacounda, which is on the other side of the country. However, rumors of deadly alms do not lead to overt xenophobic violence to the same extent as penis snatching. The fact that the almsgiver is a stranger is much more important than his ethnicity. “We don’t know who does this,” a Senegalese interlocutor told me. “The car stops, they open the window, give the package and leave. We can’t know their nationality.” The almsgiver is anonymous, because he conceals his face with a turban and remains hidden inside the car, about which it is sometimes added that it has tinted windows. The car is therefore used as a mask. This explains why many narratives omit the vehicle’s occupants, as if the car itself was the main protagonist: “the car gives alms,” it was often said. It was even called the “death SUV” (4x4 de la moit). The “death car” is a motif that recurs in many other African rumors. A few weeks before the appearance of deadly alms, another rumor was circulating in Senegal, according to which someone driving a luxury car was kidnapping children to sacrifice them. The same rumor has been circulating in Gabon for some years: a black car with tinted windows prowls around high schools and kidnaps girls who are later found dead. These stories are not totally new. During the colonial era, in East Central Africa, fire trucks and ambulances were suspected of hiding occult wrongdoings behind tinted windows (White 2000: 127–30). Strange windowless vehicles and cars with curtains over their windows were also mentioned. All these rumors draw on anxieties about situations of asymmetric anonymity: they stress the dangers of being watched and preyed upon by hidden agents who cannot themselves be seen. In some incidents related to deadly alms, cars driven by strangers are enough to attract suspicion, even though no gifts are involved. Anonymous gifts are nevertheless the focal point of most incidents, as in the following example. A man called Baldé works in Kolda and wants to send a sack of rice to his mother who lives in a distant village. He entrusts the gift to a taxi driver who was planning a journey to a village near his mother’s. When the driver arrives in the locality where he intends to drop off the rice so that Baldé’s mother can pick it up later, he meets hostile villagers who refuse to take it and order him to leave at once. Baldé, whom he calls on the phone, advises him to leave the rice in another village a few miles away. But the same scene occurs again. The taxi driver has no option but to drive back to Kolda and return the gift to its giver. The incident is the result of the presence of intermediaries in the act of giving: what was originally intended as a simple family gift is turned into a threatening anonymous gift. Not only does the giver commit the gift to an intermediary, but also the villagers are not the final recipients and are not expecting to be used as intermediaries. From their perspective, the taxi driver is just a stranger who comes to their village and insists on giving them an unsolicited gift. This is enough for the sack of rice to appear suspicious even though it does not match the common description of the deadly alms. Other incidents have been caused by an even more unsettling category of anonymous gifts: gifts without a giver. In a village near Sédhiou, a man found a ten thousand CFA francs banknote by the wayside, “just after an SUV drove by.” Instead of picking it up, he raised the alarm. A crowd gathered around the banknote, but nobody dared to touch it. As some people suggest burning it, a schoolteacher made his way through the crowd and grabbed the money. His bold gesture elicited mixed reactions. While some praised him, others condemned his foolhardiness. After the episode, local shopkeepers refused to take any money from him for fear of falling victim to deadly alms through him as the intermediary (as if money’s anonymity was contaminated by the dangers of anonymous gifts). A similar incident occurred in Diourbel. In a market, a tailor found a bag in front of the local mosque. A crowd of onlookers quickly gathered around the mysterious package. A man eventually dared to open the bag, where he found percale. People speculated: was it just a charitable gift that had been left anonymously in front of the mosque, as recommended by the Koran? Or was it the deadly alms that everybody was talking about? The latter incident highlights a tension inherent in almsgiving. Giving alms is an ordinary gesture and often a daily one in Senegal. It is indeed a religious obligation for Muslims (about ninety percent of the population of Senegal). The Wolof word for charity—sarax—is derived from the Arabic sadaqa, which is used throughout the Koran to designate alms (Weir 1995). Besides, Muslim ethics recommend that charity be given anonymously rather than publicly and conspicuously: only impersonal gifts are truly free gifts. In this view, deadly alms are nothing but a perverted form of sarax, a pious gesture turned into sorcery. In Wolof, they were indeed called sarax buy rey (“charity that kills”) or saraxu dee (“charity of death”). In fact, deadly alms bring to the fore an ambiguity underlying the moral economy of charity. Since the giver’s motives remain opaque to the recipient, anonymous almsgiving involves the threatening possibility that the selfless gift in fact conceals selfish motives and may even turn out to be detrimental to the recipient. Religious charity ought to be a gratuitous act that entails no direct reciprocity. And yet it usually implies the hope of a divine reward, either spiritual or material, in the afterlife or here below (Cruise O’Brien 1974). Even though recipients benefit from charity, they are only intermediaries or even instruments in a larger relation between almsgivers and God. This is all the more so as alms are often prescribed by marabouts in order to protect the giver against misfortune or to succeed in an undertaking.15 Thus, in the second incident described in the introduction, Mustafa’s motive for giving alms of meat was to protect his family, as he told me. He was accustomed to performing such a sarax once a year. Even though he was well aware of the rumor, he regards himself as a pious Muslim and did not expect that recipients might mistake his gift for deadly alms. Alms prescribed by marabouts are viewed ambivalently, since the almsgiver’s success is sometimes thought to be obtained at the expense of the recipient. Beggars complain that they are used as “repositories for bad luck” (dépôts de malheur).16 They are afraid of receiving gifts “weighed down” by misfortune and are therefore suspicious of the gestures and words that may go with the gift. “If someone spins the alms around his head before giving it, I never accept it,” declares a beggar. Another refuses alms that go with incantations: “I don’t know which prayers have been recited upon it. That’s why I refrain from taking it.” Since only a thin line demarcates magic from sorcery, marabouts are regularly suspected of being sorcerers (an accusation often related to the distinction between Muslim and pagan practices). They are said to perform liggéey (which means “work,” but also refers to magic) not only to protect their clients, but also to help them harm others. And this suspicion affects almsgiving as well. From this perspective, the rumor only takes to the extreme preexisting anxieties over the motives of anonymous almsgivers. Deadly alms elicit public warnings against anonymous gifts, not unlike those against anonymous handshakes in the case of penis snatching. Beggars cannot afford to stop begging, but they grow even more suspicious. “We live on alms,” says a beggar, “so we’re compelled to accept what people give us. However we won’t accept just anything anymore. When somebody gives us alms, we screen the content.” Conversely, almsgivers refrain from giving in fear of being accused. Another beggar observes: “We have been receiving fewer alms lately. The few regular almsgivers who dare to go on being generous hesitate when they are about to give. Those who used to step out of their car to greet us now just reach out and slip away” (but this behavior is not without risk, precisely because it can attract suspicion). Rumors of deadly alms thus jeopardize the everyday gestures of charity. They illustrate the moral perils of giving and receiving (Parry 1989). Deadly alms are in fact sacrifices disguised as gifts. In Wolof, sarax means both charity and sacrifice. Indeed, charity often includes the sacrifice of an animal and the gift of its meat. Deadly alms expose the occult dimension of this gift-cum-sacrifice. Meat, one important item of the deadly package, clearly alludes to sacrifice. But ultimately, the sacrificial victim is not the animal but the recipient, who is killed by the gift. “Human sacrifices” and “ritual murders” are recurring themes in this kind of rumor; the popular imagination often associates them with pacts with sorcerers and evil spirits (jinne or seytaane). Deadly alms are frequently interpreted along such lines. It is thought that the almsgiver has engaged in a pact with evil powers in order to obtain wealth or power in exchange for the sacrifice of a human victim. Instead of being the beneficiary of a gift, the recipient turns out to serve as an exchange value to pay a debt. On the occasion of almsgiving, without even realizing it, the recipient gets involved at his own expense in an opaque relationship of a totally different nature with evil powers. Deadly alms also represent a variation on the theme of the poisoned gift, a recurring topic in African witchcraft. They evoke the threatening possibility of a total reversal of the everyday logic of anonymous charity. The “murderous benefactor” (bienfaiteur assassin), as the almsgiver is sometimes called, appears as an oxymoronic figure who has something in common with the treacherous handshake of the penis snatcher. Deadly alms are thus a dramatic illustration of the crisis of the gift in contemporary Africa. This crisis affects not only kin-based reciprocity, as Filip de Boeck (2005) has shown in the case of the accusations against child-witches, but also gifts between strangers. These new forms of witchcraft bear witness to “the cracks and flaws that have started to appear in the urban gift logic” (ibid.: 209). “What poses as gift in the social interaction is no longer what it appears to be. Underneath the visible gift lurks another invisible pattern which corrupts regular patterns of exchange” (ibid.: 208). This perfectly applies to deadly alms—a sacrifice posing as gift. Unidentified phone calls Compared with penis snatching or deadly alms, rumors of killer mobile phone numbers bring to light another facet of Africa’s occult modernity. In Africa as elsewhere, mobile phones are the true emblems of modernity. They are indeed perfectly aligned with its main cultural values: mobility, individualism, permanent accessibility, and space-time compression (McIntosh 2010). However it would be overly simplistic to interpret rumors of killer phone numbers as a fear of modernity. Such rumors are not a symptom of cultural resistance to foreign technological innovation, just as rumors of penis snatching are not a symptom of cultural maladaptation to urban life. On the contrary, the rumor is evidence of the enthusiastic appropriation of this new technology: the rumor is but a by-product of the popular craze for mobile phones in Africa. The rumor appeared in Nigeria in 2004, just three years after the introduction of GSM technology in the country. Nigerians have eagerly adopted this new technology of communication, as everywhere else on the African continent, where household landlines are very scarce (De Bruijn, Nyamnjoh, and Brinkman 2009). This enthusiastic appropriation nevertheless aroused new anxieties. The year before the rumor’s appearance, Nigerians participated en masse in a one-day boycott of mobile phones to protest against the high tariffs and the poor service provided by phone operators: phone calls were often disconnected or misdirected such that people sometimes found themselves talking to unintended and unknown interlocutors (Obadare 2006). Killer numbers thus appear in an ambivalent climate of popular demand for mobile phones and widespread distrust towards phone operators. The rumor draws inspiration from Asian and American horror movies such as Ring (1998), Phone (2002), or One missed call (2004), which exploit the theme of haunted phones and which were probably circulating in Nigeria through pirated DVDs when the rumor appeared. It is not surprising that mobile phones have fed the imagination of horror movies and popular rumors: both cultural phenomena draw on the same anxieties about anonymous phone calls. Phones are “access points” through which users can be reached by others (Goffman 1971: 351). Mobile phones grant even greater access than household landlines, since individual users can be reached always and everywhere. But accessibility inevitably means potential intrusion and thus vulnerability: “points of access can easily become points of alarm” (ibid.). Phone users are exposed to unsolicited calls by unknown and invisible others. Compared with face-to-face interaction, phone communication is carried exclusively through the vocal-auditory channel without any visual access to the other party. Hearing without seeing makes identification more difficult for participants, although mutual identification is a requisite for every social interaction. This is why the first turns at talk in a phone conversation are systematically devoted to identification between parties (Schegloff 1979). But when the phone rings, before the first words are spoken, there is a sharp asymmetry between caller and answerer since the latter does not know who is calling. The person being called is therefore more vulnerable than the caller, at least in the preopening and opening sequences. Caller ID—a service usually available on all mobile phones—enables preopening identification and thus neutralizes asymmetry, even though call masking restores it (Schegloff 2002). The phone’s address book further helps anticipated identification by substituting proper names for abstract numbers. Killer numbers play precisely on these technological features common to all mobile phones. The calling number flashes on the screen but is not identified in the phone’s address book. Indeed, the danger only comes from unidentified calls; calls from friends or relatives are said to be safe. This explains why killer calls only affect mobile phones. Landline phones did not usually have caller ID: all phone calls were, therefore, initially anonymous. As a result, with landline phones the rumor could not focus on specific numbers. Killer numbers stress the contrast between identified and unidentified calls, just as penis snatching puts the emphasis on the distinction between unacquainted and acquainted people. In the incident described in the introduction, Vincent’s reaction is telling: he gets scared when he looks at the screen and sees “a number and not a name.” In a similar incident at a bus stop in Lagos, a street seller receives an unidentified phone call and raises the alarm, calling on the blood of Jesus to save him from killer calls. The scene attracts a crowd. Later, he finds out that the caller was in fact one of his friends who had just acquired a new phone number. The rumor thus focuses on the threat of being contacted by a total stranger whose identity comes down to a phone number (in Mali, the mysterious caller even uses a masked call to hide his identity). “The names of the persons to whom the killer numbers belong should have been publicly revealed,” insisted a Gabonese journalist (Gabon Flash, August 22, 2004). The fact that alleged killer numbers were often unattributed or inaccurate (with too few or too many digits, or with nonexistent prefixes) did not weaken the rumor: on the contrary, these incongruities emphasize the mystery of the phenomenon. They were also regarded as a trick from the caller even better to conceal his identity. Killer numbers thus point out the dangers of anonymity, just like penis snatching and deadly alms, but they relate to technological rather than urban anonymity. The rumor dramatizes a vulnerability intrinsic to phone communication: the banal risk of being exposed to anonymous calls becomes a mortal danger. But in a sense, it was to be expected that a technology, which is based on a disjunction between hearing and seeing and which enables long-distance communication between parties beyond physical copresence, could be considered an occult phenomenon. The enchantment of mobile phones does not proceed from a lack of understanding of modern technology but, on the contrary, emerges from an acute perception of its potentialities. Phones are witchcraft technology by design. In 2002, in Gabon, I saw several mobile phone users rejecting unidentified calls, for fear that it could be a witch’s trick to send them “night-guns” (fusils nocturnes).17 Two years before the rumor’s appearance, anxieties over the potential use of mobile phones as witchcraft were already there. Killer numbers must be seen in the context of the social use of mobile phones in Africa, just as was seen with penis snatching and urban sociality. Whereas the Internet is predominantly used to establish new contacts, notably with foreigners from Western countries, mobile phones are instead used to manage and strengthen preexisting social relationships among kinship, friendship, or business networks (Slater and Kwami 2005). In Nigeria for instance, “people make many calls just to say hello and to be in touch. To have a phone and not use it to reach out to family and friends is similar to the idea of living alone or preferring solitude and privacy to social interaction…. The vast majority of ordinary customers use a good deal of their credit making calls that are the cellular telephone version of a friendly visit” (Smith 2006: 506). The ubiquitous practice of “beeping” among African mobile phone users corroborates this analysis. Beeping consists in calling someone and hanging up at the first ring (Donner 2008). The reason for beeping is to urge the person called to call back and thus to meet the communication cost. Beeping follows rules of etiquette: in business relationships for instance, buyers are entitled to beep sellers who ought to call back, whereas in personal relationships, women beep and men call back. Beeping, as well as the transfer of communication credit by SMS (another common practice), thus inscribes mobile phone use in a preexisting nexus of social relations and exchange patterns. Such practices are restricted to communication between people who are already acquainted, who share strong ties rather than weak ones. Besides, beeping requires caller ID. This routine use of beeping and caller ID makes even more sensitive the difference between identified and anonymous calls. The threatening situation evoked by the rumor stands in sharp contrast to the ordinary social use of mobile phones. From this perspective, killer calls represent the occult flipside of beeping. Killer numbers are not regarded as the unfortunate outcome of technological hazard, but as a premeditated act of witchcraft. However, the mysterious agent hiding behind an anonymous number is not necessarily thought of as a single individual. According to many, mobile phone companies could be involved since they control telecommunication networks. These suspicions bespeak the anxieties of mobile phone users, who feel at the mercy of large-scale impersonal forces controlling complex technologies. These impersonal forces are all the more difficult to grasp and, therefore, all the more threatening as, in most African countries, since the privatization of the telecoms sector in the 1990s and 2000s, mobile phone operators have been in the hands of multinational companies of foreign origin, which operate on a transnational scale. The rumor thus exemplifies the occult suspicions that surround, in the popular imagination, the alliance between new communication technologies and global capitalism. A journalist writing about rumors of killer numbers, for instance, denounced the collusion between “Nigerian systems, leaders and corporate organisations” (Guardian, August 4, 2004). Foreign powers are sometimes suspected of hiding behind killer numbers. In Kenya, Somalia, and Mali for instance, killer numbers are said to come from abroad, so much so that all international calls become suspect. The rumor therefore evokes the threat of being exposed through the mobile phone to anonymous actors, impersonal forces, and foreign powers whose scale far exceeds the familiar circle of social relationships, of which the phone’s address book is a direct expression. What is at stake in the rumor is not an overall fear of technological modernity, but rather the wide range of relationships to which mobile phone users are likely to be exposed, without even knowing it. Just as penis snatching and deadly alms elicit recommendations to avoid shaking hands with strangers or receiving gifts from them, killer numbers elicit public warnings against anonymous calls. “Don’t answer any call if you don’t know the number; otherwise you’ll just die for nothing” (Vanguard, July 28, 2004). Mobile phone users are led to change their communication routines, at least as long as the rumor remains active. Switching off all mobile phones is the most radical response. “I have decided to disconnect the line, I don’t want any of my children to die mysteriously,” declared one Nigerian businessman in Ikeja, “and I have warned my wife against receiving unregistered phone numbers, nobody can say which one is mysterious or not. I have to take precaution. Prevention, they say, is better than cure” (This Day, July 23, 2004). But this has a high social cost, since phone users become temporarily unreachable to all, including their friends and relatives. Screening calls is a less drastic measure. “The story spread so fast and got so twisted that at a point people completely stopped picking calls from numbers which they did not have on their phonebooks” (This Day, May 24, 2008). “‘It’s better to miss an important call than for somebody to die a mysterious death just because she wants to answer a phone call,’ observed one Nigerian woman. ‘Most of my friends have their own phones and if they want to call me, they’ll use their phones,’ she submitted, just as she made another attempt to stop her phone from continuing its ringing” (Vanguard, July 28, 2004). Another protective measure consists of letting the mysterious caller speak first—a blatant violation of ordinary turn-taking in phone conversation. A Nigerian TV newscaster warned viewers: “If it is important that you have to receive all your calls, you have to allow the callers to speak first before replying them. Once this is done, you are free from their trap. But if you first say hello, you are gone for it” (This Day, July 23, 2004). The same warning is to be found in Mali: “If you pick up the call, do not speak first even if the caller greets you in Arabic, the language of the Prophet” (L’Indépendant, October 4, 2010). Though the latter statement is contradictory (speaking first alter being greeted!), it shows that the danger comes from answering the call, rather than just from picking up the receiver. The danger does not lie in the technological device by itself, but in the communication it enables. This is a matter of exposure and vulnerability. Speaking first and saying hello means exposing oneself and granting the caller access to one’s identity; whereas keeping silent means refusing contact and forces the caller to expose himself. The rumor thus highlights the hazards of anonymity and identification in phone conversations between strangers, as well as the ways these hazards are dealt with via the delicate social organization of taking turns at talk. In Vincent’s misadventure (described in the introduction), these hazards of anonymity are extended and repeated in the surreal dialogue, or rather absence of dialogue, in the aforementioned exchange between the venturous journalist and the unfortunate woman to whom the alleged killer number belonged. The journalist’s phone calls were intended to shed light on the mystery, but only resulted in even greater misunderstanding (several similar—and unsuccessful—attempts have been made in other incidents). This phone conversation, with its absurd summons and aborted beginnings, amounts indeed to a monument of incommunicability and misidentification between strangers: “– Hello, who am I speaking with? – How can you call my number and ask me ‘who am I speaking with?’” The dangers of anonymity Penis snatching, deadly alms, and killer phone numbers all illustrate the dangers of anonymity. Anonymity stands out clearly as the most distinctive common denominator of these new forms of the occult, especially if contrasted with “family witchcraft,” which represents the archetypal form of witchcraft in Sub-Saharan Africa. Witchcraft most often stems from the family. The close link between witchcraft and kinship continues in modern settings: “even in modern contexts—for instance, in the big cities—witchcraft is supposed to arise, first of all, from the intimacy of the family and the home” (Geschiere 1997: 11). In Libreville for instance, divination sessions performed by nganga (witchdoctors) aim, above all, to locate the origin of witchcraft on the victim’s “mothers’ side” or “fathers’ side.” As a result, the African witch has been predominantly conceptualized as a treacherous insider in anthropological tradition: he or she is “the hidden enemy within the gate” (Mayer 1970: 61). And yet, as Filip de Boeck rightly observes, “witchcraft is no longer something from within…. Contrary to older forms of witchcraft, the witchcraft ‘new style’ is wild, random and unpredictable, without clear direction or intention…. Because the possible sources of witchcraft are often disconnected from kinship relations, the danger may now come from anywhere. One becomes bewitched in public places like markets and shops, and through relations with unrelated or anonymous people” (de Boeck and Plissart 2004: 203). Witchcraft undeniably represents the “dark side of kinship” (Geschiere 1997: 11); but it now increasingly represents the dark side of anonymity as well, as penis snatching, deadly alms, and killer numbers paradigmatically show. In contrast with the circulation of witchcraft gossip within small-scale kinship networks, broadcast rumors bring witchcraft to a wider dimension (Stewart and Strathern 2004). This “unbounded” or “deterritorialized” witchcraft circulates through transnational rumors that are connected to the global flows of information and images of modern mediascapes (see Englund 2007 for a similar argument). These occult rumors are quintessentially modern insofar as anonymity itself is a key aspect of modernity. Penis snatching, deadly alms, and killer phone numbers draw on two different situations of anonymity, which relate to two distinctive repertoires of modernity. Face-to-face encounters with strangers, around which both penis snatching and deadly alms revolve, are typical of urban modernity (though admittedly not exclusive to it). Mediated interactions with distant and often invisible others, around which killer phone numbers revolve, are part and parcel of technological modernity (cars with tinted windows, mentioned in rumors of deadly alms, in fact fall somewhere between the two categories). Because they are based not on parochial circumstances but on very common situations, these rumors can easily apply to many different places. They must be understood in the context of the cityscapes and technoscapes constitutive of modernity in Africa (“The city” and “Technology” are indeed two key chapters of the Readings in modernity in Africa [2008] edited by Peter Geschiere, Birgit Meyer, and Peter Pels). They speak of the particular ways in which urban life and new information and communication technologies are lived and experienced on the African continent. These occult rumors are not only new imaginings for new times, but also more precisely “new imaginings for new relationships” (White 2000: 22). Inasmuch as modernity has extended the scope of human sociality in unprecedented ways through urban life and technology, it has extended as well the scope of the occult. Occult rumors are but by-products of Africa’s modernity. Penis snatching, deadly alms, and killer phone numbers give evidence of the pervasive sense of insecurity characteristic of Africa’s occult modernity. They are paradigmatic of what Adam Ashforth (2005) calls “spiritual insecurity,” a sense of exposure and vulnerability to occult forces, which relates to other dimensions of insecurity in everyday life. This spiritual insecurity is but one aspect of what Pierre-Joseph Laurent (2008), for his part, calls “insecure modernity” (modernité insécurisée). This condition of insecurity brings about a climate of generalized suspicion, which is a hotbed for witchcraft accusations and violence, since no one trusts anyone any more. Occult rumors take this collapse of trust to its extreme. “Everybody grows suspicious of everybody else,” explains a Gabonese man commenting on penis snatching. “We don’t know who is who any more,” adds another man. As penis snatching, deadly alms, and killer phone numbers all clearly show, situations of urban and technological anonymity are a locus of special vulnerability, in which insecurity is brought to the fore. Both face-to-face and mediated anonymity are potentially insecure because participants are either mutually unacquainted or invisible to one another. Anonymous situations are, therefore, fraught with uncertainty and opacity. This opacity concerns not only the participants’ identity and intentions, but also the very framing of their interaction, from which it derives its meaning. As a result, situations of anonymity often involve diminished trust between participants. In normal circumstances these hazards of anonymity can be smoothly dealt with in the course of everyday interactions. But occult rumors dramatize them and turn them into mortal dangers. Penis snatching, deadly alms, and killer phone numbers bring about a crisis of anonymity (as acute as it is short-lived). In the wake of the rumors, even the most ordinary gestures become hazardous and attract suspicion. People start screening phone calls, gifts, handshakes, and even glances from strangers. Cracks suddenly appear in the orderly give-and-take of daily interactions. Occult rumors thus feed on anxieties born out of very ordinary situations: shaking hands with strangers, accepting anonymous gifts, receiving unidentified phone calls, offering water to anonymous guests, or spotting a car with tinted windows driving around the neighborhood. From this perspective, “spiritual insecurity” might seem to be too loose and misleading a term to describe these phenomena. Admittedly, the people involved often cast penis snatching and the like in “spiritual” terms. In Francophone Africa for instance, mystique has become a catchall epithet, which encompasses everything from religion to magic, witchcraft, and sorcery. But many “mystical” phenomena in fact stem from perfectly mundane situations and could thus be more aptly analyzed as the unfortunate outcomes of an insecurity or vulnerability that pervades everyday social interactions. Such a redescription in terms of “interactional” rather than “spiritual” insecurity would enable us to deconstruct the category of “the occult” by reconnecting it to more mundane concerns and, thus, to de-exoticize witchcraft and sorcery even more convincingly (though, undeniably, Ashforth has already achieved a lot in this respect). This attempt at de-exoticizing the occult is the reason why I draw heavily on Erving Goffman and microsociology, a reference that might look odd from the perspective of the existing literature on witchcraft and modernity. Goffman is best known as the brilliant ethnographer of the daily routines of social interaction. His analysis of the public order shows us, according to Anthony Giddens, “how modernity is ‘done’ in everyday interaction” (1991: 46). The public observation of civil indifference between strangers in Western cities, for instance, represents an implicit contract of mutual acknowledgement and protection and thus serves to sustain attitudes of trust and provide a sense of security, on which modern social life depends. But, in the same movement, Goffman also brings into light the occult flipside of modernity. Indeed, he often describes everyday life as something fraught with hidden dangers. His essay on “Normal appearances,” published in Relations in public (1971), is a masterpiece of this sort. A reviewer of Goffman’s book presents the essay as “a tour de force of paranoid logic and imagination” and even views Goffman as “the Kafka of our time…because he communicates so vividly the horror and anguish—as well as some of the absurd comedy—of everyday life” (Berman 1972). Influenced by spy fiction as well as the ethology of predation, the essay indeed exposes the many dangers that lurk behind the normal appearances of public order and always threaten to undermine it. As Goffman (1971: 375) warns us: What is important is this: given that an apparently undesigned contact can turn out in retrospect to have been the first visible move in a well-designed game being played against the individual, and therefore not incidental at all, and given, further, that a genuinely incidental contact can be opportunistically exploited by bad characters—given all this—it follows that any current incidental contact that has so far not led to anything alarming might indeed do so. Strikingly, one could not find a more appropriate description of the situations in which suspicions of penis snatching, deadly alms, or killer phone numbers arise. Thus, these occult rumors do not allude to an invisible otherworld, but more mundanely to “an everyday world turned upside down,” as Richard Handler (2012: 186) characterizes Goffman’s weltanschauung. In the wake of the rumors, anonymous encounters turn out to be the opposite of what they seem to be. People start suspecting that anonymity in fact serves to hide occult relations. Alleged victims of penis snatching, deadly alms, and killer phone numbers get involved in anonymous interactions, only to realize after the fact that they have been forced into opaque relationships of a totally different nature. All the incidents triggered by the rumors thus proceed from the same situations of dysphoria caused by face-to-face or mediated encounters with strangers (on dysphoria in interaction, see Goffman 1953). It is always the interaction’s recipient and not its initiator who grows suspicious and feels threatened: the one who is accosted and whose hand is shaken, the one who is given alms, or who is called on the phone. Because of the climate of fear created by rumors, the recipient is led to interpret the stranger’s initiative in a way that is at odds with its ordinary meaning: engagement cues are misconstrued as signs of hostility. When everyday rituals are no longer able to sustain trust and provide a sense of security, normal appearances no longer hold and even the most banal events cannot be taken for granted anymore. Such a fractured world where the sense of normalcy of appearances seems to be constantly on the verge of disruption aptly describes the situation of insecure modernity in which postcolonial Africa finds itself. This is a world of endless possibilities, but ones that all too frequently turn out badly: “Nigeria is a country where anything can happen and usually does,” remarked a journalist writing on killer phone numbers (Guardian, August 4, 2004). Thus, even though at first sight Erving Goffman and Jean and John Comaroff may seem like chalk and cheese, their serendipitous encounter is not without heuristic value in casting new light on Africa’s occult modernity. It illuminates how the wide-ranging dynamics of modernity affect the minutiae of human interaction and how occult imagination can pervade the most basic aspects of everyday life. Acknowledgements The author would like to thank the three reviewers for their helpful comments and Matthew Carey for revising the language. References Agbu, Jane-Frances. 2004. “From ‘koro’ to GSM ‘killer calls’ scare in Nigeria: A psychological view.” CODESRIA Bulletin 3–4: 16–19. Akindele, Femi. 1990. “A sociolinguistic analysis of Yoruba greetings.” African Languages and Cultures 3 (1): 1–14. Appadurai, Arjun. 1996. Modernity at large: Cultural dimensions of globalization. Minneapolis: University of Minnesota Press. Ashforth, Adam. 2005. Witchcraft, violence, and democracy in South Africa. Chicago: University of Chicago Press. Balandier, Georges. 1985. Sociologie des Brazzavilles noires. Paris: Presses de la fondation nationale des sciences politiques. Bastian, Misty. 1998. “Fires, tricksters and poisoned medicines: Popular cultures of rumor in Onitsha, Nigeria and its markets.” Etnofoor 11 (2): 111–32. ———. 2001. “Vulture men, campus cultists and teenaged witches: Modern magics in Nigerian popular media.” In Magical interpretations, material realities: Modernity, witchcraft and the occult in postcolonial Africa, edited by Henrietta Moore and Todd Sanders, 71–96. London: Routledge. Berman, Marshall. 1972. “Weird but brilliant light on the way we live now.” New York Times Book Review, February 27. Bernault, Florence, and Joseph Tonda (dir.). 2000. Special issue “Pouvoirs sorciers,” Politique Africaine 79. Bonhomme, Julien. 2009. Les voleurs de sexe: Anthropologie d’une rumeur africaine. Paris: Seuil. ———. 2011. “Les numéros de téléphone portable qui tuent: Epidémiologie culturelle d’une rumeur transnationale.” Tracés 21: 125–50. Ciekawy, Diane, and Peter Geschiere, eds. 1998. Special issue “Containing witchcraft,” African Studies Review 41(3). Comaroff, Jean, and John L. Comaroff, eds. 1993. Modernity and its malcontents: Ritual and power in postcolonial Africa. Chicago: University of Chicago Press. ———. 1999. “Occult economies and the violence of abstraction: Notes from the South African postcolony.” American Ethnologist 26 (2): 279–303. Cruise O’Brien, Donal B. 1974. “Don divin, don terrestre: L’économie de la confrérie mouride.” European Journal of Sociology 15: 82–100. De Boeck, Filip. 2005. “The divine seed: Children, gift and witchcraft in the Democratic Republic of Congo.” In Makers and breakers: Children and youth in postcolonial Africa, edited by Alcinda Honwana and Filip de Boeck, 188–214. Oxford: James Currey. De Boeck, Filip, and Marie-Françoise Plissart. 2004. Kinshasa: Tales of the invisible city. Ghent: Ludion. De Bruijn, Mirjam, Francis Nyamnjoh, and Inge Brinkman, eds. 2009. Mobile phones: The new talking drums of everyday Africa. Bamenda: Langaa Publishers. Donner, Jonathan. 2008. “The rules of beeping: Exchanging messages via intentional ‘missed calls’ on mobile phones.” Journal of Computer-Mediated Communication 13 (1): 1–22. Ellis, Stephen. 1989. “Tuning in to pavement radio.” African Affairs 88: 321–30. Englund, Harri. 2007. “Witchcraft and the limits of mass mediation in Malawi.” Journal of the Royal Anthropological Institute 13 (2): 295–311. Englund, Harri, and James Leach. 2000. “Ethnography and the meta-narratives of modernity.” Current Anthropology 41 (2): 225–48. Favret-Saada, Jeanne. 1980. Deadly words: Witchcraft in the Bocage. Cambridge: Cambridge University Press. Geschiere, Peter. 1997. The modernity of witchcraft: Politics and the occult in post-colonial Africa. Charlottesville: University of Virginia Press. Geschiere, Peter, Birgit Meyer, and Peter Pels, eds. 2008. Readings in modernity in Africa. Oxford: James Currey. Giddens, Anthony. 1991. Modernity and self-identity: Self and society in the late modern age. Cambridge: Polity Press. Gluckman, Max. 1972. “Moral crises: Magical and secular solutions.” In The allocation of responsibility, edited by Max Gluckman, 1–50. Manchester: Manchester University Press. Goffman, Erving. 1953. “Communication conduct in an island community.” PhD dissertation. Chicago: University of Chicago. ———. 1963. Behavior in public places: Notes on the social organization of gatherings. New York: Free Press. ———. 1971. Relations in public: Microstudies of the public order. New York: Basic Books. ———. 1981. Forms of talk. Philadelphia: University of Pennsylvania Press. Handler, Richard. 2012. “What’s up, Doctor Goffman? Tell us where the action is!” Journal of the Royal Anthropological Institute 18 (1): 179–90. Ilechukwu, Sunny. 1992. “Magical penis loss in Nigeria: Report of a recent epidemic of a Koro-like syndrome.” Transcultural Psychiatric Research Review 29: 91–108. Irvine, Judith. 1974. “Strategies of status manipulation in the Wolof greeting.” In Explorations in the ethnography of speaking, edited by Richard Bauman and Joel Sherzer, 167–91. Cambridge: Cambridge University Press. Laurent, Pierre-Joseph. 2008. “Éléments pour une socio-anthropologie de la défiance.” In Les cliniques de la précarité: Contexte social, psychopathologie et dispositifs, edited by Jean Furtos, 33–49. Paris: Masson. Lofland, Lyn. 1973. A world of strangers: Order and action in urban public spaces. New York: Basic Books. Mayer, Philip. 1970. “Witches.” In Witchcraft and sorcery: Selected readings, edited by Max Marwick, 45–64. Harmondsworth: Penguin Books. McIntosh, Janet. 2010. “Mobile phones and Mipoho’s prophecy: The powers and dangers of flying language.” American Ethnologist 37 (2): 337–53. Meyer, Birgit. 2009. “Response to Ter Haar and Ellis.” Africa 79 (3): 413–15. Meyer, Birgit, and Peter Pels, eds. 2003. Magic and modernity: Interfaces of revelation and concealment. Stanford, CA: Stanford University Press. Milandou, Augustin-Marie. 1997 “‘Type connu ! Qui ne le connaît pas ?’ Anonymat et culture à Brazzaville.” L’Homme 141: 119–30. Mitchell, James Clyde. 1987. Cities, society and social perception: A Central African perspective. Oxford: Clarendon Press. Moore, Henrietta, and Todd Sanders, eds. 2001. Magical interpretations, material realities: Modernity, witchcraft and the occult in postcolonial Africa. London: Routledge. Moore, Sally Falk. 1999. “Reflections on the Comaroff lecture.” American Ethnologist 26 (2): 304–6. Obadare, Ebenezer. 2006. “Playing politics with the mobile phone in Nigeria: Civil society, big business and the state.” Review of African Political Economy 33: 93–111. Parry, Jonathan. 1989. “On the moral perils of exchange.” In Money and the morality of exchange, edited by Jonathan Parry and Maurice Bloch, 64–93. Cambridge: Cambridge University Press. Raheja, Gloria Goodwin. 1988. The poison in the gift: Ritual, prestation and the dominant caste in a North Indian village. Chicago: University of Chicago Press. Ranger, Terence. 2007. “Scotland Yard in the Bush: Medicine murders, child witches and the construction of the occult.” Africa 77 (2): 272–83. Rowlands, Michael, and Jean-Pierre Warnier. 1988. “Sorcery, power, and the modern state in Cameroon.” Man 23 (1): 118–32. Sanders, Todd. 1999. “Modernity, wealth and witchcraft in Tanzania.” Research in Economic Anthropology 20: 117–31. ———. 2003. “Reconsidering witchcraft: Postcolonial Africa and analytic (un)certainties.” American Anthropologist 105 (2): 338–52. Schegloff, Emanuel. 1979. “Identification and recognition in telephone conversation openings.” In Everyday language: Studies in ethnomethodology, edited by George Psathas, 23–78. New York: Irvington Pub. ———. 2002. “Beginnings in the telephone.” In Perpetual contact: Mobile communication, private talk, public performance, edited by James E. Katz and Mark Aakhus, 284–300. Cambridge: Cambridge University Press. Shack, William, and Elliott Skinner, eds. 1979. Strangers in African societies. Berkeley: University of California Press. Simmel, Georg. 1950a. “The stranger.” In The sociology of Georg Simmel, edited by Kurt Wolff, 402–8. Glencoe: Free Press. ———. 1950b. “Metropolis and mental life.” In The sociology of Georg Simmel, edited by Kurt Wolff, 409–24. Glencoe: Free Press. Simone, AbdouMaliq. 2004. For the city yet to come: Changing African life in four cities. Durham, NC: Duke University Press. Slater, Don, and Janet Kwami. 2005. Embeddedness and escape: Internet and mobile use as poverty reduction strategies in Ghana. ISRG Working Paper Series 4. London: Information Society Research Group. Smith, Daniel Jordan. 2006. “Cell phones, social inequality, and contemporary culture in Nigeria.” Canadian Journal of African Studies 40 (3): 496–523. Snodgrass, Jeffrey. 2001. “Beware of charitable souls: Contagion, roguish ghosts and the poison(s) of Hindu alms.” Journal of the Royal Anthropological Institute 7 (4): 687–703. Stewart, Pamela, and Andrew Strathern. 2004. Witchcraft, sorcery, rumors, and gossip. Cambridge: Cambridge University Press. Ter Haar, Gerrie, and Stephen Ellis. 2009. “The occult does not exist: A response to Terence Ranger.” Africa 79 (3): 399–412. Weir, T. H. 1995. “Sadaka.” In The encyclopedia of Islam. New edition. Vol. VIII, 708–16. Leiden: Brill. West, Harry, and Todd Sanders, eds. 2003. Transparency and conspiracy: Ethnographies of suspicion in the new world order. Durham, NC: Duke University Press. White, Luise. 2000. Speaking with vampires: Rumor and history in colonial Africa. Berkeley: University of California Press. Youssouf, Ibrahim Ag, Allen Grimshaw, and Charles Bird. 1976. “Greetings in the desert.” American Ethnologist 3 (4): 797–824. Les dangers de l’anonymat : sorcellerie, rumeur, et modernité en Afrique Résumé : Cet article porte sur une série de rumeurs ayant récemment circulé en Afrique. Ces rumeurs de vol de sexe, de numéros de téléphone qui tuent et d’offrandes de la mort relèvent d’un même « genre » transnational, caractéristique de la modernité sorcellaire en Afrique. On a souvent reproché à la littérature consacrée à la modernité de la sorcellerie d’adopter une perspective trop macrosociologique. Par contraste, cet article adopte une approche microsociologique et examine les répertoires interactionnels dans lesquels s’enracinent ces nouvelles formes de l’occulte. Il montre que ces dernières exploitent des inquiétudes suscitées par des situations banales : les poignées de main entre inconnus, les appels téléphoniques non identifiés, les dons anonymes. Ces rumeurs tournent ainsi autour des dangers de l’anonymat et évoquent le risque de se faire entraîner à son insu dans des interactions louches avec des inconnus. Elles s’appuient sur deux formes d’anonymat, associés à deux répertoires spécifiquement modernes : les interactions en face à face avec des inconnus, typiques de la modernité urbaine; les interactions à distance avec des agents invisibles, typiques de la modernité technologique. En contribuant à élargir sans cesse davantage les limites de la socialité humaine, la modernité s’accompagne donc d’une inquiétante extension du domaine de l’occulte. En définitive, cet article éclaire sous un nouveau jour la place de la sorcellerie dans l’Afrique contemporaine et propose de nouvelles façons d’articuler perspectives micro et macro en examinant comment les dynamiques de grande ampleur de la modernité se traduisent dans les plus menus détails des interactions sociales. Julien Bonhomme is Associate Professor in anthropology at the École Normale Supérieure (Paris, France). He conducts research on rituals, witchcraft, and rumors in Gabon and Senegal. His publications include Le Miroir et le crane: Parcours initiatique du Bwete Misoko (2006) and Les Voleurs de sexe: Anthropologie d’une rumeur africaine (2009). ___________________ 1. Accounts of the first two incidents are based on fieldwork interviews in Gabon and Senegal (fieldwork in Senegal was conducted in collaboration with Julien Bondaz). The third account is based on a newspaper article: “Two Collapse in Calabar After Receiving GSM Phone Call,” Vanguard, August 3, 2004. All names have been changed to protect identities. 2. In the scope of this article, I do not take into account Asian versions of the rumor, which differ from African ones in various respects and, for instance, are not primarily interpreted in terms of witchcraft. 3. In a seminal book, Luise White (2000) proposes to study various vampire stories circulating in colonial Africa as a “transnational genre.” 4. Elie and Mustafa are victims of physical violence, whereas Vincent and his sister are victims of an invisible violence, which could appear imaginary from an outsider’s viewpoint, but which is perceived as no less real than physical aggression by those subjected to it. 5. As well as conducting fieldwork in Gabon and Senegal, I have also collected several hundred articles from the African press on penis snatching, killer phone numbers, and deadly alms. 6. Wolof is the most widely spoken language in Senegal: it is the native language of the Wolof people (about 45 percent of the population), but also the vehicular language spoken by the vast majority of Senegalese. 7. Combined with the analysis of the newspaper corpus, the study of real incidents represents a good strategy to investigate such elusive phenomena as rumors. These ethnographic case studies are based on interviews with witnesses and protagonists (both accusers and accused), as well as fieldwork in locales where the incidents occurred (such as markets or popular neighborhoods). 8. Considerations of virility and the crisis of masculinity in Africa are obviously important to understand penis snatching, but they are beyond the scope of this paper. On this topic, see Bonhomme 2009: 37–50. 9. In contrast with penis snatchers, the Killer Granny and the Biting Mammy are often considered individuated figures (hence the use of the singular). This is also the case with the “charity killer” in rumors about deadly alms (see infra). 10. According to my data, more than 90 percent of the victims of genital snatching, but also of the persons accused, were men. 11. All three rumors elicit public statements by various authorities (government officials, psychiatrists, journalists, religious leaders, traditional healers, etc.). But only some of them warn people against the dangers evoked by the rumors, whereas many others denounce these stories as false rumors or frauds. 12. Quoted in “Panic in Khartoum: Foreigners shake hands, make penises disappear,” The Middle East Media Research Institute, special dispatch series 593, 2003. 13. The Hausas are one of the largest ethnic groups in West Africa and are located mainly in northern Nigeria and in southern Niger. Hausa communities are also scattered throughout West Africa. 14. The Fulanis are a vast ethnic group spread over about fifteen countries, principally in West Africa. Fulanis and Hausas are often categorized together in Nigeria and neighboring countries. 15. “Marabouts” (serin) here refer to local healers and diviners (who most often combine Koranic and pagan magic); but the word is also used for religious leaders of Sufi brotherhoods. 16. For similar Indian cases of gifts contaminated by the givers’ sins, see Raheja 1988; Snodgrass 2001. 17. Night-guns are considered a common technique of occult aggression in Gabon.

Berthomé, Bonhomme, and Delaplace: Preface: Cultivating uncertainty This work is licensed under the Creative Commons | © François Berthomé, Julien Bonhomme, and Gregory Delaplace. Attribution-NonCommercial-NoDerivs 3.0 Unported. ISSN 2049-1115 (Online) Preface: Cultivating uncertainty François Berthomé, École des Hautes Études en Sciences Sociales Julien Bonhomme, École Normale Supérieure Grégory Delaplace, Université Paris-Ouest Nantrerre La Défense   Instead of contrasting various situations according to their degree of uneasiness, we might better ask of the most peaceful and secure what steps would be necessary to transform it into something that was deeply unsettling. Erving Goffman, Relations in public, 1971 This themed section explores situations of interactional uncertainty, namely contexts in which the grounds of an interaction cannot be taken for granted. How to be sure, for instance, that a barbed comment is only intended to tease and is not really meant to be offensive? In the same vein, how are we to deal with “white lies” and other strategic dissimulations in flirtatious relationships? And how can we ever be sure that a benign handshake does not in fact hide malevolent intentions? These are some of the issues the contributors address in this volume. All the essays gathered here deal with opaque situations that generate uncertainty from the participants’ points of view. Starting with a minimal working characterization of interactional uncertainty—viz. situations in which (at least some) participants cannot but wonder as to the nature of their collective action—we deliberately choose not to address questions of definition, be they emic or etic. Rather, in a more experimental manner, we explore diverse—or even heterogeneous—kinds of opacities and uncertainties in a wide array of social situations and cultural environments. Opacity itself (considered as a formal property of some contexts) may proceed from various causes, such as inscrutable intentions, strategic dissimulation, double binds, or critical knowledge-gaps between participants. As to their effects, situations of uncertainty can give rise to a wide range of emotional reactions, ranging from minor discomfort or embarrassment to feelings of anxiety or insecurity. Although highly diverse, these essays nevertheless converge around two points. First, several authors show that interactional uncertainty is not always reducible to accidental misunderstandings, but can also be a constitutive or “built-in” element of various social settings. Second, many contributions refuse to consider uncertainty exclusively as a problem to be faced and solved. They show not only how social agents navigate through opaque interactions, but also how they deal with opacity as a social resource enabling them to negotiate or even create relationships. In brief, they stress the productivity of uncertainty at the heart of human sociality. Outline of a microsociology of uncertainty From a methodological point of view, we share a commitment to shedding light on situations of uncertainty through detailed ethnographies of specific situations, adopting a quasi-microscopic lens of description. Therefore, our approach may be called “pragmatic,” insofar as all the contributors resort to fine-grained analyses of contexts of interaction and communication (Severi and Bonhomme 2009). It is undoubtedly within the subfield of (mostly American) linguistic anthropology that pragmatics has established its pedigree within our discipline (e.g., Silverstein 1976; Hanks 1996; Duranti 2004), drawing on philosophers of language such as Peirce (1955), Austin (1962) and Grice (1989). Two recent contributions to this field are of special relevance to our theme. In an inspiring article about Mayan shamanism, William Hanks (2006b) carefully analyzes the orchestration of verbal and nonverbal communication involved in divinatory sessions. He convincingly argues that, in some situations, participants may achieve common ground through (and not despite) mutual opacity and significant knowledge-gaps. From another perspective, a collection of essays edited by Alan Rumsey and Joel Robbins (2008) examines native doctrines about the opacity of other minds in Pacific societies. In order to address the issue of the inscrutability of intentions, they focus on the epistemic and moral dimensions of the indigenous language ideologies that regulate the everyday use of language. While we undoubtedly build on these approaches, we nonetheless do not focus primarily on speech, language ideologies, or folk theories, but rather investigate the interactional dimensions of opacity and uncertainty. Although linguistic anthropologists regard speech as a form of social action per se, they study the articulation between language, culture, and society by means of an analytical toolkit, mainly adapted from linguistic concepts. In contrast, we draw not only on linguistic anthropology but also on interactionist sociology, a tradition that passes from Simmel to Goffman all the way through the Chicago school. By exporting analytical tools elaborated within the field of urban studies (and usually confined to Western settings) to a wide range of locales and situations, we aim to de-exoticize the anthropological study of such classic topics as ritual, witchcraft, joking relationships, and so on. One of the most effective tools developed by interactionist approaches is the concept of the frame, introduced by Gregory Bateson (1972) and consequently developed by Erving Goffman (1974). Frames refer, first and foremost, to cognitive resources mobilized by actors to stabilize their interactions and agree on the nature of their collective action. However, in Frame analysis (1974: 10, 83), Goffman himself insists that frames are exposed to “transformational vulnerability,” through operations of keying (transforming an activity into something else—as in make-believe) and fabrication (the manipulative use of keying by one of the participants at the expense of others—as with scams). The excerpt from Goffman’s Relations in public (1971: 383) we chose as the epigraph to this preface highlights the unsettling potential of such transformations. Even more radically, Goffman considers situations in which the cumulative layering of keying operations brings about structural undecidability to such an extent that nobody can trust anybody else. He gives the example of exposed double agents to illustrate this kind of interactional complexity, but mocking and hazing offer other good examples of this. This focus on moments of uncertainty echoes recent developments in French pragmatic sociology. In his latest works, Luc Boltanski (2011, 2012) has questioned the Durkheimian premises of Goffman’s microsociology but also of ethnomethodology (Garfinkel 1967), challenging their optimistic stress on actors’ abilities to manage social adjustments and repair breaches in the normative order of interaction: The main contribution of the pragmatic standpoint to sociology has been to underline the uncertainty that threatens social arrangements and hence the fragility of reality. But it stops half-way when it places too much confidence in the ability of actors to reduce this uncertainty. (Boltanski 2011: 54) Taking seriously the phenomenological background of “radical uncertainty” against which reality is socially constructed, Boltanski puts the stress on moments where uncertainty re-surfaces as an inescapable issue that social actors must confront, as opposed to everyday routine in which actors tacitly disregard misconduct and misunderstandings. While arguing for the continued relevance of frame analysis and paying special attention to the interplay of everyday “getting-by” or “making-do,” and moments of uncertainty, we nevertheless diverge from these authors on two points. First, unlike Boltanski, we do not assume uncertainty to be a given—i.e., the background against which reality is constructed—but rather as an emergent product of specific interactional dynamics. Second, unlike Goffman and most of microsociology, we do not necessarily equate the occasional surfacing of uncertainty with a breach of the Durkheimian social order. Admittedly, some contributions elaborate on the dysphoric effects of misframing, be they errors of conduct—“false notes” as Goffman (1953) called them—or ambiguities of interpretation. However, other contributions consider more constitutive forms of uncertainty (i.e., settings which not only allow for discomfort or anxiety, but which are really predicated on opacity). Thus, we propose to explore the manifold ways uncertainty might surface within or between interactional frames. Negotiating relations through uncertainty: From ceremonial settings to everyday affairs The first two contributions to our themed section take ritualized behavior as their main topic of investigation. Indeed, ritualized behavior has often been seen as an experimentum crucis for the investigation of the interactional grounds of human sociality. In ritual contexts, participants can no longer rely on the ordinary assumptions that regulate everyday interactions: for instance, ritual action can entail an opacification of their own intentionality by participants (Humphrey and Laidlaw 1994), or a paradoxical condensation of conflicting modes of relationship (Houseman and Severi 1998). Opacity or equivocality thus appears as a direct outcome of rituals’ constitutive rules, rather than as a mere failure of performance.1 Building on these research trends, Laurent Gabail describes in his contribution the relational architecture of Bassari male initiation rituals in Guinea. The initiatory transformation produces new agents—sometimes called “odd fellows”—that are characterized by their complex identity and associated with supernatural beings. The originality of Gabail’s essay is to carefully analyze the opaque interactions through which this new identity is ritually staged (e.g., masked performances or the even more unsettling behavior of ritual clowns). These clowns systematically engage in disturbing interactions with women and children. Thus, while the former playfully partake in the situation, the latter are much more ambivalent about the significance of the clowns’ behavior. In his contribution, Emmanuel de Vienne explores yet another form of ritualized behavior: he revisits the classical concept of joking relationships through an original analysis of teasing and taunting among the Trumai Indians of Brazil. In his seminal “Theory of play and fantasy,” Bateson 1972) showed that play activities involve an implicit meta-communicative message—“This is play”—that establishes a paradoxical frame comparable to Epimenides’ paradox of self-reference.

Schrempp: Copernican kinship This work is licensed under the Creative Commons | © Gregory Schrempp. Attribution-NonCommercial-NoDerivs 3.0 Unported. ISSN 2049-1115 (Online) Copernican kinship: an origin myth for the category* Gregory Schrempp, Indiana University In many traditional mythologies, kinship constitutes the privileged idiom of both unity and diversity in the cosmos. In “post-mythological” thought, categories logically conceived attempt to take over the cosmic role of kinship. I compare two accounts of the nature and genesis of categories—those by Durkheim and Mauss on one hand, and by Lakoff and Johnson on the other. Neither account severs ties with mythology or kinship; moreover, the structure of the category, like kinship, offers a mode of projecting the human as the cosmic. To the long-standing anthropological concern with the ways in which humans impose their diverse categories on the world, we should add a concern with the ways category-theorists impose their diverse worlds on the category. Keywords: kinship, category, myth, philosophy, science, Maori, Copernican Revolution That quick dismissal of the idea of a universe without life was not so easy after Copernicus. He dethroned man from a central place in the scheme of things. His model of the motions of the planets and the Earth taught us to look at the world as machine. Out of that beginning has grown a science which at first sight seems to have no special platform for man, mind, or meaning. Man? Pure biochemistry! Mind? Memory modelable by electronic circuitry! Meaning? Why ask after that puzzling and intangible commodity? “Sire,” some today might rephrase Laplace’s famous reply to Napoleon, “I have no need of that concept.” John Wheeler, Foreword to Barrow and Tipler, The anthropic cosmological principle, 1988 At rest, however, in the middle of everything is the sun. For in this most beautiful temple, who would place this lamp in another or better position than that from which it can light up the whole thing at the same time? For, the sun is not inappropriately called by some people the lantern of the universe, its mind by others, and its ruler by still others. [Hermes] the Thrice Greatest labels it a visible god, and Sophocles’ Electra, the all-seeing. Thus indeed, as though seated on a royal throne, the sun governs the family of planets revolving around it. Moreover, the earth is not deprived of the moon’s attendance. On the contrary, as Aristotle says in a work on animals, the moon has the closest kinship with the earth. Meanwhile the earth has intercourse with the sun, and is impregnated for its yearly parturition.In this arrangement, therefore, we discover a marvelous symmetry of the universe, and an established harmonious linkage between the motion of the spheres and their size, such as can be found in no other way. Copernicus, On the revolutions, 1543 Within the history of mythic visions allegedly shattered by science, the Copernican revolution holds a special, epitomizing place. And so to catch Copernicus purveying solar myths is a mythologist’s delight—a double delight in fact because there are, first of all, the wonderful solar myths that Copernicus rounds up for his reader; then, on top of these, the interesting friction that these create for our myth of Copernicus. Why, wasn’t Copernicus the one who first led us to the vantage from which, for the first time, we could see the universe as it really is, and not as our petty, self-absorbed humanity would have it? But, the sun a king? Planets as his children? The earth impregnated by the sun for her yearly parturition? This is the sort of thing that we expect from pre-Copernicans. As Thomas Kuhn has described the pre-Copernican world: Though primitive conceptions of the universe display considerable substantive variation, all are shaped primarily by terrestrial events, the events that impinge most immediately upon the designers of the systems. In such cosmologies the heavens are merely sketched in to provide an enclosure for the earth, and they are peopled with and moved by mythical figures (1985: 5). It is possible of course to shrug all this off with the observation that it is certainly “nothing more than” metaphor, perhaps created by Copernicus as a mere sop for the mythically minded who would sit in judgment of his theory—though we must also ask whether this resolution does not sound a bit too contemporary and tidy.1 Even if we wished to hold out for a clear dichotomy, there is still much to examine in the very fact that certain metaphors are so broadly compelling. It would seem that the very reason that the kinship image of the universe occurring in Copernicus, the hero of science, has not triggered expressions of amazement is that no one perceives a problem with the image; even for the scientifically minded, it is obvious, natural, and little in need of analysis. Indeed, this highly mythologized vignette—depicting the sun as a lamp in a temple and a ruler surrounded by a family of planets—may well be Copernicus’s most quoted passage. Copernicus’s solar myths suggest many things, including, minimally, that it is always possible to soften a scientific blow by—metaphorically if you will—reinvesting the world with an anthropocentric vision. His familial images—kinship, kingship, parturition, and so on—have a long history and call to mind numerous cross-cultural analogues, for in many mythologies the sun is a parent or ruler, other celestial bodies are kindred, and rituals of mating and reproduction engender the large- as well as the small-scale rhythms of the cosmos. If indeed myth is the nemesis of science, the situation is especially ironic because the myths that Copernicus cites are precisely the kind that nineteenth-century “solar mythologists” attributed to the human mind in its very infancy. The solar mythologists, a dominant voice in the study of mythology in the second half of the nineteenth century, insisted that all myths were ultimately inspired by celestial phenomena.3 In speaking of the first thoughts and words of archaic humanity, for example, the German-born patriarch of this school, Friedrich Max Müller, wrote: Every word, whether noun or verb, had still its full original power during the mythopoeic ages. Words were heavy and unwieldy. They said more than they ought to say, and hence, much of the strangeness of the mythological language, which we can only understand by watching the natural growth of speech. Where we speak of the sun following the dawn, the ancient poets could only speak and think of the sun loving and embracing the Dawn. What is with us a sunset, was to them the Sun growing old, decaying, or dying. Our sunrise was to them the Night giving birth to a brilliant child; and in the Spring they really saw the Sun of the Sky embracing the earth with a warm embrace, and showering treasures into the lap of nature (1874: 64). There is yet another dimension of irony in Copernicus’s words. Specifically, in the Judaeo-Christian cosmogony of Genesis, we encounter not a sexually generated kinship cosmos, but rather a cosmos created by a craftsman who stands over against his work—a relation that is sometimes cited as a mythical prototype of the ideal of objectivity (even while anthropomorphically inflected as the “god’s eye view”). Yet it is just this metaphor of objectivity that Copernicus’s cosmic-family deposes. This all happens of course in parallel with the many other ways in which cosmic kinship, although denied in its officially-sanctioned cosmogonic myth, finds its way into Judaeo-Christian cosmology—from images of God as Father and Redeemer as Son, to the ideal of the nun as the consecrated bride of Christ, along with many other cults of folk religion that incorporate a sexually-generative dimension in cosmology. One also recalls St. Francis’s famous Hymn to the Sun which alternates between male and female siblings: Brother Sun, Sister Moon and Stars, Brother Wind and Air, Sister Water, Brother Fire, and Sister Earth, who is also referred to as Mother. The Creator in Copernicus’s image is enmeshed within, rather than standing outside of, the cosmos as a would-be object of knowledge. We are left, then, with the irony that, in terms of images or emblems of objectivity, Copernicus’s cosmic-family image is actually a step backward from the image it replaces, the divine Craftsman of Genesis. Or to put it differently: when epistemologists invoke the metaphor of the god’s eye view, they do not seem to have in mind the god Zeus, with his perpetual struggles to find solace from his wrangling kinfolk, who are tutelary beings of the various regions of the cosmos. Zeus’ great affliction as depicted by Homer—an affliction which the Judaeo-Christian Creator had to some degree escaped, at least until Copernicus!—is, in a nutshell, that he cannot extricate himself from his cosmos, which is to say, from his family. In regard to our cosmic ambitions, Zeus and the Judaeo-Christian God of Genesis form an interesting pair, reflecting, respectively, our conflicting desires to be enmeshed within and to stand contemplatively without our cosmos. Mythological cosmologies of kinship and of the body In this essay I will explore the differences between a mythological worldview and a scientific worldview from the standpoint of two topics—kinship and the category—that are often implicated in attempts to account for the transition from one worldview to the other, and which consequently serve as the site of numerous origin myths that sustain both. The image of a cosmos organized and held together by kinship is found in many mythologies and/or religious traditions. Although my biggest concern is kinship cosmology in the “post-mythological” world, it is important, as a preliminary step, to emphasize the complexity of those cosmologies of kinship that are typically called mythological. To this end I will summarize some of the patterns that I have noted in research on the cosmological traditions of Polynesia and especially of the New Zealand Maori. Those familiar with the ancient Greek mythic cosmogony recounted by Hesiod in his Theogony will note many parallels between this and the Maori story.4 Maori cosmology forms an epitomizing case for familial cosmology, since the Maori universe is conceptualized in most versions as a kin group. Perhaps the single most oft-quoted passage in the literature on the Maori is a passage from ethnographer Elsdon Best: When the Maori walked abroad, he was among his own kindred. The trees around him were, like himself, the offspring of Tane [ancestral God of forests, and in some accounts, humans]; the birds, insects, fish, stones, the very elements, were all kin of his. ... Many a time, when engaged in felling a tree in the forest, have I been accosted by passing natives with such a remark as: “Kei te raweke koe i to tipuna i a Tane.” (You are meddling with your ancestor Tane) (Best 1924: 128–29). The specific characteristics and meanings of Maori kinship, as well as the nature of the Maori commitment to the kinship cosmos (is this a literal belief? a poetic trope?) pose complex issues. In lieu of a complete analysis, I would like to offer some speculations about the way that this central postulate—that the cosmos is a kin group, the “kinship postulate” for short—is expressed and constructed in Maori narrative. First and most important, the kinship postulate seems to operate as a kind of generative principle for a large group of etiological stories, of which there may be a theoretically limitless number. Maori cosmogony begins as a version of the widespread story of the origin of the universe from the mating of sky and earth, the first male and female, as primal parents. In some Maori accounts the parents are pressed closely together until the children separate them and then go on to become fish, trees, plants, and humans and/or the tutelary deities of these entities; they spread through the world as the variety of forms comprised in nature. Maori narratives often proceed as though basing judgments of kinship descent more on functional dependency (especially of smaller creatures on larger ones) than on morphological similarity. For the natural habitat of a given entity is said to be its parent: shellfish are the children of the sea, birds the children of trees, insects the children of plants. Thus a consistent bio-cosmological principle—habitat as parent—stands behind the idea that earth and sky, the physical cosmos, are the parents of humans. While there are numerous forms of such interdependency, the web of kinship also comprises domination, jealousy, incompatibility, and indeed, war. One can surmise that it is just because of the rich variety of kinds of relations (and the possibilities for multiple overlays of different, even contradictory, relations) found in kinship that the Maori have settled on it as their basic cosmological image, an image that finds a place for any entity. One of the more discordant patterns displayed within these kinship stories has to do with the incompatibility of different life strategies displayed in the cosmos. The primordial example occurs in the story of the separation of Sky and Earth itself. In one of the versions of this story, that recounted by the Arawa chief Te Rangikaheke,5 one of the children is opposed to the idea of separating the parents. When his siblings overrule him and separate their parents, pushing Sky above, this dissenting child becomes Wind and, with his brood of children—who are various forms of storms and pernicious meteorological conditions—he continually takes revenge on his siblings, including the humans, even to this day. In responding to the very first attack of Wind and his brood, there is dissension among the other siblings. Some think the sea would provide superior shelter, while others argue that greater security is to be gained on land. The argument ends in stalemate: one group heads into the ocean and becomes the fish, the other group heads inland and becomes the plants. In splitting up to form the two tribes, the fish and plants both insultingly predict that the other will end up as food for humans. This scenario also allows a space for interstitial species—species difficult to classify. In the separation into land and sea creatures, one little band gets confused and heads in the wrong direction: these are the lizards—and thus we have an explanation of a particular land creature that bears many of the physical characteristics of the sea part of the family. Countless other narratives reproduce the pattern of kinship disjunction, from stories about the mountains, which lived together until splitting up to occupy different tracts, to stories recounting the many finer differences that can be noted among insects. There is this story about how the mosquito and sandfly tribes split over a debate about strategy: The sandfly said to the mosquito: “Let us go and attack man.” The mosquito remarked: “Let us await the shades of night ere we attack, wait until evening, that we may hum in his ears and make them tingle. . . . In this way we can safely attack, but if we attack in daylight, many will perish. If you assail man by day, then you of us two will surely be slain.” Arose the sandfly: “Let us ever give battle in the light of day. In daylight will I go forth. Though I perish in myriads, what matter it, so long as I draw blood. But you, O mosquito! Attacking by night, shall be destroyed by smoke.” Even so the mosquito consented, saying: “As you will” (Best 1925: 993). One notes in this account the same pattern as found in the primordial separation of land and sea creatures: differentiation amidst arguments and insults, speciation through incompatibility of lifestyles and distinct acts of will. These processes form sub-patterns within the larger kinship postulate, and they also demonstrate the intrinsic richness of this postulate—most especially in its capacity to portray unity and disunity at the same moment and within the same frame of reference. The accounts of the speciation of the natural world rely upon many of the idioms and much of the specific terminology associated with human kinship and with processes of tribal formation (notably the term iwi, tribe, for both human political units and natural species). Natural speciation provides a model for human tribalization and vice versa. But besides operating as a generative principle—of narratives, and of models of social process—the kinship postulate also operates as a principle of incorporation. The example most to the point here has to do with the well-known European fableof the lazy grasshopper and the industrious ant, a version of which appears in one of Elsdon Best’s collections of Maori narratives: The ant spoke to the cicada saying: “Let us be diligent and collect much food during the warm season, that we may even retain life when the cold season arrives.” But the cicada replied: “Nay! Let us rather bask in the sun’s rays on the warm bark of trees.” Even so did the ant toil throughout the summer at gathering and storing food in secure places. Meanwhile the cicada said: “What a fine thing it is to bask in the warm sun and enjoy oneself. How foolish is the ant who toils incessantly.” But when the cold season came, and the warmth went out of the sun, the cicada perished miserably of cold and hunger, while the ant was warm and well fed in his snug home underground (Best 1925: 991). The European version of this story is a fable—that is, a terse story told to illustrate a specific moral. The Maori version is not a fable, but an origin myth, or part of one. The Maori version does not announce a specific moral; instead, like the story of the mosquito and sandfly and many other stories of the Maori, it accounts for and dramatizes the different behavioral characteristics of different species. The European story has thus been incorporated in such a way as to be consistent with other Maori stories detailing the unfolding web of Maori cosmic kinship. I earlier alluded to Te Rangikaheke’s account of the separation of the first parents, Sky and Earth, in which the dissenting child goes to the sky, becomes the wind, and with his meteorological brood makes war on the children who remain on the earth. The attacks occasion a series of debates among the earth-inhabiting siblings about the means of defense: some run to the sea (and become the fish) some burrow in the land (and become the plants). But there is an exception: the being who will become the progenitor of humans, Tu, refuses to be intimidated; he stands strong, rather than hiding. The battle between Tu (the human; the term literally means “stand”) and the elements ends in a stalemate—an apparent allusion to the present ongoing relationship between humans and “the elements.” When this is achieved, Tu turns toward the siblings who, in seeking protection, had abandoned him—and in retribution Tu makes them his food. These events provide paradigmatic justification for humans eating the other things of the cosmos. At points it is almost as if the truly significant units of Maori genealogies are the nodes of innovation and diversification in strategy which the above scenarios illustrate. At any rate it is important to note that these nodes, dialogical in form, provide a particularly open format for the incorporation of new negotiations—such as those involved in the dialogue of the ant and cicada. The deep and intense implication of the idiom of kinship in Maori narrative—in expounding both cosmic interconnection and diversification—is captured by distinguished scholar J. Prytz-Johansen in a quaint image; he comments, “If one could picture to oneself a person like KANT among the old Maoris—which indeed is difficult—one should not be surprised if to the fundamental categories of knowledge, time and space, he had added: kinship” (Prytz-Johansen 1954: 9). I can think of no better analogy for describing how the kinship postulate operates as a generative and organizing principle for Maori narratives: it challenges the storyteller to create a story in broad conformity with other such stories; and, as a principle that is (to fill out the Kantian metaphor) a priori as well as synthetic, it confirms in advance that such a story is possible. But there are other ways in which one could explore the relationship of the kinship postulate to narratives in the Maori case. One of these, briefly, consists in the fact that the kinship postulate has implications for the Maori organization of genres. There are Maori narratives that are reminiscent of familiar Western folkloric genres (myths, local legends, epics, proverbs, and so on). Yet cosmic kinship seems to add another level of integration which links these narratives as one large narrative—because the characters from all of these narratives have a place in one encompassing cosmic genealogy, not totally unlike the ideal of “universal history” that has periodically appealed to Western historians. Rather than clear-cut myths versus historical legends, Maori narratives of the past flow into one another, beginning with something like myths and turning into something like legends—genealogy precluding absolute ruptures between genres. The kinship postulate also has implications for the very concept of narrative. The genealogy that links the various narratives itself forms a narrative of sorts, so that a Maori cosmologist can tell the story of the origin of the universe and of human history merely by reciting a genealogy. For example, the first origins of the cosmos—the story of the children of Sky and Earth groping around in the dark between the yet-unseparated parents—are portrayed through a series of terms that has the appearance of a series of names in a genealogy; yet, in their literal meanings, the terms describe a series of sequential states of the early cosmos: Nothing, Night, Searching, Growth, and so on. They tell the same story as the narrative of the separation of Sky and Earth through a series of highly condensed allusions.6 The foregoing provides one example of the elaboration of a kinship idiom in traditional cosmology, specifically that of the Maori; many other, highly-varied examples can be found in the world. Before moving on to consider “categories,” however, one more point should be added about cosmologies of kinship. Specifically, kinship, while designating a level of organization above that of any particular body, is yet ultimately inseparable from the idea of the body; bodies are the elements linked through webs of kinship, and indeed the body in its procreative power is the ultimate source of such webs. In the Maori case the interdependence of body and kinship as images of the cosmos is revealed, among other ways, in the fact that the first generative pair, Sky and Earth—sometimes with the implication that together they form a single body7—can be invoked to summarize the entire cosmos. Gaia plays a similar role in Greek mythology. Another example, indeed one of the best-known instances in Western cosmology, is the mythico-mathematical cosmos described by Plato in his Timaeus (32–34 [1981: 44–45]), in which the universe is characterized as a unitary living body. And these are not isolated instances. Although kinship provides what is probably the single most recurrent image or idiom through which mythologies portray the structure of the cosmos, the image of the cosmos as a body is not uncommon. Note, finally, that Copernicus (in the opening epigram above) alludes approvingly to both images, referring to the sun as the mind of the universe (which is presumably the body) and the head of a royal family (with the planets as children). Philosophies of familiar form: Kinship and the life of the category in the post-mythological world The era of the presocratic philosophers, beginning around the sixth century BCE, is characterized in part by a critical confrontation of the mythological understanding of the world by methods and perspectives that came to be known as philosophy. And chief among the tools of philosophy is the category (a notion especially important to Aristotle and having a forerunner in Plato’s notion of Forms). As a gross generalization, cosmic persons (gods and heroes and their genealogical interconnections) are to mythic understanding as categories are to philosophical understanding. I will explore this contrast, emphasizing not the ways in which philosophy differs from myth, but rather the parallels and overlaps between the two. I do not disagree with the general proposition that new ways of understanding the cosmos appeared on the scene or at least rapidly picked up momentum in the era of the presocratics and classical Greek philosophers. But, in the company of many other scholars, I am also intrigued at the ways in which the transition is less than neat and complete (with some sympathy for the view that it will never be so). The love of illustration drawn from Hesiod and Homer (even amidst skepticism toward them) found in Plato and Aristotle, as well as myth-like formulations (such as the image noted above from Plato’s Timaeus) are merely one early indication of the continuity of myth within the new philosophical spirit. I will pursue the lack of completeness, in the transition from myth to philosophy, with respect to one particular theme. Specifically, in contemporary epistemology, and in certain theories of the category in particular, we encounter images of kinship and the body that cannot but recall the former life of these images in the world of mythology. These images raise fundamental questions about the central if not centric engagement of kinship and the body within arguments about the very structure and nature of “the category.” Although the term “category” now tends to be used mainly to denote a class, or group of things classified together, the term was used by Aristotle (Categories [1938]) to designate ten ultimate predicates which summarized the possible attributes of things. Aristotle’s categories are Substance, Quantity, Quality, Relation, Place, Time, Position, Condition, Action, and Affection. Aristotle seems to have arrived at his categories through an ascent in abstraction to the most general and encompassing possible predicates. The eighteenth century philosopher Immanuel Kant drew on Aristotle’s categories and introduced some modifications in order to construct his table of “a priori categories of the understanding.” For Kant the categories constituted mental molds through which reason unifies and imposes form on the data of sensory experience; by “a priori” he meant that the categories were prior to any such experience. Though Kant regarded knowledge as possible only through these categories, he rejected Aristotle’s seeming conviction that the categories were transparent to being. According to Kant, the world we know is always and only the world as it is knowable to us through the a priori categories. Aristotle and Kant tended to regard such categories as givens of consciousness, saying little about how humans came to possess them. One late nineteenth-century Darwinian response to the Aristotelian/Kantian categories, however, was to argue that they might be accounted for by bio-evolutionary process: adaptively advantageous mental molds might be selected for by Darwinian processes—Herbert Spencer made such an argument.8 This Darwinian response in one sense, of course, contradicts Kant by seeking categories that are a posteriori, that is, products of experience over time. But evolutionarily advantageous categories would become part of the built-in mental structure that any particular human now inherits; in this sense the categories would “become” a priori. Kant had used the term “pure reason” to denote that which reason, by virtue of the a priori categories, is able to know prior to any sense experience. This new Darwinian project amounts to a search for an “impure”—that is, experiential—origin of “pure” reason. The work of French comparative sociologists Emile Durkheim and Marcel Mauss in the early twentieth century also holds a place within the tradition of the search for the impure origin of the categories. And I will argue that two influential contemporary scholars, George Lakoff and Mark Johnson, represent the continuation of this tradition. My analysis will focus on a comparison of these two pairs of scholars—Durkheim and Mauss, and Lakoff and Johnson (for convenience sometimes shortened to Durkheim/Mauss, Lakoff/Johnson). Both pairs pay homage to Aristotle, and, though not adopting his formulations in detail, ground their projects in part on the assumption that there exist basic, universal mental molds something like those that Aristotle long ago set out.9 Durkheim and Mauss find the source of the categories in the experience of society; Lakoff and Johnson find it in bodily experience. The Aristotelian/Kantian categories are thus contested territory: the issue is, what kind of experience provides the model for the categories? The contesting claims will be the focus for the analysis that follows. Before beginning, a terminological problem should be mentioned. Specifically, both Durkheim/Mauss and Lakoff/Johnson allude to the idea of categories in the Aristotelian/Kantian sense described above—that is, in the sense of very general mental molds or predicates; and both choose as foci at least some of the categories set out by Aristotle and Kant. But both Durkheim/Mauss and Lakoff/Johnson are interested in telling us also about the origin of “the category” in another sense, that is, in the more specific and contemporary sense of a class (or set or kind). Here the question is more specific: what is the source of our ability to create classes of things, that is, to group together disparate entities into kinds or sets. It is all too easy to pass from one sense of category to the other, in part because the idea of class (or category, in our contemporary sense) is but a short step from some of the Aristotelian categories (notably Substance and Quality) and their Kantian derivatives, and in part because both senses of category are relevant to both Durkheim/Mauss and Lakoff/Johnson. But my purpose is to compare Durkheim/Mauss and Lakoff/Johnson in terms of the general drift of their arguments. For this purpose, the distinction between these two senses of category often is not critical; where it is I have tried to make the sense clear through the context. The origin of categories Because many of the issues here are complex, I will summarize my arguments in advance, and then go on to fill them out by looking in more detail at the writings of Durkheim/Mauss and Lakoff/Johnson. My arguments can be summarized in three main points: First, the image that informs Durkheim and Mauss’s account of the origin of the categories is the image of kinship (either directly or as implicit in the idea of a “tribe” or “family”; Durkheim and Mauss regarded such kin-based units as the elementary forms of society). The image that informs Lakoff and Johnson’s account of the origin of the categories is the human body. As indicated above, these two images—kinship and the body—are also two of the most recurrent images drawn upon in portrayals of the cosmos within mythology. My first main point will be that the properties that make these two images good choices for myth-makers portraying the cosmos are the same properties that make them good choices for contemporary cognitive theorists seeking either the experiential origin of, or an origin myth for, categories or “the category.”10 Both kinship and the body are ever-present in human experience. Both kinship and the body offer potent models of unity in diversity—models of a whole made up of many interrelated parts. Images of kinship or the body in mythologies serve to organize and classify the cosmos; theorists attribute this same function to these images in accounting for the origin of the categories. My second main point will be that theories of the category offer as much potential for cosmological projection—that is, for imposing prosaic, local images onto the cosmos—as do traditional mythological portrayals of the cosmos. It is not difficult to recognize the process of such projection in a mythology that portrays the cosmos, as many in fact do, as one big tribe or as a gigantic living body. But, if Durkheim/Mauss are right, then, we have all been unconsciously imposing on the cosmos the form of the tribe or family every time we recognize a “set”—for all sets must, at some level, bear the imprint of their original model. By contrast, if Lakoff and Johnson are right, then what our sets reflect is not the tribe but the body as container. If either the Durkheim/Mauss or Lakoff/Johnson theory of the origin of the categories is wrong (and it would seem that at least one of these theories has to be wrong, since they are contradictory), or if both theories are wrong, then the situation is even more interesting. Consider this: for those who have been convinced by either of these theories of the category, one of these images—the family or the body—takes on a cosmic stature just by virtue of this conviction. To argue that we are all necessarily projecting, through the intermediary of the category, a particular form on all cognized entities, ipso facto instantiates that image cosmologically by making it into an object of belief, though perhaps less obviously than do traditional origin myths when they project such images directly as the structure of the cosmos. It is as if, in attempting to free ourselves from our parochializing inclinations, we forsake a mythological kinship-cosmos or body-cosmos for a cosmos based on a certain kind of abstract entity, the category—and the category meanwhile becomes the new object of our anthropocentric desires. In the tradition of Aristotle and Kant, categories are the enabling, mediating structures of all knowledge. A human image projected onto, or as, the category thus achieves, though by a less direct route, the same totalizing status as a human image that is painted by mythologies directly onto the canopy of the heavens. My third main point will be that juxtaposing two different theories of the “impure” origin of reason serves to bring to light the difficulty of adjudicating between contesting claimants. Both theories locate the source of the categories in those human experiences that they regard as most basic and formative. Opinions about what experiences are most basic and formative are academic discipline-centric. Durkheim and Mauss, who were attempting to forge a discipline of “comparative sociology,” saw experiences of social collectivity as the linchpin; Lakoff and Johnson, who are developing a more psychologically-oriented cognitive perspective see the experience of the body as the linchpin. Both pairs of scholars try to substantiate their claims by showing that the categories conform in outline to the experiences they posit as formative. The problem is that the categories, as predicates or molds of maximal generality, conform in outline to virtually all experience—making it difficult to adjudicate between different claims to primacy. The deep differences in portrayals of the universal condition of all knowledge given to us respectively by Durkheim/Mauss and Lakoff/Johnson originate in different kinds of centric inclinations applied to the Aristotelian categories; by the former, the categories are sociomorphized, by the latter, they are (in the narrow sense of the term) anthropomorphized.11 Viewing mythological or folk portrayals of the cosmos as projections of local frames of understanding and practice is a rather standard anthropological move. Even the most vocally anti-positivistic of contemporary social theorists will often quietly accede to the Durkheimian/Freudian dictum that, whatever its professed object, the “real” object of mythico-religious cosmological discourse is humans and their inner-worldly concerns. Part of the force with which the cosmological “projectionist” thesis took hold within ethnography at the turn of the century had to do with the fact that amidst intense opposition on many other points, psychoanalytic schools (deriving especially from Freud) and sociological schools (deriving especially from Durkheim) were in fundamental agreement about where the true object of cosmological discourse is not to be found, namely, the heavens. For Freud the foundational experience was the generic individual psyche as situated within the dynamics of familial relationships, especially the relation of parent and child;12 for Durkheim it was the experience of the social collectivity through its preeminent structures, symbols and rituals. Freud and Durkheim together form a sort of Copernican revolution within ethnology, in the sense that they charge this enterprise with laying bare the all-too-human desires and designs that impinge on cosmological discourse. Now to the specifics: I will compare what Durkheim and Mauss, and Lakoff and Johnson have to say regarding the origin of the categories. Since both pairs of scholars have written a great deal, my discussion is selective, aiming to bring out the general tenor of both positions as well as some pertinent sub-themes within each. Durkheim and Mauss’s perspective can be found in two condensed statements from Durkheim’s Elementary forms of the religious life: At the roots of all our judgments there are a certain number of essential ideas which dominate all our intellectual life; they are what philosophers since Aristotle have called the categories of the understanding: ideas of time, space, class, number, cause, substance, personality, etc. They correspond to the most universal properties of things. They are like the solid frame which encloses all thought; this does not seem to be able to liberate itself from them without destroying itself (1965: 21–22). How does Durkheim explain the origin of these categories? The category of class was at first indistinct from the concept of the human group; it is the rhythm of social life which is at the basis of the category of time; the territory occupied by the society furnished the material for the category of space; it is the collective force which was the prototype of the concept of efficient force, an essential element in the category of causality. However, the categories are not made to be applied only to the social realm; they reach out to all reality (1965: 488). Let us turn for a comparison to Lakoff and Johnson; for present purposes it will be most useful to focus on two works, Johnson’s The body in the mind (1987) and Lakoff and Johnson’s Philosophy in the flesh (1999). These works present an account of the origin of categories that is largely devoid of reference to the sort of social or collective forms that play the dominant role in Durkheim and Mauss (although, as we shall see, the family makes an appearance when Lakoff and Johnson turn from epistemology proper to moral philosophy). Lakoff and Johnson employ the concept of “image schemata,” defined by Johnson as “a recurrent pattern, shape, and regularity” which arises “chiefly at the level of our bodily movements through space, our manipulation of objects, and our perceptual interactions” (1987: 29). As “structures for organizing our experience and comprehension” (1987: 29), the function of the image schemata is roughly comparable with the functions that Aristotle and Kant, and Durkheim and Mauss following them, attributed to the categories. And, more importantly, some of the specific image schemata that Lakoff and Johnson focus upon are just those potential predicates that Aristotle, Kant, and Durkheim and Mauss call categories. To facilitate comparison I will focus on three such categories: class, cause, and time—in each case filling out Durkheim and Mauss’s arguments, then comparing them with the arguments of Lakoff and Johnson. Class Durkheim and Mauss’s most elaborate line of argument has to do with what one might call the morphological congruence between the idea of “society” and the idea of a “class.” One of the ways in which, they argue, any “class” of things is like a society is that one perceives in any class some sort of internal cohesion—some sort of affinity among the members of that class. The loftiest moments in Durkheim and Mauss’s writings occur when they address the issue of social sentiment, the collective “effervescence” that affectively binds together members of society. Such sentiments reach peaks of intensity in social rituals, fusing individuals into a single entity “this moral being, the group” (Durkheim 1965: 254). The experience of this all-embracing, all-powerful sentiment—this absolute basis of society—is seen by Durkheim and Mauss as the experiential model for the internal cohesion we ascribe to any logical “class” of things. Another way in which, for Durkheim and Mauss, any class of things is morphologically congruent with a society is that any class, like any society, implies a border. In an intriguing argument, Durkheim and Mauss point out that although logic has nothing per se to do with space, we nonetheless tend to diagram logical relations spatially (i.e., with what are now called Venn diagrams). Logical diagrams often involve circles within circles, just as tribal structures often involve sub-clans within clans. This proclivity suggests to Durkheim and Mauss that the tribal territory, border, and concentric structure are the experiential sources of the space, border, and concentric structure of formal logic. It is certainly not without cause that concepts and their interrelations have so often been represented by concentric and eccentric circles, interior and exterior to each other, etc. Might it not be that this tendency to imagine purely logical groupings in a form contrasting so much with their true nature originated in the fact that at first they were conceived in the form of social groups occupying, consequently, definite positions in space? And have we not in fact seen this spatial localization of genus and species in a fairly large number of very different societies? (1972: 83) They summarize: “Logical relations are thus, in a sense, domestic relations” (1972: 84). Let us turn now to see what Lakoff and Johnson have to say about the origin of the category of class. If Durkheim and Mauss find the model for the idea of any clearly-bounded class of things in the formative experience of being a member of a tribe or family, Lakoff and Johnson find it in the experience of one’s body—specifically of one’s body as a container or as an entity that deals with other containers. This bodily experience gives rise to what they call the “CONTAINER schema.” The formulations concerning this schema that offer the most dramatic contrast with Durkheim and Mauss are to be found in Johnson’s The body in the mind; Johnson writes: Our encounter with containment and boundedness is one of the most pervasive features of our bodily experience. We are intimately aware of our bodies as three-dimensional containers into which we put certain things (food, water, air) and out of which other things emerge (food and water wastes, air, blood, etc.). From the beginning, we experience constant physical containment in our surroundings (those things that envelop us). We move in and out of rooms, clothes, vehicles, and numerous kinds of bounded spaces. We manipulate objects, placing them in containers (cups, boxes, cans, bags, etc.). In each of these cases there are repeatable spatial and temporal organizations. In other words, there are typical schemata for physical containment (1987: 21). Paralleling the quest of Durkheim and Mauss, Johnson pushes further, claiming that the very principles of formal logic derive from just this bodily experience. His arguments, like those of Durkheim and Mauss, appeal mainly to morphological congruence. For example, Johnson argues that the P/ P formula of formal logic is congruent with the container schema that we derive from the bodily experience of boundedness; thus bodily experience is the likely source of formal logic, which extends and formalizes it: It follows from the nature of the CONTAINER schema (which marks off a bounded mental space) that something is either in or out of the container in typical cases. And, if we understand categories metaphorically as containers (where a thing falls within the container, or it does not), then we have the claim that everything is either P (in the category-container) or not-P (outside the container). In logic, this is known as the “Law of the Excluded Middle” (1987: 39). In Philosophy in the Flesh, Lakoff and Johnson add that the container-schema forms the basis of the philosophical notion of “essence”; their argument, like that of Durkheim and Mauss, implicates the space, border, and concentric structure of formal logic.13 For the sake of imposing sharp distinctions, we develop what might be called essence prototypes, which conceptualize categories as if they were sharply defined and minimally distinguished from one another.When we conceptualize categories in this way, we often envision them using a spatial metaphor, as if they were containers, with an interior, an exterior, and a boundary. When we conceptualize categories as containers, we also impose complex hierarchical systems on them, with some category-containers inside other category-containers (1999: 20). In Philosophy in the flesh, we encounter a further interesting twist in the appeal to morphological congruence. In this ambitious work, Lakoff and Johnson aspire to reconceptualize not just logic, but the history of philosophy from the standpoint of “the striking claim that the very structure of reason itself comes from the details of our embodiment” (1999: 4). They attempt to do this by revealing the image schemata at the basis of various philosophical doctrines. Aristotelian logic and theory of essences, they argue, are governed by the CONTAINER schema; their presentation begins, “Aristotle gave us the classical formulation of what we will call ‘container logic,’ which arises from the commonplace metaphor that Categories Are Containers for their members” (1999: 380). This is a novel and intriguing attempt at substantiating claims about the experiential source of logic. That is, in addition to demonstrating the morphological congruence between the experience of the body and the operations of logic, Lakoff and Johnson attempt to demonstrate morphological congruence by means of a reading, from their own perspective, of the philosopher credited with the invention of formal logic. To summarize: Durkheim and Mauss give us the in-the-tribe/outside-the-tribe origin of logic; Lakoff and Johnson give us the in-the-body/outside-the-body origin. Cause In summarizing Durkheim and Mauss’s views on the origin of the category of “cause,” I can do no better than to repeat a part of the long passage quoted above: “it is the collective force which was the prototype of the concept of efficient force, an essential element in the category of causality” (Durkheim 1965: 488). “Collective force” is the power of the group over the individual. For Durkheim and Mauss this power manifests itself most efficaciously not through physical coercion but through symbols and ideas that create and arouse a sui generis energy and sense of moral authority within a social collectivity—a dimension of consciousness that transcends individual interests. It is the feeling of the power of social sentiment and moral authority that is the experiential source of the idea of cause—or the notion that the things of the cosmos are not isolates but rather exert influence on one another. Let us now consider Lakoff and Johnson on the category of cause, which, paralleling Durkheim and Mauss, they refer to as force, or, more specifically, the “image schema of FORCE.” Contrary to Durkheim and Mauss’s affective, emotional, moral force emanating from the collective, Lakoff and Johnson find the experiential source of force in physical force. Johnson writes: Though we forget it so easily, the meaning of “physical force” depends on publicly shared meaning structures that emerge from our bodily experience of force. We begin to grasp the meaning of physical force from the day we are born (or even before). We have bodies that are acted upon by “external” and “internal” forces such as gravity, light, heat, wind, bodily processes, and the obtrusion of other physical objects. Such interactions constitute our first encounters with forces, and they reveal patterned recurring relations between ourselves and our environment (1987: 13). The phrase “publicly shared meaning structures,” near the beginning of this passage, offers an opportunity to clarify a point of potential confusion between the theories considered here. One does find numerous references to a social dimension for the image schemata throughout the writings of Lakoff and Johnson. But it is important to realize that the social dimension enters the picture for them in a very different way than it does for Durkheim and Mauss. For Durkheim and Mauss the formative experiences that give rise to the categories are all intrinsically social: they are various dimensions of the irreducible experience of social collectivity. For Lakoff and Johnson the formative experiences belong to the body: there is nothing to suggest that a body in isolation from society would not come up with the image schemata discussed above. Although, practically speaking, it would seem necessary that there be a second body to care for the infant body undergoing the formative experiences, the experience of a social relationship with that second body is not portrayed as otherwise intrinsically necessary for generation of the image schemata. The category of class, for example, derives preeminently from the experience of the first body in itself as a container; along with this are mentioned environmental interactions that apparently need not be with animate beings. A necessarily social dimension enters Lakoff and Johnson’s perspective not on the level of formative experience, but in the claim that at least a part of embodied experience is universal: it is the universality of bodily experience that gives rise to publicly shared meaning structures. Lakoff and Johnson’s main discussion of “cause” in Philosophy in the flesh is found in their Chapter Eleven (“Events and Causes”) (1999: 170–234); they address the issue of whether the multiple senses of cause yield an underlying unity; and they answer in the affirmative. In the opening part of their presentation they say: At the heart of causation is its most fundamental case: the manipulation of objects by force, the volitional use of bodily force to change something physically by direct contact in one’s immediate environment. It is conscious volitional human agency acting via direct physical force that is at the center of our concept of causation. . . . Prototypical causation is the direct application of force resulting in motion or other physical change (1999: 177). In closing their discussion of cause, they provide a final terse example: The central prototypical case in our basic-level experience gives us no problem in answering the question. He punched me in the arm. He caused me pain. Yes, causation exists (1999: 233). To summarize these theories of the origin of the category of “cause”: for Lakoff and Johnson, the origin lies in the experience of physical force exerted by or against the body; for Durkheim and Mauss, the origin lies in the sense of a sui generis transcendent energy or power felt in collective milieux, eminently in social ritual. Time Argument by morphological congruity takes an interesting turn in both Durkheim and Mauss and Lakoff and Johnson, in both cases involving an appeal to the reader to engage in introspection. Compare the following two passages on the category of time, the first by Durkheim, the second from Lakoff and Johnson: Try to represent what the notion of time would be without the processes by which we divide it, measure it or express it with objective signs, a time which is not a succession of years, months, weeks, days and hours! This is something nearly unthinkable. We cannot conceive of time, except on condition of distinguishing its different moments (Durkheim 1965: 22). Lakoff and Johnson: Try to think about time without thinking about whether it will run out or if you can budget it or are wasting it.We have found that we cannot think (much less talk) about time without those metaphors. That leads us to believe that we conceptualize time using those metaphors. . . . What, after all, would time be without flow, without time going by, without the future approaching? (Lakoff and Johnson 1999: 166) The two accounts are so parallel that one wonders which is the case: that Lakoff and Johnson have unwittingly reinvented Durkheim’s rhetoric, or that the latter regard the former as so broadly known that readers can be trusted to recognize the allusion. Consider what happens in these parallel accounts. In each, the appeal to the inconceivability of purely abstract time leads the analysts to propose a more concretized portrayal—a portrayal of measured or differentiated time—but in each case with a selected version of concretized time, one that supports the larger claims of the particular authors. Durkheim slips in the calendar, that is, socially constituted units such as weeks and days, which lead directly into his overriding concern with social/sacred ritual; whereas Lakoff and Johnson slip in metaphors and conceptualizations (“budgeting” of time, “flow”) whose primary contexts are typically a generic individual negotiating quotidian life. The most abstract and fluid of all concepts, “time” is entirely congruent with either concretization. How it all began Continuing with the temporal theme, but at another level of analysis, it is important to note that, in arguing for their respective versions of experiential origins of the category, Durkheim and Mauss and Lakoff and Johnson both focus on the morphological congruity between the structure of the category, on one hand, and, on the other, their favored realm of experience: the social collective for Durkheim and Mauss, the generic body for Lakoff and Johnson. But these structural arguments lead into theories of origin that invoke or allude to a variety of temporal or quasi-temporal frameworks, including the individual’s psychic ontogeny, the history of Western academic thought, the social evolution of human institutions, and the biological evolution of the human species. Often several temporal frameworks are alluded to at once, with a suggestion of the recapitulation of similar patterns within different time scales. These quests for temporal founding moments sometimes have the feel of a quest for an origin myth. As if to make up for not offering a full-blown, heroic origin myth of the kind we find in Hesiod or the Maori, the formulations of Durkheim/Mauss and Lakoff/Johnson are myth-like in two different ways: first, as discussed above, in that they project, by mediation of the category, a particular local image onto the cosmos; and secondly, in that they attempt to give us some version, even if vague, of a temporal moment of origin—the moment when that particular image came to be so favored. For example, along with overtures to Darwinian biological evolutionism, one encounters in Lakoff and Johnson allusions to ontogenic development; some of the clearest occur in the passages from Johnson’s The body in the mind that were cited above: We begin to grasp the meaning of physical force from the day we are born (or even before) (1987: 13). And: From the beginning, we experience constant physical containment in our surroundings (those things that envelop us) (1987: 21). But just what is our experience of force and envelopment before we are born, that is, the experience of the womb? Is it the body’s first experience of itself as a container (perhaps a container interacting with another container)? Or does the womb provide the first experience of genealogy, encompassment by tribal territory, and the individual’s dependence on the social? Such appeals have mythical dimensions; they attempt to construct a plausible first experience and are necessarily speculative. Even “empirical” research into child development cannot escape the necessity of inferring the experience of those who are as yet unable to articulate it. Throughout the history of social thought, the minds of infants have furnished an irresistible object of learned projection, an ontogenic alternative to the mythical/historical first caveman stepping out of the cave. The fact that infants are like us yet different—ourselves at a distant, less developed stage—together with their limited capacity of responding (at least on such matters as the relative priority of physical versus moral experience) places them in the company of other favored objects of projection, including ancient ancestors, culturally exotic “others,” nonhuman animals, and celestial bodies. Durkheim and Mauss adopted the framework provided by the socio-cultural evolutionism of their day as their arena for speculating on category origins; and this framework, too, was inclined toward the position that the moral and intellectual development of the individual recapitulates human socio-cultural evolution. In assessing the possible arguments for one view or the other, it is important to mention an asymmetry that might seem to favor the Lakoff/Johnson position. Sometimes Lakoff and Johnson are, and sometimes they are not, careful to distinguish the several possible senses of their mantra, “embodied.” Even among committed Cartesian dualists I suspect one could find some who would assent to the proposition that all cognition is embodied in some sense. To the extent that one accepts that the site and mechanism of the categories must be physical/neural structures, the categories must be embodied; and in Philosophy in the flesh and other works Lakoff and Johnson offer glimpses into contemporary neural theory and how it might relate to their theories of metaphor—although the relation is presently so distant that the former realm of theory sheds little light on the latter.14 constraints atsome point would have undercut a function that surely figured in the evolution of the sensorium: that of successfully informing the body about the world outside of itself. On this point—the world outside—Durkheim and Mauss if anything have the edge, positing a more clearly relational experience, rooted in the collective, as generative model. If Lakoff and Johnson favor ontogenic appeal and Durkheim and Mauss cultural-evolutionary appeal, they converge in the dream of utilizing their respective theories about the origin of categories to put all of academia on a new footing—a more modern version of the ambition encapsulated in Kant’s famously arrogant title, Prolegomena to any future metaphysic. Not only will they show the necessity of their perspective, they will enshrine it at the base of the intellectual world, structurally and temporally. The crown or regina scientiae will go to the discipline that succeeds in accounting for the origin of the categories. Both pairs—Durkheim and Mauss, and Lakoff and Johnson—specifically focus on philosophy, acknowledging the historical prestige of this discipline while challenging what they see as an inclination toward a view of reason as pure. Both pairs of contenders seem to think that Aristotle got the categories formally about right; and that what is lacking is a proper accounting for their origin and development. Both Durkheim/Mauss and Lakoff/Johnson also envision a major rewriting of the history of philosophy from their perspectives, a rewriting that begins with a new account of the origin and nature of the fundamental categories. For Durkheim and Mauss, the Aristotelian categories offered an organizational plan for comparative sociology. The history of the Aristotelian categories would be cast as the continuation of a sort of pre-history of the categories, in which each Aristotelian category would be seen to be prefigured in the social forms and practices of tribal societies. From tribal classifications and practices it would be possible to reconstitute the social ground and prehistory of the category of class; from food customs, the category of substance; from exchange customs the category of relation and so on.16 Calling attention to perduring questions about the origin of Aristotelian categories, Durkheim and Mauss conclude their investigation this way: As soon as they are posed in sociological terms, all these questions, so long debated by metaphysicians and psychologists, will at last be liberated from the tautologies in which they have languished. At least, this is a new way which deserves to be tried (1972: 88). The social evolutionism that undergirds Durkheim and Mauss’s plan—in which contemporary tribal societies are thought to offer something like the prehistory of the Aristotelian categories—is not found in Lakoff and Johnson; indeed such social evolutionist doctrines have by now been generally discarded. Yet Lakoff and Johnson’s plan is in many respects parallel to Durkheim and Mauss’s: Lakoff and Johnson seek the basic character of the Aristotelian categories in non-technical, everyday understandings of the world, as revealed in contemporary popular metaphors. Lakoff and Johnson then analyze the history of such categories in Western philosophy as a continuation and special case of such everyday understandings. They account for different philosophical schools through the different embodied image schemata and metaphors that these schools choose to emphasize—in a word, this is the project of Lakoff and Johnson’s Philosophy in the flesh. In sum, both Lakoff/Johnson and Durkheim/Mauss aspire to rescue and reframe Western philosophy by reconceptualizing its root doctrines from the standpoint of their respective versions of the experiential origins of reason, emphasizing especially the great founding figure of Aristotle. These two projects—Durkheim/Mauss’s and Lakoff/Johnson’s—overlap in opposing the view of philosophy as a pure, self-contained discourse, unencumbered by the constraints of the mundane world. That is, Durkheim/Mauss and Lakoff/Johnson both insist that the basic conceptual apparatus of higher-level thought—philosophy, logic, science and math—grow out of, and never entirely sever connections with, everyday knowledge and practice—inflected, by Durkheim and Mauss, as our pre-scientific, tribal past, and by Lakoff and Johnson, as contemporary everyday understandings of our life in the world. It is important to note that at this level Durkheim/Mauss and Lakoff/Johnson are allied against the idea of “pure” reason. But there is not just one form of everyday understanding; and we see a definite selectivity operating in the two projects, even though both choose data that happen to be of special interest for mythologists and folklorists. For their examples of pre-scientific, everyday understandings Durkheim and Mauss draw on myths and rituals from tribal societies, which they considered to be models of coherent social integration; while Lakoff and Johnson emphasize everyday metaphors (many of which have a proverb-like flavor) primarily drawn from their own society. Both in terms of the kind of societies involved (“tribal” in contrast to “modern”), and in terms of the speech genres that are considered (myth and ritual in contrast to everyday metaphor) there is a link between what the scholars would like to show (collectivity in contrast to the body) and the kinds of everyday understandings they have selected as their data. Myths and rituals, the data that Durkheim and Mauss draw upon, have a social orientation and functionality—they portray and uphold the overall worldview and social structure of the societies in question. By contrast, Lakoff and Johnson, in surveying the range of metaphors, particularly emphasize everyday phrases and proverbs or proverbial sayings of the kind relevant to any individual’s negotiation of the everyday world in pursuit of specific purposes: time is money, budget one’s time, life is a journey, and so on. It is worth noting that studies by folklorists of proverb and proverbial speech typically call attention to the pragmatic character of these folk genres; though proverbs are social in the sense of being broadly disseminated among members of a society or language group, their advice is typically directed to, and consulted by, an individual facing a concrete dilemma. There is of course a major quirk in all of this, specifically that many images can be used to reference either the social collective or the generic individual depending upon how one reads them. For example, Lakoff and Johnson, in their broad survey of metaphors, come upon instances of the metaphor whose source image for Durkheim and Mauss served as exemplar of the collective—specifically, the family and kinship. But when this happens, Lakoff and Johnson invariably read these in terms of the generic body. The most striking example occurs in chapter 14 (“Morality”) of Philosophy in the flesh. Here, Lakoff and Johnson turn to the metaphor of the “family” as the overarching metaphor that binds particular moral metaphors into a “coherent moral view” (1999: 312), asserting that this metaphor, enlarged to “The Family of Man,” provides the basis of universal morality. Since the metaphor projects family moral structure onto a universal moral structure, the moral obligations toward family members are transformed into universal moral obligations toward all human beings (1999: 317). In this passage, the origin of universal morality is said to lie in moral obligations that belong to family structure. The passage is highly reminiscent of Durkheim and Mauss, positing the experience of the immediate social collective as the ground of universal moral order—a moral order which is projected outward from the first and primary experience of moral structure (the “family” for Lakoff and Johnson, the “tribe” for Durkheim and Mauss). At this point, Durkheim and Mauss see the matter as finished, but Lakoff and Johnson do not, for their larger perspective demands that the body be the generative foundation, not just for categories but for morality, whether local or universal. For them, universal morality reduces in its principle to family morality, which in turn must be subjected to another reduction, which they attempt in the following passage: Our very idea of what morality is comes from those systems of metaphors that are grounded in and constrained by our experience of physical well-being and functioning. This means that our moral concepts are not arbitrary and unconstrained. It also means that we cannot just make up moral concepts de novo. On the contrary, they are inextricably tied to our embodied experience of well-being: health, strength, wealth, purity, control, nurturance, empathy, and so forth (1999: 331).17 Note that this list proceeds from qualities (health, strength) most easily attributed to an individual body, to those (nurturance, empathy) that involve a relation between bodies. This ordering is the equivalent, in the realm of morality, to the ordering of physical experiences offered in the passage quoted above in which Johnson offers “our bodies as three-dimensional containers” as his first example of the embodied experience of containment, and then moves outward to offer other examples of embodiment which involve our relationships with things outside of ourselves such as vehicles and other containers. In both realms—morality and physical experience—we start with a bio-psychological sense of body as sentient organism and “self” and we move outward into a world of bodies in interaction. Apparently, either end of such continua counts as “embodied.” But the fact that what is entailed in “embodiment” is so stretchable poses a vexing point about metaphor in general, and about the relation of Lakoff and Johnson to Durkheim and Mauss. Specifically, one with the will to do so can bring virtually any two terms into a metaphorical relation: metaphorizing means making a choice—in many cases, a political choice—from a pool of possibilities. We certainly could, in the manner of the examples above, extend the notion of embodied experience to include the experience of containment within the tribal territory. And if embodied experience includes such qualities as nurturance and empathy, why not say that the experience of collective effervescence in ritual, which for Durkheim epitomizes the irreducibly “collective,” is an “embodied” experience? By this route we could arrive at the conclusion that the difference between Lakoff/Johnson and Durkheim/Mauss is non-existent. But such a conclusion is contrived: the very ordering of Lakoff and Johnson’s litanies suggests a hierarchy of what counts as embodied. The body’s awareness of itself as container, or of its own strength and health, are offered first; however far the notion of embodied experience may be stretched, the stretching is outward from an anchored beginning point. The founding and final message that comes through, at least when one considers the full corpus of Lakoff and Johnson’s work is: the body.18 The anchoring point and final message in Durkheim and Mauss is: society. Yet it is interesting to note how it is possible to start with a certain frame, the body for Lakoff and Johnson, and draw in other frames, such as the collective dimension of kinship. Since the body and its genealogy are mutually implicating, analytical passage from one to the other is always possible—no less in theorizing about the category than in traditional mythology: recall the Maori mythic cosmos described above, in which the cosmos is portrayed at different moments as both a genealogy and a unitary body. To Lakoff and Johnson’s referral of kinship to the body, moreover, one finds a fascinating reciprocal in one of Durkheim’s students, Robert Herz, who finds a way to refer the overtly bodily to the irreducibly collective. Specifically, Hertz adopted the same imperative—to reveal the collective nature of human life and thought—embraced by Durkheim’s other students. But unlike those other students, who chose topics with plainly evident collective dimensions—such as social rituals and forms of ritualized exchange—Hertz landed a topic whose immediate referent is the body as physical organ: the asymmetry that favors one side of the body over the other (in humans usually the right over the left). To the given physical asymmetry of right and left, Hertz argued, nearly every society has added a collective symbolic dimension—in the form of connotations of sacred for the right and profane for the left (obvious examples are visible in derivatives of the Latin terms dexter and sinister). That the body is given in nature as asymmetric is beyond human volition; but that this given condition of the body is universally seized upon as a moral symbol is a matter of collective sentiment. Hertz suggests that “[i]f organic asymmetry had not existed, it would have had to be invented” (1973: 10).19 Thus, we see in Hertz’s reading of the social collective in the body something like the reciprocal to Lakoff and Johnson’s reading of the body in the genealogy. If (as noted above) one could deny the differences between Lakoff/Johnson and Durkheim/Mauss only at the cost of considerable contrivance, it must also be acknowledge that considerable contrivance went into the creation of those differences. Creative incorporation in both directions—toward the body in Lakoff and Johnson, toward the moral collective in Durkheim and his students—is a fascinating business, illustrating the complexity one encounters in these heady claims about the ultimate ground of knowledge. In the end there may be no image that is irreducibly collective or bodily, yet images can be tweaked to conform to analytical—and of course, political—aspirations.20 What comes through respectively in Durkheim and Mauss, and Lakoff and Johnson, are very different origin myths of the category, which spill over into different theories and rhetorics of moral obligation. There are a number of interesting ways in which the worlds of Durkheim and Mauss and Lakoff and Johnson permeate their origin myths of the category. “He punched me in the arm. He caused me pain. Yes, causation exists.” Although Lakoff and Johnson’s little scenario falls well short of Freud’s famous primal patricide-cum-totemic meal (Totem and taboo )—an act of violence which, for Freud, marked the origin of human consciousness as we now experience it21 In the atmosphere of philosophical and economic individualism and social alienation that followed World War I, Durkheim and his students sought to prove the necessity of sociality to human life, and therefore of sociology as a discipline, by demonstrating that the possibility of any general concept rests upon the prototype experience of collective force that primordially combines the many into the one. In the early twentieth century, compelled by the vision of the failure of civilization, Durkheim looked into the Aristotelian/Kantian categories—those virtual forms, or empty molds—and in them he thought he saw the shape of a tribal camp, the palpable image of a coherent, functioning society. By contrast, in the late twentieth-century United States Lakoff and Johnson look into those same forms, and in their illustrations we see some of the common, prosaic terms of our world: the body, containers, getting into cars, getting jostled. Compared with Durkheim and Mauss’s, most contemporary readers would find Lakoff and Johnson’s account more compelling; and I suspect that among the reasons is the greater compatibility of the latter with elements of regnant worldview, including our contemporary obsessions with individualism, “personal space,” and “body image.” The possibility for conflicting interpretations regarding prototypical experiences for fundamental categories inheres in the very notion of fundamental categories. However Aristotle arrived at his ten, the process certainly involved a deliberate search for those mental molds of most general applicability. And here we return to the basic problem: that to the extent that these fundamental categories are a successful winnowing of the most general properties of things, and molds of universal applicability, there is nothing they are not congruent with. As congruency is one, if not the main, basis on which these pairs of scholars argue for prototypicality, we are left with no shortage of candidates for the experiential prototype of the categories. To the well-established anthropological research question of how different peoples impose their different categories on the world, we should add the question of how different category-theorists impose their different worlds on the category. Having focused on a comparison between the perspectives of Lakoff and Johnson and Durkheim and Mauss, I would like now to comment on what I see as a kind of aberration, or perhaps transformation, that begins before but crescendos after Philosophy in the flesh, in a series of popular political writings in which Lakoff ends up less at odds, than as convergent, with Durkheim and Mauss. Specifically, the role of family structure in the genesis of morality takes a new form in Lakoff’s political writings. In these he finds the difference between American conservative and liberal politics to lie in the contrast between two different models of family structure, which he terms, respectively, “Strict Father morality” (conservative) and “Nurturant Parent morality” (liberal). Lakoffs notion that different political philosophies arise from different ideal images of family structure not only starts with the collective dimension of human existence, but, more specifically, it recalls Durkheim and Mauss’s claim that different intellectual universes arise out of different social structures (or “social morphologies,” as Durkheim and Mauss called them). It is as though kinship, having shrunk from the tribe to the immediate domestic sphere in the modern world, rebounds back to the tribe in the underlying, energizing metaphors of national political debate. The argument about contrasting family models is put forward by Lakoff in a series of political writings ranging from works that interdisciplinarily inclined academics or serious general readers might consult, to tracts that would have special appeal for grass-roots organizers; these writings include Moral politics (1996), Don’t think of an elephant! (2004), Whose freedom? (2006), Thinking points (2006b), and The political mind (2008). It would be interesting to further explore the stylistic and strategic variations among these political works, but more to the point here are the qualities that unite them with one another and distinguish them as a group from earlier works by Lakoff and Johnson. There is a shift in perspective as Lakoff moves into political pundit mode. With one important exception, which I will discuss below, “the body” or “embodiment” as source of cognitive form and moral imperative—in other words, the linchpin of Lakoff and Johnson’s earlier cognitive theorizing—recedes to a nebulous background, giving over its here-is-what-to-remember urgency to the two models of the family and their influence on politics. Why the shift? One possible reason has to do with the fact that Lakoff’s intentions are now more applied than theoretical: the underpinnings may not be necessary for arguments aimed at the broadest possible political audience. Or it may be that the underpinnings are no longer the real underpinnings. I cannot help but wonder whether the body recedes because it is not the most fruitful source image for the political message that Lakoff seems to want to send: politics is about the relationship between and among people (between and among bodies, if you will), for which the image of a family provides a more potent generative bedrock—especially for the values of empathy and nurturance, which Lakoff puts at the top of the liberal (Nurturant Parent) political philosophy that he wishes to promote, and lower in the conservative (Strict Father) political philosophy, which he wishes to challenge.23 This is speculative, of course, but not more so than the theories about category origins we have been considering. In part just because of the speculative character of their ventures, it is a stretch to see in either Durkheim and Mauss or in Lakoff and Johnson a scientific theory of the origin of categories. What we encounter is more on the order of origin myths, full of brilliant invention and experiential appeal, while infused with scientific rhetoric and the politics of the academy and the nation. The earliest recorded origin myths (the Babylonian Enuma Elish, for example, which served to legitimate the rule of the political paramount) as well as the earliest philosophical analyses of myths (Plato’s Republic is the perennial example) are intensely political; there is no reason to suppose that contemporary forms of mythologizing should be any different. The exception to the recession of the body in Lakoff’s political writings lies in his analysis of a value that, in the earlier analyses considered above, does not figure among those ascribed to the body or to either the Strict Father or the Nurturant Parent family, namely, the value of freedom. Lakoff’s analysis of freedom is developed most fully in Whose freedom? The battle over America’s most important idea (2006), in which he claims that “the idea of freedom is felt viscerally, in our bodies, because it is fundamentally understood in terms of our bodily experiences” (2006: 29). Following the procedure laid out in Philosophy in the flesh, Lakoff begins with a discussion of the bodily experience of freedom, emphasizing especially freedom of motion, and then proceeds to derive from it philosophical inflections of the idea of freedom, emphasizing especially freedom of will and its framing within the tradition of faculty psychology: In the Enlightenment, there was an elaborate metaphorical folk theory, called faculty psychology, in which the mind was a kind of society, with members who were individuals with different jobs. Among the members of the society of mind were Perception, Reason, Passion, Judgment, and Will. Perception gathered the sense data from the outside world; Reason figured out the consequences; Passion was the locus of desire; Will controlled action; and when Passion and Reason were in a standoff as to what Will should do, Judgment made the decision (2006: 33). What is important about the “society of mind” metaphor in Lakoff’s political theorizing is that it and the related metaphor of the “body politic” (2006: 36) open a direct route from politics to the body. This new route, based on the image of political society as a system of mutually-sustaining parts analogized to the parts of a body or a mind, bypasses the social collectivity with most immediate impact on the formative experiences of maturing individuals in American society, namely the family. This new route thus might seem to preclude, or at least evade, a possible Durkheim-inspired alternative explanation of the experiential origin of the concept of freedom. Immediately following the sentence, quoted above, announcing the rootedness of freedom in bodily experience, Lakoff says this: The language expressing the metaphorical ideas jumps out at you when you think of the opposite of freedom: “in chains,” “imprisoned,” “enslaved,” “trapped,” “oppressed,” “held down,” “held back,” “threatened,” “fearful,” “powerless.” We all had the experience as children of wanting to do something and being held down or held back, so that we were not free to do what we wanted (2006: 29). Note that most of these opposites of freedom are terms whose first referent is human social relations, and that all of them have social relations as at least one possible referent. The terms are accompanied by a “just-so” story, locating the birth of the idea of freedom in the child’s experience of its opposite, that is, constraint. The fact that Lakoff gives us a list of qualities denoting malignant, freedom-denying forms of social relations, opposing these to the body’s visceral experience of freedom, stacks the argument in favor of the body and against society. But one could easily reverse all of this, as many thinkers have, and portray the body as a source of desires that, unmitigated, work against the ideal of the well-being of all. Society, by contrast, is the source, means, and guarantor of freedom: it offers the possibility of enacting protections against unscrupulous individuals, mechanisms to adjudicate grievance, and, in language and accumulating cultural traditions, sources of intellectual and aesthetic stimulation and choice through which individuals make themselves into distinctive and effective persons. Lakoff’s origin story about the child’s first experience of constraint could be recast as the child’s primordial experience of a protective nurturance that guarantees freedom from fear and the chance to mature. Much of this theme—that is, the freedom-enhancing side of protective social constraint—can actually be found in Lakoff’s fuller analysis of freedom,24 and it all fits rather well with the Nurturant Parent liberal political model that he advocates. Once again, there is an analytical choice of what to single out as experientially primary. But to go through Lakoff’s arguments concerning freedom and counterpose a full Durkheimian alternative—that is, an account of the origins of freedom in irreducibly collective experience—would merely net another rehearsal, in a new register, of what has already been said above in the context of space, time, and causality. Some cosmological projections may in the end have little to do with cosmology in any direct sense. At the present state of the art, our most potent and encompassing projections may be those that take the form of a claim that certain experiences are the prototypical source of categorical structure in general and thus the framework of all knowledge. Our presently most potent form of projection, in other words, is perhaps not the experience we project onto all others, but rather the experience that we project as the experience that we project onto all others (for novelty’s sake, I shall avoid labeling this as meta-projection). Rejection of the Kantian tendency to assume that the synthetic principles of cognition are inscrutable in favor of the view that they have a discoverable origin, results in these principles themselves, and even the generic structure of the category—such a prize these now are!—turning into a projective screen for the world as we know it as well as our hopes, dreams, and desires for it—a screen as alluring as the starry sky above. Surviving and Prevailing: The Wittgensteinian Revolution In the foregoing analysis I have focused on issues connected with one of two main usages that theorists employ when they refer to Aristotelian categories; one usage designates a list of highly-abstract, maximally-applicable predicates or frames (“time,” “space,” “cause,” etc.) judged by Aristotle as intrinsic to the making of propositions; the other designates one of the specific frames that inheres in Aristotle’s list: the notion of a “class” or “set” (or “category” in a narrower sense). In category theory these two usages often flow into one another. By way of a complement to the foregoing, I will offer in this final section a brief excursus into a third main usage of the term “Aristotelian” in relation to categories. A category is Aristotelian in this third sense when it is a category (or set or class) that admits no gradation of membership and thus operates binarily: any entity either is or is not a member of a particular category. This is the type of “category” used by Aristotle in his logic (and hence its designation); it corresponds to what Lakoff and Johnson in Philosophy in the flesh call an “essence” prototype.25 In recent times a good deal of attention has been paid to sets that allow graded membership and “fuzzy” boundaries and which are thus different in principle from such binary or “Aristotelian” categories. Particularly influential in promoting the recognition and importance to human cognition of non-Aristotelian categories has been the work of Eleanor Rosch (1978), though founder’s credit is often accorded to the philosopher Ludwig Wittgenstein, who wrote: Consider for example the proceedings that we call “games”. I mean board-games, card-games, ball-games, Olympic games, and so on. What is common to them all?—Don’t say: “There must be something common, or they would not be called ‘games’”—but look and see whether there is anything common to all.—For if you look at them you will not see something that is common to all, but similarities, relationships, and a whole series of them at that. . . .I can think of no better expression to characterize these similarities than “family resemblances”; for the various resemblances between members of a family: build, features, colour of eyes, gait, temperament, etc. etc. overlap and criss-cross in the same way.—And I shall say: “games” form a family (1968: 31–32). In an earlier work, Lakoff (1987), taking his cue from Wittgenstein and championing the work of Rosch, has portrayed the contrast between the category based on “family resemblances” and the clearly bound Aristotelian category as the basis for a major revolution in cognitive science.26 But I am intrigued by the substrate that survives this purported revolution—for once again both antipodes implicate the image of a family. Aristotle’s works on logic and classification are laden with the terminology of kinship (starting with genos or kin/family, rendered in Latin and English as “genus”), yet when Wittgenstein offers an alternative to Aristotle the alternative too is portrayed through an image of kinship, that is, “family resemblance.” Another dimension of the appeal to kinship in the image of “family resemblance” confirms a continuing relationship between kinship and categories in recent times. Specifically, theoretical developments within the comparative study of kinship systems—a specialization that declines and then recrudesces within social and cultural anthropology—have undergone some theoretical turns that parallel the broadening of theories of categorization inspired by Wittgenstein and Rosch. The boom in comparative kinship studies in the mid-twentieth century was given impetus by the idea promoted by Durkheim that kinship systems are for societies what skeletal structures are for living organisms; as organisms are scientifically classified and compared through skeletal types, societies can be classified and compared cross-culturally through kinship-system types. A.R. Radcliffe-Brown (1965) adopted this idea and developed an influential set of methods and theories about kinship, for the most part assuming that kinship systems would be unilineal (that is, based on tracing descent dominantly through either the male line or the female line). Partly as the result of the arguments of Polynesianists, notably Raymond Firth (1963; cf. Schwimmer 1978 for Maori), kinship studies more recently have accepted the possibility that the principles on which kin-groups are based may not be as neatly unilineal as once assumed: individuals are often found to claim filiation of great genealogical depth from both parents (sometimes compounding at ascending generations)—making clan membership fuzzy, negotiable, and shifting. This realization appeared just as cognitivists were recognizing that folk categories in general often allow inclusion through a plurality of routes rather than by binary criteria. Thus in these two realms—cognitive theories of folk categories, and anthropological theories of kinship—the findings converged: humans create classes characterized by flexibility, gradation of membership, and “fuzzy boundaries.” It seems that as goes theory of kinship, so goes theory of the category—or vice versa. There are some interesting moments in these kinship debates. One of these concerns anthropologist Claude Levi-Strauss’s discussion of the “house” in the context of kinship; “house” makes reference to a seeming kin group constituted through several inconsistent modes of filiation. As noted above, under the formative influence of Durkheim’s work on kinship, elaborated by A.R. Radcliffe-Brown, the anthropological study of kinship had come to be dominated by the ideal of the clearly bounded Aristotelian category, or in other words, the assumption that all kinship groupings necessarily follow binary Aristotelian logic (any individual is a member of either this clan or that). Considering (as discussed above) that Durkheim wanted to show that Aristotelian logic derived from the forms and practices of social life—for him epitomized in the tribal genealogy—one can see why: if Aristotelian logic is to be traced to tribal kinship practice, then tribal kinship had better be Aristotelian in principle!27 As noted above, anthropologist Claude Lévi-Strauss caused a flurry—within the rarefied milieux of kinship theory—by invoking as a kinship term a concept that relates to kinship only metonymically—the term “house”—in order to designate a type of social grouping, encountered in various parts of the world, which is based on several different modes and degrees of filiation rather than a unilineal principle of descent. It is surely an indication of the degree to which anthropological studies of kinship had been skewed in an Aristotelian/binary direction that Lévi-Strauss resorted to a term with no intrinsic connection to kinship (i.e., “house”) in order to capture the possibility of fuzziness that to the naive outsider—the philosopher Wittgenstein—seems to have been the very soul of kinship—why else would the latter have adopted “family resemblance” as the emblem of classificatory fuzziness?28 Another quirk can be found in anthropologist David Schneider’s colorful and radical critiques of the anthropological study of kinship (1968; 1984). Schneider was so struck by the variability of so-called kinship systems (and by theoretically anomalous points, some of them quite charming, such as the insistence of some informants that the dog was a member of the family) that he was led to question the very existence of this particular category of ethnographic phenomena—wondering whether kinship was an ethnographers’ illusion. But it is important to note, once again, that this skepticism developed in the context of an anthropology predisposed to search for an Aristotelian category—an Aristotelian essence of kinship. One wonders whether it would have made a difference if Schneider had, instead, approached the variability of kinship systems in terms of the possibility of discovering areas of graded overlap in the manner of Wittgenstein’s “family resemblances.” That would have been an interesting situation: kinship salvaging itself as a topic of theoretical discourse by serving as its own naive metaphor. This position has much to recommend it; and it is partly in view of this possibility that, in the face of Schneiderian skepticism, I still dare to think that kinship is a genuine topic of ethnographic exploration (even while, admittedly, remaining fuzzy about what I would expect to encounter within it). Schneider was no mythologist, yet other thinkers noticed the implications for mythology of his radical critique of kinship theory. Judith Butler (2000), for example, enlists Schneider’s critique in an influential and thought-provoking re-evaluation of the sagas of Oedipus and Antigone around their confused kinship. Though in contemporary Western society our most widely shared public cosmologies tend not to be heavily reliant on kinship images (with exceptions of course: contemporary scientific cosmology is replete with kinship metaphors: “baby universes,” “still-born universes,” and so on), we nevertheless have maintained, even in our most rationalistically inclined intellectual pursuits, numerous forms of speculation to the effect that the cosmos, or its building-block, the category, and kinship somehow go together; like Zeus, we are not able to let go of our kinship web. These speculations range from the claim that the cosmos or the category is a projection or formalization that abstracts the structure of kinship, to the recurrent, sometimes almost subliminal, expectation that kinship will serve as a reliable metaphor for the creation of order in general. From this point of view alone it should come as no surprise that Lakoff, in the most popularizing of his works, carries his theoretical speculations about the genesis of moral reasoning in kinship into the more specific claim that the political polarity in contemporary American politics is best grasped through contrasting models of familial order (the aforementioned “Strict Father” and “Nurturant Parent” models). As noted above, Durkheim and Mauss suggest that, formal structure aside, our very predisposition to approach the cosmos as though its entities are related in a web of mutual influence, rather than as discrete isolates, flows from the experience of being born into a human situation with this character.29 Finally, it is almost as though Kant anticipated Durkheim’s sociologizing move, on this point, by naming one of his fourteen a priori categories—specifically his category of reciprocal-causality—as “community” (Gemeinschaft), thus pre-positioning a moral concept for projective extension to the physical world. It is difficult to imagine any method that could definitively move any such ideas beyond the realm of conjecture or projection. But, given the constancy, beginning even with Copernicus (who led us out of our anthropocentrism) with which our intellectual abstractions have been attended by images of kinship, it would be equally difficult to find (should we want to) a method that would cause the latter images to definitively depart—a method that would, in other words, fully put to rest our Zeusian affliction. The borderline of the great Copernican divide, the point at which we learned to think of the universe scientifically rather than mythically—as a mechanism rather than as the rest of our family—is, logically and temporally, a fuzzy one. Fuzziness as an intellectual mood has, as one of its grammatical manifestations, double-negative constructions, for it is somehow less committal to say that “X is not without” than to say that “X is”. I therefore find it not inappropriate to conclude with a choice Lacanian triple-negative: “Not that what isn’t Copernican is absolutely unambiguous” (1991: 7)—which I take to be a French critical theory way of saying that what we mean by Copernican may not be entirely clear. Bibliography Allen, N.J. 2000. Categories and classifications: Maussian reflections on the social. New York: Berghahn Books. Aristotle. 1938. The categories. Translated by Harold P. Cooke. London: William Heinemann Ltd. Aveni, Anthony. 1994. Conversing with the planets. New York: Kodansha. Barrow, John. 1995. The artful universe. Oxford: Oxford University Press. Best, Elsdon. 1924. The Maori, Vol. I. Wellington: Harry H. Tombs ———. 1925. Tuhoe. Wellington: Reed. ———. 1976. Maori religion and mythology, Part 1. Wellington: Government Printer. Butler, Judith. 2000. Antigone’s claim. New York: Columbia University Press. Copernicus. (1543) 1978. On the revolutions. Baltimore: Johns Hopkins University Press. Curnow, Jenifer. 1985. “Wiremu Maihi Te Rangikaheke: his life and work.” Journal of the Polynesian Society 94: 97–147. Detienne, Marcel. 1986. The creation of mythology. Chicago: University of Chicago Press. Dietz Moss, Jean. 1993. Novelties in the Heavens. Chicago: University of Chicago Press. Dorson, Richard. 1965. “The eclipse of solar mythology.” In Myth: A symposium, edited by Thomas Sebeok, 25–63. Bloomington: Indiana University Press. ———. 1968. The British folklorists. Chicago: University of Chicago Press. Durkheim, Emile. 1965. The elementary forms of the religious life. New York: Free Press. ———. 1982. The rules of sociological method. New York: The Free Press. ———. 1984. The division of labor in society. New York: The Free Press. Durkheim, Emile and Marcel Mauss. 1972. Primitive classification. Chicago: University of Chicago Press. Firth, Raymond. 1963. “Bilateral descent groups: an operational viewpoint.” In Studies in kinship and marriage, edited by Isaac Schapera, 22–37. London: Occasional Papers of the Royal Anthropological Institute. Freud, Sigmund. 1950. Totem and taboo. New York: W.W. Norton & Company. ———. 1967. Moses and monotheism. New York: Vintage Books. Girard, René. 1987. Things hidden since the foundation of the world. Stanford: Stanford University Press. Hertz, Robert. 1973. “The pre-eminence of the right hand: a study in religious polarity.” In Right & Left, edited by Rodney Needham, 3–31. Chicago: University of Chicago Press. Hesiod. 1953. Theogony. Indianapolis: Bobbs-Merrill. Johnson, Mark. 1987. The body in the mind. Chicago: University of Chicago Press. Kant, Immanuel. 1965. The critique of pure reason. New York: St. Martin’s Press. (Based on the 2nd edition, published originally in German in 1787). Kardiner, Abram. 1945. The psychological frontiers of reality. New York: Columbia University Press. Kuhn, Thomas. 1985. The copernican revolution. Cambridge: Harvard University Press. Lacan, Jacques. 1991. The seminar of Jacques Lacan, Book II. New York: Norton. Lakoff, George. 1987. Women, fire and dangerous things. Chicago: University of Chicago Press. ———. 1996. Moral politics: how liberals and conservatives think. Chicago: University of Chicago Press. ———. 2004. Don’t think of an elephant! White River Junction: Chelsea Green Publishing. ———. 2006. Whose freedom? The battle over America’s most important idea. New York: Picador. ———. 2006b. Thinking points. New York: Farrar, Straus and Giroux. ———. 2008. The political mind: Why you can’t understand 21st century politics with an 18th century brain. New York: Viking. Lakoff, George and Mark Johnson. 1980. Metaphors we live by. Chicago: University of Chicago Press. ———. 1999. Philosophy in the flesh. New York: Basic Books. Lévi-Strauss, Claude. 1988. The way of the masks. Seattle: University of Washington Press. Malinowski, Bronislaw. 1944. A scientific theory of culture. New York: Oxford University Press. McManus, I.C. 2002. Right hand, left hand: The origins of asymmetry in brains, bodies, atoms, and cultures. Cambridge: Harvard University Press. Müller, F. Max. 1874. “Comparative mythology.” In Chips from a German workshop, Vol. II, 1–87. New York: Scribner, Armstrong, and Co. Needham, Rodney. 1972. “Introduction.” In Primitive classification, by Emile Durkheim and Marcel Mauss, vii–xlviii. Chicago: University of Chicago Press. ———, ed. 1973. Right and left. Chicago: University of Chicago Press. Numazawa, Franz Kiichi. 1984. “The cultural historical background of myths of the separation of sky and earth.” In Sacred narrative, edited by Alan Dundes, 182–92. Berkeley: University of California Press. Plato. 1981. Timaeus. Translated by Desmond Lee. Harmondsworth: Penguin. Prytz-Johansen, Jorgen. 1954. The Maori and his religion. Copenhagen: Ejnar Munksgaard. Radcliffe-Brown, A.R. 1965. Structure and function in primitive society. New York: The Free Press. Rosch, Eleanor. 1978. “Principles of categorization.” In Cognition and categorization, edited by Eleanor Rosch and B.B. Lloyd, 27–48. Hillsdale: Lawrence Erlbaum Associates. Schneider, David. 1968. American kinship. Englewood Cliffs: Prentice-Hall. ———. 1984. A critique of the study of kinship. Ann Arbor: University of Michigan Press. Schrempp, Gregory. 1983. “The re-education of Friedrich Max Müller.” Man 18: 90–110. ———. 1992. Magical arrows. Madison: University of Wisconsin Press. ———. 1998. “Mathematics and traditional cosmology: notes on four encounters.” Cosmos 14: 211–226. Schwimmer, Eric. 1978. “Lévi-Strauss and Maori social structure.” Anthropologica 20: 201–22. Spencer, Herbert. 1897. The principles of psychology, Vol. II. New York: D. Appleton and Company. Spiro, Melford. 1978. Burmese supernaturalism. Philadelphia: Institute for the Study of Human Issues. Taylor, John. 1991. Linguistic categorization. Clarendon: Oxford University Press. Turner, Mark. 1987. Death is the mother of beauty. Chicago: University of Chicago Press. Tylor, E.B. 1929. Primitive culture, Vol. I. London: J. Murray. ———. 1970. Primitive culture, Vol. II. Gloucester: Peter Smith. Wheeler, John. 1988. “Foreword.” In Barrow John and Frank Tipler, The anthropic cosmological principle, vii–ix. Oxford: Oxford University Press. Wittgenstein, Ludwig. 1968. Philosophical investigations. Oxford: Basil Blackwell. La parenté Copernicienne : un mythe d’origine pour la catégorie Resumé : Dans de nombreuses mythologies traditionnelles, la parenté constitue l’expression privilégiée de l’unité ainsi que de la diversité de l’univers. Dans la pensée « post-mythologique », des catégories conçues logiquement tentent de s’arroger le rôle cosmique de la parenté. Je compare deux exposés de la nature et de la genèse des catégories, celui de Durkheim et de Mauss d’une part, et celui de Lakoff et Johnson de l’autre. Aucun de ces exposés ne rompt le lien avec la mythologie, ou avec la parenté. De plus, la structure de la catégorie, tout comme la parenté, offre un mode de projection de l’être humain en tant qu’unité cosmique. À la déjà longue préoccupation anthropologique concernant la manière dont les humains imposent leurs catégories diversifiées sur le monde, devrait s’ajouter le souci des manières dont les théoriciens de la catégorie imposent leurs mondes diversifiés sur celle-ci. Gregory Schrempp is Associate Professor of Folklore and Director of Mythology Studies at Indiana University (Bloomington). He is author of Magical arrows: the Maori, the Greeks, and the folklore of the universe (1992). This essay will appear, in fuller form, as a chapter of The ancient mythology of modern science, by Gregory Schrempp, forthcoming from McGill-Queen’s University Press (Spring 2012). ___________________ *. This essay is dedicated to the memory of David Schneider who, in his courses in the anthropological study of kinship at University of Chicago, constantly challenged the idea of kinship as a system isolated from other dimensions of culture and worldview. David, in some ethereal sense I hope this will remove my grade of “Incomplete” in your course. 1. Jean Dietz Moss discusses this passage from Copernicus in the context of a rhetorical analysis: “His exercise of analogy here, glorifying and personifying the sun, is intended to induce his readers to shift their emotional allegiance from what he considered an erroneous explanation of the cosmos to a more accurate one. An elegant rhetorical appeal to the emotions such as this in a work devoted to science, although strange to us, certainly would not have surprised his audience” (1993: 1). I disagree with this assessment of the source of strangeness in the passage from Copernicus. The strangeness does not flow from its status as a work devoted to science—for works by present-day science writers, and especially popular works on cosmology, are replete with appeals to emotion and we do not find it strange. Whatever strangeness we feel flows rather from the fact that this is a passage from Copernicus, to whom we have assigned a definite mythic-hero role: that of leading us out of anthropocentrism. The solar myth that Copernicus cites is not anthropocentric in the sense of placing the earth in the physical center of the cosmos, but it is anthropocentric in a way that is equally important, namely, in the assumption that the design of the physical universe confirms human moral values and rules that are encoded in kinship and in religious belief. 2. Cf. Detienne’s (1986) discussion of the prejudicial attitudes that have often propelled the idea of “myth.” 3. On the solar mythology movement, see Richard Dorson (1965, 1968) and Schrempp (1983). 4. Although Polynesia is geographically very distant from the Mediterranean, numerous variants of the story of separation of sky and earth as primal parents of the cosmos occur between these two areas, suggesting that a distant historical connection between the Greek and Maori stories is possible. Numazawa’s (1984) survey of this myth gives some indication of its geographical distribution. 5. See especially Jenifer Curnow (1985). There is also a summary and commentary on this account in Schrempp (1992: 58–59). 6. Such genealogies raise important issues concerning what we mean by narrative. Genealogy might be thought of, in certain contexts at least, as a limiting case of narrative. Any continuum of human action can, in the recounting, be thickened with lavish detail or thinned to the minimum elements that are capable of conveying a sense of a continuum of world and action, i.e., the names of the main players listed in historical sequence. We might think of genealogy as a highly presupposing narrative—a narrative that, relying on the audience’s foreknowledge of details, pares the recital to a minimal image, which can always be fleshed out. Maori genealogies are often prefaced by and punctuated with images of organic development such as the growth and spreading of vines—not unlike the Western inclination to think of genealogies as forming “trees.” Numerous examples of Maori cosmogonic genealogies can be found in Elsdon Best (1976). 7. In some Maori accounts it is ambiguous whether the first being of the cosmos is single or dual (see Schrempp 1992: 60–74). 8. For Spencer’s response to the Kantians see especially his appendix “Our space-consciousness—A reply” in his The principles of psychology (1897: 651–669). More recently, John Barrow (1995: 28) also proposes to Darwinize Kantian epistemology. 9. In both Durkheim/Mauss and Lakoff/Johnson one finds recognition of both a universalist level and a culturally relative level of analysis. In both projects the universal level amounts to a set of categories reminiscent of Aristotle’s categories, but both projects also recognize culturally divergent shapings of fundamental orientations. For example, some societies are organized on the principle of “dual organization” or “moieties” (in which all members of a society belong to one of two macro-clans). Such societies tend to carry this principle of organization into classification in general—so that everything in the cosmos, not just humans, belongs to one of two cosmic clans. Durkheim and Mauss treat the principle of moiety organization as an epistemologically-generative social form that is relative to social morphology rather than as a human universal. For their part, Lakoff and Johnson approach “Time is money,” or more broadly the view of time as commodity, as a realization of time which not all societies share (1999: 163–64). Kant, Durkheim/Mauss, and Lakoff/Johnson are unified in the fact that they are all convinced of the existence of universals in human knowledge even while insisting that such knowledge in some sense is relative to the nature of the knower. 10. Indeed at some points in Primitive classification it is impossible to tell whether Durkheim and Mauss are talking about the cosmos or the category: they seem to think of the image of society as inscribed simultaneously onto the cosmos—as when a society portrays the spatial organization of the cosmos on the model of the organization of its sub-clans—and onto the form of the category. The data of myth and ritual favored by Durkheim and Mauss no doubt contributed to oscillation between the category and the cosmos as the object of “this anthropocentrism, which might better be called sociocentrism” (1972: 86). 11. The contrast here necessitates the use of “anthropomorphize” in the narrow sense, to mean the imposing of human body form. This differs from my general preference for a broad usage, in which anthropomorphizing could also include the projection, onto nonhuman objects, of intangibles such as human language or of forms that humans have created as extensions of themselves, such as villages. The problem in the present instance is that since humans create their societies, according to the broad usage sociomorphizing would be part of anthropomorphizing rather than offering a contrast to it (cf. Durkheim and Mauss’s comment on “anthropocentrism” and “sociocentrism” in the previous note). As I suggest at a later point, the inelegant term “corpomorphic” might (as compared with “anthropomorphic”) be more to the point - although this usage too is ambiguous, since corpus is often cast as metaphor of the social (as in a “corps” or “corporation”). 12. We have, specifically, in Freud and in the various divisions of his followers, a great variety of formulations on the general theme that the structure, moral topography, and major “personalities” of the cosmos are projections of experiences based on one or another relation of familial kinship. The posited dynamics can be very intricate; for example in Moses and Monotheism Freud implicates the projection of stress points in familial relations not only in accounting for the shape of particular cosmological beliefs, but in accounting for a shift from one belief system to another: The ambivalence dominating the father-son relationship shows clearly, however, in the final result of the religious innovation. Meant to propitiate the Father Deity, it ends by his being dethroned and set aside. The Mosaic religion had been a Father religion; Christianity became a Son religion. The old God, the Father, took second place; Christ, the son, stood in his stead, just as in those dark times every son had longed to do (1967: 111). Freud’s general approach to cosmological projection has given rise to many variants, and it sometimes appears to be complemented by a sort of rebounding projection—of astronomical images into the sphere of psychoanalytical terminology and metaphor. Jacques Lacan is the exemplary case of this tendency—which can be seen for example in his discourses on the phallus and the meteor, why planets can’t talk, and the meaning of Freud’s proclamation of psychoanalysis as a Copernican revolution (e.g., 1991: 3, 13, 16, 224, 234–40). The tendency itself might be seen, from a slightly more distant perspective, as a new variation on an age-old fascination with the possibility of formal parallelism or mutual causal influences between heavenly constellations and human ones (see e.g. Aveni 1994). The American anthropological “culture and personality” movement of the mid twentieth century added to the psychoanalytic movement a heightened sense of cultural relativism by emphasizing the variability of psychodynamic “constellations” comprised in the child-rearing practices of different societies; such variability, it was argued, would be reflected in projective mechanisms, including cosmology (see Kardiner 1945; Spiro 1978). 13. The fact that Lakoff and Johnson here propose a specific “essence” prototype for the clearly bounded category, may have to do with previous work on “fuzzy” categories based on the principle that Wittgenstein dubbed as “family resemblance” (see especially Lakoff 1987). Grouping by “family resemblance” is a possibility that is never considered by Durkheim and Mauss in the pre-Wittgensteinian milieu. 14. As is the case with many works of cognitive science, the neural level is present as a sort of Promised Land toward which researchers are heading. The homage to neurons characteristic of Lakoff and Johnson is often found as well in the context of Artificial Intelligence research. For although in the latter endeavor computer emulation of intelligent processes offers a way of doing cognitive science without directly studying neurons, such efforts are sometimes accompanied by the assumption that successful emulations will turn out to have counterparts in neural structures. 15. I mean the projection of the human body as model onto the rest of the world. 16. On the place of Aristotle’s categories in the organizational plan of Durkheim’s school, see N.J. Allen (2000: 32–35 and chapters 5 and 6) and Schrempp (1992: 160–68). Unlike the American anthropology of the same period (in which each student tended to focus on a particular tribe or clan), in Durkheim’s program there was a tendency for students to choose one Aristotelian category and pursue it cross-culturally. 17. In Lakoff and Johnson’s collection of Metaphors we live by (1980) we find very few metaphors linked to kinship, and those we do find are similarly explored from the perspective of their relation to a generic individual. For example, birth metaphors (such as “our nation was born out of a desire for freedom”), which might easily be invoked to emphasize the conceptual depth at which we rely on kinship idioms, and hence a sense of the collective as a model for understanding the world, are assimilated instead to “a gestalt consisting of properties that naturally occur together in our daily experience of performing direct manipulations” (1980: 75). The same is broadly true of Mark Turner’s exploration of the kinship metaphor in Death is the mother of beauty (1987). 18. The differences between Durkheim/Mauss and Lakoff/Johnson as to the generative source of morality offer some interesting parallels to debates in the early twentieth century between Durkheim and anthropologist Bronislaw Malinowski on the perspective of “functionalism,” or the idea that social practices should be analyzed in terms of the contribution they make to maintaining a system. Durkheim thought that society should be the telos of the concept of “function,” while Malinowski (see especially 1944) thought that the well-being of individuals in society should be the telos. 19. Assessments of Hertz’s work are found in Needham (1973) and McManus (2002). The asymmetric dualism of the body is seized upon by Hertz to give voice to a deeper asymmetric dualism that runs throughout Durkheim and all of his students’ work, and towards which contemporary cognitive scientists are generally skeptical. Specifically, Durkheim sought to recast the age-old tradition of mind/body dualism sociologically—as the duality of collective vs. individual consciousness. He regarded the duality of individual and society as the most basic duality, the one that underlies the range of specific dualistic formulations found through the world—body vs. mind, matter vs. spirit, sensation vs. intellect, percept vs. concept, the symbolism of left and right. In Durkheim’s view, it is in the constitutive nature of society to draw such distinctions, always asymmetrically—one pole predominating over the other just as the social predominates over the individual. Epistemology and moral philosophy derive, according to Durkheim, from the same source: intellect towers over the senses, the concept over the percept, just as the social norm towers over individual desire. 20. That images are susceptible to such tweaking would be unsurprising to those who emphasize the “fuzzy” character of categorical boundaries; see the final section on the “Wittgensteinian Revolution.” 21. Freud’s scenario of the original patricidal totemic meal is put forward in Totem and taboo (1950: 140–41). See also Girard’s (1987) treatment of myths of founding violence, and my (1998: 215) comments on Girard. Lakoff and Johnson develop their claims about embodiment and moral reasoning at the level of a generic ego in its immediate kin relations, and only then do they project these onto larger social-political realms. This part of their approach recalls Freud. 22. Certainly there is a scalar difference between category and cosmos; moreover, the former is the medium through which humans act and interact in the immediate world, while the latter is a sort of abstraction from and summing up of that world and those interactions. But the differences of scale and abstraction are matters of degree, not kind: cosmos and category ultimately belong in the same discourse. 23. There is an interesting parallel between the way Lakoff describes the morality of the Strict Father family and the way that Lakoff and Johnson describe the morality of the body, which is another factor leading me to ask whether, in challenging the Strict Father political model, Lakoff may not also be distancing himself from the body as the source of political morality. Specifically, Lakoff puts strength at the top of Strict Father family values and empathy below these (1996: 379–83; cf. 2008: 77–82), a hierarchy that would seem to correlate with the order in which the moral values of the body are listed by Lakoff and Johnson: “health, strength, wealth, purity, control, nurturance, empathy, and so forth” (1999: 331; the full text is quoted above). “Nurturance” and “empathy” are last in this list, except for “and so forth.” In other words, it appears that body-morality more comfortably fits the Strict Father (conservative) political model that Lakoff rejects than the Nurturant Parent (liberal) model that he endorses. This assessment rests on the assumption that the list’s ordering implies a hierarchy or, as I have characterized it above, a movement out from a center; and I do think a reader could reasonably take away that impression. By contrast, a Nurturant Parent family, according to Lakoff, puts empathy and nurturance first (1996: 381). I am reading a lot from a little here, but this is unavoidable since Lakoff and Johnson give us little more than the passage quoted above on how to get from kinship morality to the body. 24. See especially his chapter 3 (“The logic of simple freedom”). 25. Lakoff (2008: 79–81) suggests a linkage of the binary Aristotelian category to the absolutism of Strict Father (conservative) morality, as discussed above. Seemingly by implication, fuzzy categories belong to Nurturant Parent (liberal) morality. 26. There is a terminological problem in whether one should refer to a Wittgensteinian grouping of family resemblance as a “category.” I do follow this usage, a practice which of course leads in the direction of setting up the concept of “category” itself as a fuzzy category. Various interrelationships have been suggested between the Aristotelian and Wittgensteinian category; John Taylor (1991: 68–74) for example presents interesting arguments to the effect that both necessarily have a place in our cognitive life. His nomenclature—”folk” vs. “expert”—may be unfortunate, however, contradicting many academic and nonacademic usages of both terms—including the realm of “high-tech,” where, for example, recent computer software has turned to “fuzzy” logic (thus, in Taylor’s terms, “folk” logic) in the design of so-called “expert systems.” 27. That Wittgenstein chose the term “family resemblance” to stand for just that part of classification that does not conform to the Aristotelian category offers an interesting possible challenge to Durkheim and Mauss, since the latter attempt to derive Aristotelian logic precisely from the kinship of tribal structure. Following Wittgenstein’s metaphor, one might argue that Durkheim and Mauss’s proposed source offers a bad model of that for which it is argued to serve as prototype, namely, the Aristotelian category. However, even if at some levels of focus the difference between Aristotelian categories and Wittgensteinian categories or fuzzy sets is critical, it is not so critical from Durkheim’s perspective. In the milieu of social evolutionism, the working assumption was that classification evolved from an original state in which the human mind was able to make no distinctions whatever. Durkheim was attempting to account for the transition from total indistinction to some form of grouping; in this context, the difference between an Aristotelian and Wittgensteinian category is a mere detail. 28. Lévi-Strauss (see 1988: 186–87) in the end does not go beyond, nor admit even a possibility of going beyond, the kinship of the clear Aristotelian category, for the “house societies” are seen by him as ultimately not based on kinship, but merely as invoking a rhetoric of kinship. 29. There are long-recognized problems of circularity in Durkheim and Mauss’s arguments (e.g., see Needham 1972: xii–xv). How is the grouping of humans, which will serve as the prototype of the idea of a category, possible without an idea of category in the first place? However, the Durkheimian/Maussian arguments have more plausibility if we envision an ongoing dialectic, rather than a relation of temporal precedence and antecedence. Humans simultaneously deal with different levels and scopes of classifications—from familial relations to universe—and these levels may exercise various kinds of shaping influence on one another.

Stépanoff: Human-animal “joint commitment” in a reindeer herding system This work is licensed under the Creative Commons | © Charles Stepanoff. Attribution-NonCommercial-NoDerivs 3.0 Unported. ISSN 2049-1115 (Online) Human-animal “joint commitment” in a reindeer herding system Charles Stépanoff, École Pratique des Hautes Études This study explores the grounds and paradoxes of cooperative interaction in a reindeer herding system in Southern Siberia. While the majority of human activities are joint activities where goals or actions of participants require transparency and common knowledge, this article asks to what extent it is possible to build a cooperative interaction with minimal shared knowledge and poor means of communication. The article shows how, despite a lack of a clearly shared plans of action, herders are able to induce reindeer to come back spontaneously to the camps through nonverbal communication, even though reindeer graze freely and autonomously most of the time. Herders come to rely on reindeer’s cognitive skills and desires and, more generally, on animal autonomy in order to keep their herd engaged with them. Paradoxically, humans can domesticate reindeer only if they keep them wild. Yet, in spite of a relation marked by communicational opacity and radical asymmetry, reindeer and men are able to maintain an ongoing cooperative context that allows them to carry out extremely complex joint activities, such as riding. Keywords: reindeer herding, anthropology of cooperation, animal behavior, Tozhu-Tuva The majority of human activities are joint activities: they involve multiple participants who carry out actions together. Participants must understand the common goal of the activity and coordinate their actions. For example, assembling a stand together requires each interactant to adapt his or her own gestures to the gestures and intentions of the other interactants. The choice of tasks and the coordination of movements are accomplished by acts of communication, essentially through spoken language, but also through gestures and embodied displays. Participants commit themselves to the activity by these acts of communication. As Herbert Clark argues (1996, 2006), to be successful, a joint action requires not only the private commitment of each participant, but also a “joint commitment” toward a common goal. In social life, joint activities are not always as symmetrical and transparent as, for instance, shaking hands or assembling a stand together are. Cooperative activities often involve people belonging to different gender categories, age groups (parents and children), or hierarchical positions (bosses and apprentices). These categories are regularly imbued with unequal skills and knowledge, social status, and power. Expert and nonexpert interactions are typically asymmetric: parties have significant gaps in their relevant knowledge. However, on the basis of the study of a divination ritual involving a shaman and a patient, William Hanks (2006: 325) has shown, “sociality need not entail common knowledge, but still gives rise to joint commitment.” In such an interaction, the expert’s challenge is to “induce commitment” of the nonexpert and to bring him to “accept what he cannot know or even understand” (ibid.: 324). To what extent is it possible to build a cooperative interaction with minimal shared knowledge and poor means of communication? Human societies are not composed only of humans; they also include numerous types of animals, such as pets and livestock. In herding societies in particular, people and animals are intimately bound together, they live together and influence each other; they form what could be called “hybrid human-animal communities” (Lestel, Brunois, and Gaunet: 2006). Recent studies have highlighted the ability of some domestic species, such as dogs and goats, to react to social cues presented by humans without prior learning, which seems to be evidence that they were selected for their social-communicative skills during the process of domestication (Hare et al. 2002; Kaminski et al. 2005). Even between humans and insects, one can observe the emergence of presignaletic forms of communication in learning contexts (Renesson, Grimaud, and Césard, this volume). The limits of communicational processes are currently being explored thoroughly, however in the everyday life of human-animal communities, communication is not a goal in itself, it is oriented toward the realization of tasks. In such communities, can some human-animal interactions be identified as forms of “cooperative activities”? On which common ground could a kind of interspecific cooperation be built? Real cooperation must be something both different from biological symbiosis and from communicational exchange. Cooperation implies, at least, a triadic relational scheme involving two agents looking together toward a common goal and involving themselves in its accomplishment. But what can “joint commitment” or “coengagement” of an animal and a human be? In this paper, I will examine how a long-standing social cooperation can arise in a situation of communicational opacity and radical asymmetry. I will explore these questions using the example of a specific Siberian reindeer herding system.1 The paradox of reindeer herding Reindeer herding has recently given rise to much anthropological and ecological research that challenges the comprehension of human-animal relationships. Recent evidence suggests that a system cannot be correctly described from an anthropocentric point of view exclusively concerned with the actions of humans toward or upon animals as the object of such actions. Herders themselves do not interpret their activity in this way: they know that animal behavior, memory, wills, and the social organization of herds are crucial (Paine 1988, 1994). Robert Paine called the capacity of humans to decipher reindeer behavior, and the reactive ability of reindeer to memorize herders’ order of things “reciprocal learning.” Beach and Stammler (2006) have highlighted what they call the “circularity of wills” between humans and animals. It has been reported in the past that Nenets herders “follow the deer,” and indeed this sometimes occurs, but the actual situation is more complicated. As Hugh Beach and Florian Stammler argue, reindeer internalize the patterns of movement dictated by herders, so that “herders follow the reindeer that follow the desires of the humans” (2006: 7). This kind of integration constitutes what they call “symbiotic domestication.” For Kirill Istomin and Mark Dwyer, drawing on Paine’s concept of reciprocal learning, reindeer and humans construct not only their wills, but also their behaviors, in interaction with each other (Dwyer and Istomin 2008; Istomin and Dwyer 2010). In each herding system, humans act according to their expectations and hypotheses relating to reindeer behavior. As for reindeer, they adapt their behavior to herding techniques. Herding systems, Istomin and Dwyer argue, produce a loop of mutual adaptation of human and animal behaviors; however, this loop is also subject to perturbations under the influence of technological and sociopolitical changes. As many authors have noted, a particularity of reindeer husbandry is that reindeer are herded in regions where they can actually survive without the intervention of man. Throughout the Arctic world, many populations of former domestic reindeer are known to have become feral (Baskin 2009). Iceland is populated by thousands of wild reindeer descending from a small group of domestic reindeer imported in the eighteenth century (Thórisson 1984). These facts give a significant advantage to the reindeer. While representatives of native peoples often claim that “without our reindeer, we will perish,” reindeer can manage without humans. Siberian oral traditions recount cases of groups of herders that starved to death because their herds were driven away by huge wild reindeer herds (Gurvich 1977: 49–50). In the present day, it is clear that the survival of some indigenous peoples of Siberia is directly linked with the situation of their reindeer herding.2 Thus, the paradox of reindeer herding is that, compared to other domesticated species, humans can domesticate reindeer only if they keep them (in the) wild. Therefore, the reindeer retain an element of choice: even in the most controlled systems, they can find opportunities to abandon humans and go live without them in the tundra or the taiga. Consequently, herders are forced to be especially aware of reindeer behavior; moreover, they often try to understand and manipulate the motivation behind this behavior, namely reindeer will and desires. Indeed, the existence of a sustainable reindeer herding system is virtually impossible without a kind of cooperation by the reindeer themselves, an engagement by the animals in the human-animal association. Reindeer are thus a good example of the participation of animals in their own domestication (Digard 1990). The herder-reindeer relationship is a fragile and unsettling combination of domination and reciprocity, betrayal and trust, incomprehension and collaboration. Herders must both recognize animal autonomy and keep control of their interaction with them. This balance of herd control is framed by two threats, one internal and one external. If herders excessively concentrate their herd in order to control them more easily, they subject the animals to the risk of epizootics; on the other hand, if the herders’ pressure becomes too loose and reindeer scatter, the risk of predators and the flight of reindeer increases (for more detail on this double threat, see Stépanoff 2012). The key to the herders’ success is their ability to get reindeer to engage in an interactional field and to have the desire to renew this engagement regularly, with as few external constraints as possible. Although reindeer do not seem to share plans and goals with humans, a kind of particular engagement by the reindeer seems to be necessary to support the human-animal relationship, and to achieve such delicate human-animal cooperative activities such as riding or moving to a new campsite through the forest. This cooperation is obviously made difficult by the fact that the participants belong to different species: knowledge is highly asymmetric, communication is poor, and the interests of the participants may sometimes be contradictory. Hence the question: What are the herders’ strategies to get reindeer to engage in a sustainable cooperative interaction? Although the reindeer-human balance is highly dependent on daily contact, there are few ethnographic descriptions of the concrete interactions between men and animals. Recent theoretical studies on reindeer herding have mostly been carried out among Nenets, Komi, and Sámi herders (i.e., on market-oriented pastoral systems of the Arctic tundra, where herders manage herds of thousands of animals). In these systems, direct interaction between humans and most of the reindeer is rather rare and basic. My study deals with a case of reindeer herding among the Tozhu of Southern Siberia. This case of Tozhu herding is part of a hunting-oriented subsistence economy: small herds, ranging from a dozen to 150 reindeer, are herded in the taiga without a goal of meat production. Herders know all their reindeer individually, many of which have a name. I will argue that many of the herders’ everyday actions aim to keep their reindeer engaged within a cooperative framework with them, and that herders attempt to ensure the integration of reindeer in the nomadic community through three central means: (1) cultivating attractiveness for reindeer, (2) using the cognitive skills of reindeer, and (3) favoring hierarchy in the herd. Supporting an ongoing relationship with animals The Tozhu are hunters and reindeer herders of the Sayan Mountains (Southern Siberia) where they neighbor other reindeer herders from historically related populations: the Tofa (or Tofalar), the Soiot of Buriatia, and the Dukha of Mongolia (in Mongolian, Tsaatan). The Tozhu, Tofa, Dukha, and Soiot dialects are very close to Tuvan, the official language of the Tuva Republic in the Russian Federation. Sayan is the southernmost area where there is reindeer herding, and it is also the oldest known, as petroglyphs attest to its existence in the Bronze Age (Kyzlasov 1952). Of course, herding techniques have undergone many changes since then, mostly under the influence of Turco-Mongolian steppe pastoralism. Tozhu herding overcame two tragic crises during the twentieth century: collectivization in 1949 and privatization during the 1990s, which was accompanied by a 90 percent fall in reindeer livestock (see Donahoe 2004). The Tozhu system is typical of taiga reindeer herding. For Tozhu living in the taiga, domestic reindeer are used as beasts of burden, and from spring to fall seasons, does’ milk is a major part of their diet. Castrated male reindeer (chary) are used for daily transport (for riding and as pack animals). Tozhu herders do not use sleds or snowmobiles, which would be impossible in the mountainous taiga where they live. The Tozhu avoid slaughtering their reindeer and they do it only if absolutely necessary, for example when game is scarce. If they need to kill a reindeer, they choose an ill, wounded, old, or rebellious animal. While in tundra herding systems large reindeer herds are regularly watched over—sometimes twenty-four hours a day—Tozhu reindeer graze freely in the vicinity of the campsite, yet frequently also as much as several kilometers away. Tozhu herders sometimes go out to bring their reindeer in, but they never simply stand near to keep watch over them. In winter, reindeer come more or less regularly to the camp of their own accord—some every morning, some less frequently. Everyday, herders check which reindeer come, and if some are missing for too many consecutive days, they try to find them out in the taiga and drive them back to the camp. These searches can be difficult and dangerous and often remain unsuccessful for several days. Finding three reindeer in the dense mountainous taiga is obviously far more difficult than finding a herd of a thousand reindeer in the open landscapes of the tundra. The Tozhu herders must use their hunting skills in order to track and analyze the animals’ footprints. Unsurprisingly, herders prefer it when all the reindeer come back by themselves, but they are also aware that the reindeer might make another choice.3 Some herders report that, after the collapse of the State farm (sovkhoz) system in the 1990s, the population of wild reindeer (taspanan in Tuvan) increased sharply at the expense of lost domestic reindeer from “bad” or “lazy” private herders. In neighboring areas of the Sayan chain, individual herding skills had been so deeply eradicated during the socialist period, that attempts to reintroduce private herding were disastrous. Among the Soiots of Buriatia, two attempts to reintroduce private herding at the end of 1990s ended with the loss or death of most of the reindeer because herders were unaware of the pasture and mobility needs of the reindeer (Baskin 2009: 245–246). Among the Tofa, Donahoe has shown that due to herders’ lack of experience and the lack of attention they paid toward reindeer behavior and needs, domestic herds almost completely disappeared within a few years (Donahoe 2004: 136–137). For example, one of the last Tofa herders kept his riding-deer tied up for several days at a time in the village (ibid.: 137–138), a practice which would never be observed among the Tozhu because, according to them, it causes hunger and stress for the animal. In these conditions, it is clear that the Tozhu herding system’s survival is, to a large extent, dependant on the capacity of herders to develop among their reindeer a tendency to come back to the camps, a willingness to maintain contact with humans. Herding with delicacies In summer, reindeer are harassed by mosquitoes (ymyraa) and botflies (maas), especially when there is no wind. These insects exhaust reindeer and cause diseases, and to avoid them, herders choose windy places for their summer camps. Sometimes herders create smoke in order to drive off insects and the reindeer gather in a compact group near the smoke and stay there all day. At night and on windy days, as soon as the mosquitoes disappear, the reindeer move away to graze freely. However, this technique for keeping the herd close is only effective in the summer. All year round, salt and human urine play a more central role in maintaining the fragile link between humans and animals. In winter, when some reindeer come back to the campsite in the morning, people go out of the house or the tent and distribute salt to them. While giving salt, they tie up some of the mothers and leaders. Other reindeer will stay near those tied up. This practice is not novel. At the beginning of the twentieth century, Douglas Carruthers (1914) observed among Tozhu reindeer that “for salt they have a very keen desire, keener than I have noticed in other animals; here it is a common sight at evening to see the women feeding their pet deer with salt out of little leather bags as they come to the tents for the night” (1914: 231). Similarly, Ørjan Olsen ([1915] 1921: 94) noticed that Tozhu reindeer like salt very much, and added this interesting hypothesis: “this is perhaps the reason why they stay so close to humans.” Nowadays, some officials in the agricultural administration in the Tozhu region explain that salt is necessary for the physiology of reindeer. However, ethologists consider salt to be rather dangerous for their health if it is distributed in winter, as the Tozhu do, because the animals can get thirsty and then swallow too much snow (Baskin 2009: 17). Herders do not claim that they distribute salt for the sake of the animals’ health: they know that wild reindeer find salt by themselves in natural fields and in plants, and infer that domestic reindeer can also do this. Herders explain that they actually give salt because otherwise their reindeer would not come back to the campsite of their own accord. Indeed, the presence or absence of salt in the settlement has a strong influence on reindeer behavior. I observed that in the camp of a poor herder who had run out of salt, his reindeer did not come back for several days running, while some reindeer came back to the settlements of other herders who gave out salt every morning. For herders, running out of salt can lead to the herd returning to the wild. It is worth noting that the presence of salt also has important connotations in Tozhu cooking. The liver and fat of game animals are eaten raw and not salted. It is forbidden to offer salted food to the master-spirits of the landscape through the fire, probably because salt, as a characteristic feature of the human diet, is not suitable for nonhumans. In cooking, as in herding, salt seems to underline the boundary between wild and domestic domains. The attractive power of human urine In the Tozhu system, urine is a crucial component of the familiarization of reindeer with the human body and its odor. Although reindeer fear the smell of man, they are attracted by the presence of salt in human urine (Baskin 2009: 76, 80). Tozhu men are used to urinating near the house, often on a hollow stump, or even in a urinal specially constructed for reindeer: a tree trunk with a trough carved in it, adapted to height of reindeer mouths (figure 1). These urinals are constructed very close to the home, about ten-fifteen meters away, although among Western (steppe) Tuvans, it would be not decent to urinate so openly. In winter, urine freezes immediately and is thus conserved in these urinals, so that when they come to the campsite, reindeer can easily lick it. Reindeer also lick frozen urine and dishwater on the ground, breaking the ice with their hooves. Figure 1: Reindeer rush up when a man comes to the urinal. As soon as they arrive at the campsite, reindeer observe people’s behavior closely—they wait for salt or urine. When a man seems to be about to urinate, some reindeer immediately gather around him, sometimes running up to him. In 1914, Olsen had already observed this behavior among the Tozhu. He points out, “When they see a man urinating, reindeer come from everywhere to be sure to find their diet. By virtue of [their] intimate relationship with men, they are so gentle that you can often catch them with your hands” ([1915] 1921: 95). Brian Donahoe (2004: 133) even noticed that herders “discriminately distribute their urine (an important source of salt for the deer) to their favorites or to those they feel need it most.”4 Obviously, urine strongly consolidates intimate relationships and bodily attachment between reindeer and humans. Familiarizing fawns and creating addiction The relationship between a mother doe and her fawn is established during the first three days of the fawn’s life. Initially the fawn does not specifically recognize its mother: it is attracted by any large and dark form. If this relationship is disturbed, the mother may refuse to feed her fawn (Baskin 2009). Tozhu herders seem to have an intimate knowledge of this development process. They avoid contact with the fawns during the first days of their life. After that, the herders drive the does and their offspring to the settlement, where the fawns are tied up for several days near the tent. Their mothers are released to graze during the day, but they do not go far as their fawns stay at the campsite. This is both a means of familiarizing the fawns and of controlling the herd. In the evening, the does are brought in and tied up near their fawns to suckle them and to be milked by the herders. While they are tied up without their mothers, young fawns receive salt, or even sugar if there is some, from the hands of the herders. Herders do not distribute sugar and salt to fawns for the sake of their health, but, as they say, because “everybody likes sugar!” They recognize that the operation of tying up fawns is quite tiring, however they also know that this critical time has a strong influence on the further development of the reindeer. Indeed, while fawns lick the herders’ hands, they learn to associate pleasure with the scent of man and the contact of human skin. Herders endeavor to ingrain in reindeer from birth a physiological need for salt and to transform it into a kind of addiction, so that reindeer will look for more salt than they need and will be willing to come back to their settlements regularly in quest of pleasure. It should be noted that people do not distribute salt just because they are preoccupied with the necessity of maintaining control on their herd. Visitors to a camp like to give salt to local reindeer too, even though they are not theirs. From the observation of herders’ behavior it seemed clear to me that from the human side, there is also a kind of pleasure in distributing salt: perhaps the satisfaction of seeing a reindeer give in out of greed, slowly overcoming its fear and surrendering to humans, the pleasure of a special, fragile, and daily reinforced human-animal contact and communication (figure 2). Figure 2: Attracting a deer with salt (1), offering salt (2), and tying up (3). How do you get a reindeer to carry you? In large-scale herding systems, many of the techniques of maneuvering reindeer herds are based on reindeer’s fear of humans. Reindeer are often frightened in order to be “pushed” in the right direction (Baskin 2009: 110–121). Riding and pack animals, which are the most used to humans, are released into the herd after work. Once in the herd, they adopt the behavior of the other reindeer and they tend to flee from humans, so that it is often difficult to catch them again. According to Baskin, a minimum of two herders is needed to catch a deer: one disturbs the group where the deer is located and forces them to run, while the other herder is hidden and tries to catch it with a lasso (ibid.: 122–125). Unlike some tundra herders, the Tozhu have no herding dogs; they have lassos, but they are rarely needed, apart from using them for reindeer with a particular problem, for instance if they are frightened or wounded. Usually a single herder can catch a riding reindeer out at pasture without any trouble. He follows the tracks and when the reindeer are in sight, he calls them, shouting “oh! oh!” and shows a little bag of salt that he shakes in his hand. The cries “oh! oh!” are neither human nor reindeer language. They are an element of the different vocalizations and words the Tozhu use to address their reindeer, something similar to what Eduardo Kohn (2007) calls a “transspecies pidgin,” a specific language adopted by humans in order to communicate with a domestic species (cf. Fijn 2011). Of course, the reindeer will have detected the herder’s approach much earlier through the noise he makes and his smell, so that the aim of the cry is not to let them know that he is there, but to attract their attention and interact with them. The reindeer then stand and look at him—a field of reciprocal attention has been opened. After some hesitation, one reindeer decides to approach, and then all the reindeer begin to gather around the herder. Then he gives salt to the animal he needs, and puts a rope around it without any problem. The field of reciprocal attention has evolved into a field of cooperative interaction. Basically, in Tozhu herding, catching a reindeer and getting it to work initially require that the reindeer agree to engage in an interactional field with humans. Cries open a communicative field in which the reindeer is invited to engage itself. Once the deer is engaged in the interaction, the herder demands more and more engagement and coordination from it: he leads it to the camp, saddles it, and finally, standing on a small support, he deftly mounts. The support, usually a tree stump, permits the rider to mount without harming the animal’s back. In human interactions, people first negotiate to commit together to a common activity toward an overall goal, before committing to the segmental actions required for the achievement of the activity according to a means-ends analysis. For example, the joint activity of “assembling a stand together” can be broken down into actions such as “arranging the parts,” “assembling the parts,” and so forth. As Clark argues, commitments are hierarchical and stack vertically from a basic goal up to particular tasks (Bangerter and Clark 2003; Clark 2006). In reindeer-human interactions, joint activity is not structured in such a vertical hierarchy. On the contrary, it seems to progress horizontally from simple contact to more complex interactions. The reindeer first engages in a simple exchange of salt through a tongue-hand contact, after which it is induced into a more goal-oriented cooperative activity, such as going together with a man on its back through the taiga. Riding on mountainous terrain covered by deep snow is a relatively difficult activity, which requires good coordination between human and animal. It is virtually impossible to achieve it with horses, which are not adapted to this region in winter. A bad coordination between the rider and the mount leads to a loss of balance, and the rider will fall off or the deer will be injured. Breaking a reindeer’s back is an accident herders fear. Uneven ground, tree branches hampering the rider, unexpected slippery ice, cause ceaseless shifting of the load on the fragile spine of the animal. The deer adapts its position and effort in order to maintain its balance as well as its rider’s. The rider helps it by modifying his seat and his posture, and by leaning on the stick he holds in his right hand. This stick is an important element in the technique of reindeer riding, which distinguishes it from horse riding, and reveals the rider’s active participation in the maintenance of balance (figure 3). Lightly tapping the reindeer’s rump with the stick and a special “ah ah” cry signal to the deer that it must move forward. Figure 3: Riding a reindeer. The rider ensures his balance and helps the reindeer’s efforts by using his stick. In his left hand, he holds the rein attached to a rope around the reindeer’s neck. Maintaining an attachment When reindeer arrive at the campsite in the morning, herders who have time (and energy) tie up the leaders and some other does; they then release them in the evening. The technique of catching deer requires, first of all, distributing salt, or making the gesture of offering salt, then deer let themselves be caught by the neck and tied up (figure 4 and figure 5).5 The rope is then tied to a post or to the trunk of a young larch (shet) lying on the ground. Figure 4: Tying up a reindeer standing near the urinal. Figure 5: The boy catches the deer by the neck with the right hand, with which he was just giving salt. In the left hand he holds a salt-bag. At first glance, tying up reindeer could seem to be a strange and risky practice. Reindeer could dislike being caught and forced to stay all day in a place where they have nothing to eat or drink (in the campsite, for example, lichen and grass are quickly crushed and covered by excrement). They could get hungry or contaminated by infectious diseases like foot rot. So why do the Tozhu regularly tie up their reindeer? It is not in order to protect them against predators, since wolves prefer to attack at night. Some herders claim that if reindeer stay in camp all day and are released at night, they will not go graze too far away. However this is not the main reason. All herders agree that if reindeer are not tied up, they tend to become less “smooth” (chaash), less familiar with humans, and they might disappear into the forest: “You teach by tying up” (baglap doredi’r). Thus it appears that keeping reindeer tied up all day at the campsite has no other goal than to support the relationship of intimacy between reindeer and their herders. It is a kind of reflexive interaction intended to allow future interactions to happen. By attaching their reindeer, herders force them to prefer the contact of salt to free pasturing, to dominate their hunger, to keep them accustomed to human smells, voice, contact, and to remain “attached,” in a figurative sense, to humans. 6 It supports a cooperative context, inside which reindeer can be induced to having more active participation in specific activities, such as riding and milking. Individualities among the herd A wild reindeer herd is not an undifferentiated mass. It has an internal social structure, albeit a quickly evolving one. Among wild reindeer in the tundra, the position of the leader is quite fragile and temporary. In a situation of collective hesitation the first reindeer that takes the decision to cross a river or to flee becomes the leader. In one herd, 13–37 percent of reindeer are potential leaders, each of which can become an actual leader in specific circumstances (Baskin 1976; Baskin 2009: 102–104). How can reindeer be integrated into a human-animal community if they have their own social structure? Everywhere in Northern Eurasia, herders are aware of the internal organization of herds and use this to manage reindeer movement (Baskin 2009; Paine 1988; Takakura 2010). I will argue that Tozhu herders not only observe and use hierarchy, they also produce it. Bucks and leaders The Tozhu castrate most males and keep only one buck for a maximum of thirty females. A big and strong reindeer is selected to remain entire. Herders rely on bucks’ strength to defend the herd against wolves. This brings to mind an observation made by Olsen among the Tozhu in 1914: “even when wolves are numerous, little is done to help reindeer, which manage on their own” (Olsen [1915] 1921: 101). Two-year-old castrated males, chary, are trained to become pack animals: they learn to accept quietly being tied up, loaded, and ridden. They are the most “tame” individuals in the herd. Chary have “nicknames” (chola), such as kara chary “black ox” or ak chary “white ox.” Among wild reindeer not all males have equal access to mating, as dominant males have priority. The position of the dominant male is the result of fighting and intimidation; it is neither irreversible (i.e., it can change every year) nor exclusive. In the middle of the mating season, dominant males may become exhausted so that mating opportunities for the low-ranking males increase (Hirotani 1994). Castration creates an irreversible and exclusive supremacy for the entire buck. It amounts to rigidifying a situation of domination that is fluid in the wild. The “herd effect” or “gregarious behavior” is the process through which the decision of one experienced reindeer becomes the decision of a whole group of reindeer. It is crucial for herders to control which reindeer’s behavior will be followed by the herd. In large-scale herding, herders endeavor to frighten the most timid reindeer, generally adult does, in order to put them in the position of leaders. Young deer do not fear humans enough and they do not flee straight from them. Herders frighten those reindeer that will have a clear and predictable movement of flight and will be followed by the rest of the herd in the direction expected by herders (Baskin 2009). In contrast to large-scale methods and to the organization of wild reindeer herds, Tozhu herders select leaders among the less timid and most familiar animals, and they actively strengthen their position. Leaders are elderly does, whose main function is to lead their offspring and kin to human settlements regularly (figure 6). They have nicknames, such as koigunak “rabbit” or ulug myndy “great doe.” In Tuvan, they are called bashtanchy (or “leaders”), from the verb bashta- (“to lead, to drive”). Figure 6: A bashtanchy (leader doe). Oleg is a rich herder who possesses around 150 reindeer (Donahoe interviewed him in 2000, see Donahoe 2004:174). He has three leaders, each followed by a group of around fifty animals. Even when the whole herd is gathered, these groups remain separate entities. Leaders drive their groups to graze in different directions. Some lead them to the campsite every morning, some disappear for a long time—it depends on the leader’s choice, namely its liking of salt. In the evening, when the reindeer are released, they are driven by herders to a distance of 50–100 meters from the campsite, after which the leader drives the herd on its own to a suitable pasture. Leaders also open up the way in the snow, making it easier for younger deer. Leaders are easily recognizable by the bell hanging from their necks. Bells are the key device used by herders to stabilize their status. Each bell has its own sound, so that reindeer can easily recognize and follow their leaders. Some herders report that they accustom fawns from birth to hearing the sound of their leader’s bell and to pay attention to it. And as reindeer are generally grouped around their leader, bells also help herders to find lost reindeer. Bells also contribute to the protection of reindeer, as they are supposed to frighten wolves. Thus, being around the leader is safer for other reindeer, which certainly contributes to strengthening its authority and herd cohesion. Herders emphasize that they do not choose or train their leaders to be leaders. The leader, as they claim, “appears by itself, it is such from birth” (typtyp keer, bodu-la törümelinden). A herder explained: “They cannot be trained, they have it in their blood [hanyndan]. It is the same among humans; some are smarter than others.” In their understanding, herders do not create but recognize the leaders. Practices, however, are fairly interventionist. Although herders make use of some dispositions to leadership, they also actively modify them. First, leaders are sorted by the herders; “bad” leaders, those that have a harmful influence on other reindeer, leading them to flee from humans, are slaughtered. Second, herders strongly influence the relational structure of the herd, strengthening and stabilizing the dominant position of the leader. While the leader is a temporary position in the fluid structure of wild herds, in Tozhu herding, being a leader is a long-standing status integrated in a stable hierarchical organization—in a way quite similar to human social organization. Consecrated individuals It is important to emphasize that the human strategy of selecting bucks and leaders and stabilizing their hierarchical position is associated with a theory about the innate qualities of individuals. I have shown elsewhere that the innatest and essentialist mode of comprehension is also active in the strategy of identification of other “singular” (onzagai) beings, such as sacred reindeer and shamans (Stépanoff 2011). Herders are very sensitive to the individual idiosyncrasies of their reindeer. They assume that “each reindeer has its own different character” (azhy-chaŋ). The Tozhu have an institution that highlights their attention to individuality in the herd: the consecration of particular reindeer called ydyk to the master-spirits of the landscape (oran eezi). Most households have one ydyk reindeer in their herd. Such an ydyk is described by herders as an animal “that is never touched” (shuut degbes)—it is not ridden, not loaded, not milked if it is a female, not killed, and not eaten. However, it is not considered as idle. On the contrary, as a herder said, “the ydyk reindeer watches over [harap-körüp turar] and protects the household. If there is an ydyk animal, it is said that people and reindeer will not fall ill.” It is interesting to observe that in this herding system, where herders barely watch over their herd, they assign the task of protecting humans and reindeer to their ydyk reindeer. Reindeer to be consecrated are “animals that somehow distinguish themselves”—they have a particular coat, white or black, green eyes, or a “particular character” (aazhy-chaŋy aŋy). The inferential process determining the selection can be described as such: visible atypical features are interpreted as indices of an uncommon personality, an individual essence. From this essence, this reindeer is expected to be graced with other aspects, such as occult powers and relationships with invisible agencies. The ritual of consecration is performed in a singular human-animal relational configuration. The ritual is carried out by a shaman or an elder of the household. The reindeer is fumigated with juniper and colored ribbons are tied around its neck. Milk is poured over the animal. Then it is led three times around a fire. Next, the person performing the ritual catches hold of its head and forces it to bow three times before the sun and to “pray.” During this the ritual performer pronounces: “May the herd grow, may there be a lot of milk, may alien things (öske chüveler) not attack.” These words are attributed to the reindeer; it prays for the happiness of its herd and against wolf attacks (designated by the euphemism “alien things”), just like a shaman does for his or her patients. In this ritual, the reindeer is given the position of an actor and is responsible for the relationships between the herd and the entities of the environment (sun, wolves, etc.). The practice of consecrating reindeer is important for the problem of cooperation between man and reindeer. From a psychological point of view, this institution actively contributes to spreading and reinforcing the following: (1) the idea that the herd is structured by deep essential differences that exist between individuals and that these differences should be noticed and enhanced by humans (In other words, humans’ psychological essentialism contributes to reinforcing a social organization that is much more volatile among wild herds.); (2) a model of a human-animal relationship which attributes autonomy and responsibility to reindeer. A cooperation through nomadic space Reindeer mostly pasture freely over vast territories around human encampments that are moved seasonally. During the year, there are many possibilities for humans and reindeer to lose each other. It is important that humans and reindeer have common perceptions of the landscape and common routes to meet up again and maintain their association. According to herders, reindeer remember landscapes and roads very well, even better than humans do. If a reindeer taken to a village in summer is then let free, it will be able to find its own way back to the campsite, even if it is many kilometers away. However, in winter, reindeer do not suffer from the heat in villages as they do in summer, so they may decide to stay. If a rider gets lost, or if he is drunk, he can simply let his reindeer go free and it will lead him back to his campsite. Nomadic routes belong to common geographical knowledge shared by reindeer and humans. The nomadic route is neither a product of human or animal volition, it is the result of long-standing relationships among humans, animals, and their common environment. A nomadic route is the spatial projection of the reindeer-human coengagement in the landscape (see figure 7). Figure 7: Routes in the Sayan Mountains. Satellite view: © Google 2012, Terrametrics 2012. The spatial pattern of routes It is possible to distinguish ecological factors (animal behavior, biotopes, climatic conditions) and nonecological factors (political environment, cultural traditions) that influence the movement pattern (Dwyer and Istomin 2008), however many of these factors are closely interconnected. Anatoly Khazanov (1994: 38) stressed that, within one migration route, it is important to distinguish the interseasonal movements, which are determined by natural seasonal cycles and the intraseasonal movements that depend on the herd’s size, and a series of other nonecological factors. Spatial patterns of routes are a combined response to different parameters such as: Traditional land use rights. Reindeer health and reproduction needs. Herd control needs. Other occupations, such as hunting, fishing, or going to the village of Adyr-Kezhig. Land rights. In the Tozhu area, the overall geographic setup of the annual route of a herder family is primarily determined by the land use rights and by reindeer needs. Of course, these factors are themselves connected, as hereditary territories are adapted to reindeer behavior. Although present-day herders may have been employed in different places during the Soviet era, after the fall of the sovkhoz, they came back to places where their families had roamed before. Oleg says that he lives on his “ancestors’ territory” (ögbeler churtu). This territory is structured by “roads” (oruk) that connect different river valleys. The herd’s needs. The needs of the herd are generally fairly similar to those of wild reindeer, thus their respective routes are often quite close. In winter one can observe wild reindeer tracks just a few kilometers away from those left by domestic reindeer (herders are able to distinguish wild and domestic reindeer tracks). In winter, reindeer need places with rich lichen cover, and not too much snow or wind. In the Tozhu region, such places are found in low valleys with taiga vegetation. Summer pastures must be places with cool temperatures and a strong wind in order to avoid insects and diseases. In Tozhu, high open pasture with mengi “everlasting snow,” without trees, where the wind blows day and night, is ideal. Consequently, Tozhu migration patterns come under “vertical nomadism,” going from low taiga pastures in winter to the summits in summer (figure 8). Control needs. These depend on the herd’s size. A large herd has several leaders which go off to graze in different directions. Therefore controlling its movement is more difficult than with a small herd. In a valley, when herders release reindeer in the evening, they drive them in the direction of an uphill slope, or toward rocks, or up to high snow, so that it will be less tiring—and therefore more tempting—for them to come back to the campsite the following morning than to move away from it.7 Other occupations. Hunting and fishing do not have an equal influence on all the households and their nomadic patterns. Comparisons show that poor herders with small herds are freer to change their itinerary in order to go hunting or fishing. Figure 8: Distances and altitudes of a migration route. Hot months are spent in the high altitude tundra. Routes and reindeer memory Herders with small herds often take two different routes to climb up to summer pastures and to come back down in autumn. For example, one poor herder sometimes changes his summer pasture and goes to another valley in order to find more game of different species. He says that he finds it interesting to explore new hunting and fishing territories. But, such freedom in the choice of his itinerary is not possible for a rich herder like Oleg, who owns around 150 reindeer. It is important for him that his reindeer remember a single route; this is why he takes the same route in the same valley every year to reach summer and winter pastures. Oleg is one of the most experienced and respected herders of the high Yenisei Valley. Donahoe published data on his route in the year 2000 (Donahoe 2004: 174), which enables us to assess the evolution of his movement pattern over ten years. Formerly living only in a tent, in March 2000 Oleg was building a log cabin on his winter pasture near a lake in the Serlig-Khem region (ibid.). In February 2011, when I visited him, he already had five such cabins; each winter he stays two to three months in one of them depending on the weather conditions that year. The rest of the year, Oleg lives in a tent. Interestingly, in his case, the construction of log cabins does not imply a reduction of mobility. The cabins are situated at various altitudes between about 1100 meters and 1400 meters, within a ten kilometer radius. Knowing exactly the different average snow coverage at each one, Oleg chooses the most convenient one each year. In addition to this climatic factor, he does not go back to the same cabin two years running, in order to enable the lichen to regenerate. Oleg selects a winter pasture that insures an ideal thickness of snow, around 40–60 centimeters. Snow should not be neither too thick, so that reindeer can dig for lichen with their hooves, nor too thin, which would enable them to move too easily and for wolves to approach and attack reindeer. In 2010, the total route length was 125 kilometers, with a difference in height of about 1000 meters between the winter and summer campsites. Oleg avoids wide rivers such as the Yenisei, where reindeer can move easily on flat ice with thin snow cover and that is frequented by wolves. At the end of the 1990s, Oleg had only about twenty reindeer, which he had received from the former sovkhoz after privatization. At that time, he used to follow different seasonal routes: he went up to summer pastures through the Terben Valley and back down in fall following the Cholos and the Serlig-Hem. However, when his herd grew, he decided to use the same corridor both in spring and in fall, the Terben Valley. He explains: Having only one route is better for the reindeer: they remember it…. That is my habit, and my reindeers’ habit: they know the route even better than I do. In March, I send them to the summer pasture and they will go there anyway, I don’t need to look for them. When we move, if a couple of reindeer stay behind it doesn’t matter, they will get to the campsite later. If I wandered away from this route, my reindeer would get lost…. A reindeer is clever like a person. That is my secret. Oleg’s strategy is to adopt a single route, so that reindeer can easily internalize it. The reindeer will go in the right direction without requiring much effort from the herder to get them to do it. This kind of movement management relies on animal memory and on the internal organization of the herd, because the route is memorized by experienced leaders who drive the herd to the right place. As Oleg and his family do not have the necessary strength, technologies, or the desire to control their large herd actively, they have to rely on reindeer capacities and their willingness to collaborate in the human-animal nomadic movement. The consequence is that the new route must be adapted not only to the reindeers’ physiological needs, as it were, but also to their cognitive skills. On the other hand, by using this single route, Oleg can no longer allow himself to go off in search of new hunting and fishing places. The growth of his herd has lead him to reduce the scale of nonherding elements connected to the migration, thus decreasing the share of hunting and fishing in the household income. The household becomes more dependent on its herd; in order to get cash or meat, it will be easier for them to sell or to kill reindeer. However Oleg continues to hunt whenever he can. (It is easier in winter when the herd’s requirements are low.) In purely pastoral contexts, it is well known that large herds require more pasture and more mobility than small ones. This is not only caused by the dietary needs of the livestock, but also by the behavior of the animals (e.g., cows walk faster in large herds than in small ones). However, this is not true in the case of the Tozhu because, besides reindeer herding, other factors such as hunting and fishing induce herders to move. Small herds of ten reindeer can walk very far if their herder wants to hunt. The size of the herd does not influence the length of the route, but its form—the larger the herd, the more stable and simple the migration route should be in order to maintain herd control.8 This case is particularly interesting as it gives a clear example of a smooth transition from a mixed system where hunting is dominant to a herding-oriented system. In this transition, a key factor is played by the requirements of reindeer cognition. It is possible that in other places throughout Eurasia, the growth of herds has increased the importance of animal “memory requirement” in herd management and in movement patterns, which has rendered impraticable an erratic mobility dedicated to hunting. The timing of the itinerary We have seen the importance of different factors influencing the choice of the spatial aspect of the itinerary, but a nomadic route is not only spatial. It also has a calendar, a series of decisions of moving, which varies from year to year. It has been observed among Nenets herds, for example, that reindeer tend to migrate when they feel the need. As Dwyer and Istomin (2008: 529) put it: “The herders move when reindeer no longer want to stay on the pasture.” It means that in the decision process of moving to a new camp, pasture and climate conditions are not directly evaluated by the herders but by the reindeer themselves (ibid.). This analysis is clearly relevant to Tozhu herding. The Tozhu are proud to be more mobile than Western Tuvan pastoralists, who move camp only two or four times a year. The Tozhu explain that they actually cannot stop being mobile, because “the reindeer move themselves; if you stay, your reindeer will go away.” Tozhu herders observe that, at the end of April, as soon as it becomes “hot,” reindeer cannot stay in the same place, they begin to head off toward the calving area and the herders must follow them. At the beginning of summer, reindeer tend to go higher up in the mountains. In fall, they have a very strong tendency to leave the high mountain pastures and to descend into the valleys (in fall many mushrooms appear in the forest and reindeer, who love them, go “mushroom picking”). Another reason is the snowfall. Reindeer go down to the forests, where the snow is thinner than on high pastures and will not prevent them from finding food. Therefore, reindeer have a tendency to follow the route expected by humans, although the timing of their migration cannot be controlled by humans, as it depends on complex ecological factors. As in Dwyer and Istomin’s study (2008), herders do not make their decision to move after an evaluation of different ecological factors. They observe the behavior of their reindeer and they move when they notice that “cattle cannot stay in place” (mal turbas, and when it becomes too difficult to control the herd. In short, (1) the choice of migration routes and fixed campsites is made by herders, who evaluate the quality of pastures and the presence of snow; (2) the timing of movements depends mostly on reindeer behavior. Herders consider that the decision to leave a camp is actually taken by the reindeer. While humans master the spatial aspects, reindeer master the timing. This does not mean that reindeer have no idea regarding the question of the direction of the migration, or that humans have no idea about the timing of the migration. However, humans tend to impose their view about the direction, and to listen to reindeer about the time to go. This is understandable as it is much easier to orient the movement of a herd than to prevent it leaving when it has already decided to migrate. Discussion: Animal autonomy, distributed cognition, and joint commitment This study confirms Istomin and Dwyer’s (2010) hypothesis that reindeer behavior should not be considered as a natural given, as it largely depends on local herding technologies, and as such has cultural foundations. Tozhu reindeer behavior is as far from tundra reindeer behavior as their herders’ techniques are from those of tundra herders. The human-reindeer relationship is characterized by fundamental opacity and asymmetry. Humans ride and kill reindeer; the opposite is not true. Humans can claim that they have a much better knowledge of reindeer intentions than vice versa, and they are certainly right. Herders make sure that reindeer understand their intentions when they ride them or drive them in the taiga. However, a reindeer does not know that the herders want to slaughter it, until the split second when it feels the knife in its neck. While among western Tuvans slaughtering a sheep is a public and festive event carried out near the home, among the Tozhu, men usually kill the deer in isolation, out of the sight of the other reindeer, but also of women and children. As a result, while in large-scale herds reindeer develop panic behavior related to the experience of slaughtering places, Tohzu reindeer do not show fear connected with the danger of being killed by their herders. On the other hand, it does not seem possible for a reindeer to hide its intention to kick a human or to flee from him. Thus humans can relatively easily make inferences about reindeer intentions. Herders attribute intentions and desires to their animals, they try to anticipate and modify them, and take them into account when taking decisions. Humans are able to select which of their intentions they communicate to reindeer, which means that they have representations of the representations reindeer have of their intentions. There is no evidence of the opposite being true for reindeer. Although asymmetric and opaque, this relationship is not without reciprocity. Discussing Tim Ingold’s (2000) concepts, Donahoe (2004) rightly observes that “respect of autonomy” and “trust” are crucial in the Tozhu attitude toward domestic reindeer. This is true at different levels of herd control: protection, direction, cohesion, in which herders try to lead reindeer to engage themselves. Most of the time, the Tozhu do not watch over their herd, but rely on the reindeers’ capacity to gather around their leader and to go in the right direction, and on their desire to keep contact with humans in order to have access to salt and for protection. This kind of joint management, which is central to the Tozhu herding system, is founded on animal autonomy. The principle of animal autonomy implies that humans expect and encourage in animals the presence of skills that enable them to play an active role in their relationship with humans and with the environment. Humans must deeply engage in their association with reindeer. They must completely adapt their way of life to reindeer needs. In order to be able to use reindeers’ strength, milk, and, sometimes, flesh, humans give them their urine. Reindeer can lead herders into potentially critical situations if they leave the herders, and opportunities of this happening arise every day. This gives strategic importance to the animals’ will and desires within this herding system; herders cannot ignore this. Furthermore, herders’ bodily engagement (giving reindeer salt and urine, milking them, tying them up repeatedly) is necessary in order to build up an intimate relationship with their reindeer. Only in such a relationship will reindeer then engage themselves. In the Tozhu system, herders share some important cognitive tasks with reindeer such as the evaluation of complex ecological factors, the memorization of the route, and decision-making about migration. So, is it possible to qualify this herding system as a “joint activity” with shared intentionalities and goals? In this relationship, human and animal knowledge is highly asymmetrical, although complementary. The herder-reindeer relationship is close to what Hanks (2006: 302) calls “induced commitment,” by which the nonexpert, although he or she does not understand what the expert is doing, “is drawn into the process as an active co-participant.” Certainly, although they do cooperate, there is no evidence that reindeer share common plans with humans and that they feel a motivation to help humans to reach them. For example, a reindeer and its rider can both have the intention: “We come back to the campsite,” and the reindeer will actively and deliberately contribute to achieve it, even if the rider is drunk and unable to drive it. If the drunken rider happens to fall, the reindeer will stand by and wait for him so that he can easily mount. This seems to indicate that the reindeer’s intention does include the rider and that it is aware of its participation in a joint action. However, in this case we may have two coordinated but separate intentions, rather than a joint plan of action in which two intentionalities merge. Only humans seem to be able to form representations on others’ representations. According to Michael Tomasello (2006), only humans engage in acts of shared intentionality, where participants integrate others’ intentions and representations and coordinate their actions to pursue a shared goal. So in what kind of opaque and shaky cooperation are reindeer and herders engaged? Our data show that, in the Tozhu system, humans cannot accomplish their goals if their reindeer do not support them and regularly reinforce the relational field constituting the frame of these goals. Reindeer “induced commitment” is subtly aroused and controlled by herders; it is based on animal memory, desires, and motivation to migrate, which are modified to include humans. What is shared in such a human-animal community is not so much a strict and explicit plan of actions, but rather a common intention to maintain a cooperative context—a desire to carry on living and carrying out undefined actions together. Acknowledgments I am grateful to the editors, the authors and the anonymous reviewers of this issue for their precious comments. I am also indebted to Brian Donahoe for the texts and information he communicated to me. References Bangerter, Adrian, and Herbert H. Clark. 2003. “Navigating joint projects with dialogue.” Cognitive Science 27 (2): 195–225. Baskin, Leonid Mironovich. 1976. Oleniprotiv volkov. Moscow: Znanie. ———. 2009. Severnyi olen’: Upravleniepovedeniem ipopuliaciiami: Olenevodstvo, ohota. Moscow: Tovarishchestvo nauchnyh izdanii. Beach, Hugh, and Florian Stammler. 2006. “Human-Animal Relations in Patoralism.” Nomadic peoples 10 (2): 6–29. Carruthers, Douglas. 1914. Unknown Mongolia. A record of travel and exploration in North-West Mongolia and Dzungaria. T. 1. London: Hutchinson & Co. Clark, Herbert H. 1996. Using language. Cambridge: Cambridge University Press. ———. 2006. “Social actions, social commitments.” In Roots of human sociality: Culture, cognition and interaction, edited by Nicholas J. Enfield and Stephen C. Levinson, 126–150. Oxford: Berg. Digard, Jean-Pierre. 1990. L’homme et les animaux domestiques: Anthropologie d’unepassion. Paris: Fayard. Donahoe, Brian. 2002. “Hey, you! Get offa my taiga! Comparing the sense of property rights among the Tofa and Tozhu-Tyva.” Max Planck Institute for Social Anthropology Working Paper Series 38. ———. 2004. “A line in the Sayans: History and divergent perceptions of property among the Tozhu and Tofa of South Siberia.” PhD Thesis. Bloomington, IN: Indiana University. Dwyer, Mark J., and Kirill V. Istomin. 2008. “Theories of nomadic movement: A new theoretical approach for understanding the movement decisions of Nenets and Komi reindeer herders.” Human Ecology 36 (4): 521–33. ———. 2009. “Komi reindeer herding: The effects of socialist and post-socialist change on mobility and land use.” Polar Research 28 (2): 282–97. Fijn, Natasha. 2011. Living with herds: Human-animal coexistence in Mongolia. Cambridge: Cambridge University Press. Gurvich, Iliia S. 1977. Kultura severnyh iakutov-olenevodov. Moscow: Nauka. Hanks, William F. 2006. “Joint commitment an common ground in a ritual event.” In Roots of human sociality: Culture, cognition and interaction, 299–328. Oxford: Berg. Hare, Brian, Michelle Brown, Christina Williamson, and Michael Tomasello. 2002. “The domestication of social cognition in dogs.” Science 298: 1634–36. Hirotani, Akira. 1994. “Dominance rank, copulatory-behaviour and estimated reproductive success in male reindeer.” Animal Behaviour 48 (4): 929–36. Ingold, Tim. 2000. The perception of the environment: Essays in livelihood, dwelling and skill. London: Routledge. Istomin, Kirill V., and Mark J. Dwyer. 2010. “Dynamic mutual adaptation: Human-animal interaction in reindeer herding pastoralism.” Human Ecology 38 (5): 613–23. Kaminski, Juliane, Julia Riedel, Josep Call, and Michael Tomasello. 2005. “Domestic goats, Capra hircus, follow gaze direction and use social cues in an object choice task.” Animal Behaviour 69 (1): 11–18. Khazanov, Anatoly M. 1994. Nomads and the outside world. 2nd ed. Madison: University of Wisconsin Press. King, Alexander D. 2002. “Without deer there is no culture, nothing.” Anthropology and Humanism 27 (2): 133–64. Kohn, Eduardo. 2007. “How dogs dream: Amazonian natures and the politics of transspecies engagement.” American Ethnologist 34 (1): 3–24. Kyzlasov, Leonid R. 1952. “Drevneishee svidetel’stvo ob olenevodstve.” Sovetskaia ètnografiia 2: 39–49. Lestel, Dominique, Florence Brunois, and Florence Gaunet. 2006. “Etho-ethnology and ethno-ethology.” Social Science Information Sur Les Sciences Sociales 45 (2): 155–77. Olsen, Ørjan Mikael. (1915) 1921. Los Soyotos: Un pueblo primitivo, nomades Mongoles pastores de renos. Translated by Erling Marius Nilsen. Madrid: Calpe. Paine, Robert. 1988. “Reindeer and caribou Rangifer tarandus in the wild and under pastoralism.” Polar record 24 (148): 31–42. ———. 1994. Herds of the tundra: A portrait of Saami reindeer pastoralism. Smithsonian series in ethnographic inquiry. Washington: Smithsonian Institution Press. Petri, Berngard È. 1927. Olenevodstvo u karagas. Irkustk: Izd. Vlast’ truda. Rastsvetaev, Mihail K. 1933. Tungusy miamial’skogo roda. Social’no-èkonomicheskii ocherk s prilozheniem tungusskih biudzhetov. Leningrad: Izdatel’stvo akademii nauk SSSR. Stammler, Florian. 2006. When reindeer nomads meet the market: Culture, property and globalisation at the end of the land. Münster: Lit. Stammler, Florian, and Hugh Beach. 2006. “Humans and reindeer on the move.” Nomadic peoples 10 (2): 628. Stépanoff, Charles. 2011. “Saillances et essences: Le traitement cognitif de la singularité chez les éleveurs de rennes tozu (Sibérie méridionale).” L’Homme 200. 175–202. ———. 2012. “Entre piétin et loups: Menace interne et menace externe dans l’élevage de rennes des Tozu.” Cahiers d’anthropologie sociale 8: 137–51. Takakura, Hiroki. 2010. “Arctic pastoralism in a subsistence continuum: A strategy for differentiating familiarity with animals.” In Good to eat, good to live with: Nomads and animals in Northern Eurasia and Africa, edited by Florian Stammler and Hiroki Takakura, 21–42. Sendai: Tohoku University, Northeast Asian Study Series. Thorisson, Skarphedinn. 1984. “The history of reindeer in Iceland and reindeer study 1979–1981.” Rangifer 4 (2): 22–38. Tomasello, Michael. 2006. “Why don’t apes point?” In Roots of human sociality, edited by Nicholas J. Enfield and Stephen C. Levinson, 506–524. Oxford: Berg. Vainshtein, Sevyan. 1961. Tuvintsy-Todzhintsy. Moscow: Izd. vost. lit. ———. 1980. Nomads of South Siberia: The pastoral economies of Tuva. Cambridge: Cambridge University Press. ———. 2009. Zagadochnaia Tuva. Moscow: Domashnaia gazeta. Un “engagement conjoint” homme-animal dans un système d’élevage de rennes Résumé : La présente étude explore les fondements et les paradoxes de l’interaction coopérative dans un système d’élevage de rennes en Sibérie méridionale. Bien que la majorité des activités humaines soient des activités conjointes où les objectifs comme les actions des participants exigent transparence et connaissance commune, cet article se demande dans quelle mesure il est possible de construire engagement conjoint transspécifique avec un partage minimal des connaissances et de faibles moyens de communication. L’article montre comment les éleveurs parviennent à inciter les rennes, qui pâturent librement la plupart du temps, à revenir spontanément aux campements, ce grâce à une communication non verbale et en l’absence d’un plan d’action clairement partagé. Pour maintenir un engagement du troupeau envers eux, les éleveurs s’appuient sur les compétences cognitives et les désirs de leurs rennes, et plus généralement sur une autonomie animale. Paradoxalement, les humains ne peuvent domestiquer les rennes que s’ils les maintiennent à l’état sauvage. En dépit d’un rapport marqué par l’opacité communicationnelle et l’asymétrie radicale, rennes et hommes entretiennent un contexte coopératif renouvelé qui leur permet de mener des activités conjointes complexes telles que la monte. Charles Stépanoff is Assistant Professor (maître de conférences) at the École Pratique des Hautes Études (Paris). He holds the chair “Religions of Northern Asia and Arctic.” His research fields are Tuvan shamanism, anthropology of ritual, and human-animal cooperation. Since 2011, he is the coordinator of the International group of research (GDRI) “Nomadism, society and environment in Central and Northern Asia” (France, Russia, Kirghizstan). ___________________ 1. After studying western Tuva (Siberia) from 2002, I conducted fieldwork in the Tozhu area (eastern Tuva), in July-August 2008 and February 2011. Tozhu reindeer husbandry has been described by explorers Ørjan Olsen ([1915] 1921) and Douglas Carruthers (1914) and by a Soviet ethnographer, Sevyan Vainshtein (1961, 2009). Recently Brian Donahoe has conducted thorough anthropological research on Tozhu and Tofa reindeer herding institutions and their relationships to the environment (Donahoe 2002, 2004). 2. See King 2002: “Without deer there is no culture, nothing,” the words that grace this title are from a herder in Kamchatka. In the Taimyr Peninsula, from the 1950s onwards, many domestic reindeer were driven away by an increasing population of wild reindeer. The Soviet authorities considered hunting to be more profitable than herding in this area, thus herding declined and eventually collapsed, causing both a cultural and demographic catastrophe for the indigenous peoples of the peninsula (Baskin 2009: 161). 3. Although domestic reindeer have a different name (iv) from wild reindeer (taspanan) and are called mal (or “cattle”), Tozhu partly treat them as wild beasts. For instance, Tozhu say that only wild game (an) liver must be eaten raw. However, when they kill a domestic reindeer, they eat its liver raw as well. They would not eat the liver of cow or a sheep in this way. 4. According to the ethologist Baskin (2009: 76), tundra herders sometimes attract their sledge reindeer by pretending to urinate. 5. Formerly, the Northern Yakut also used urine and salt to attract reindeer in winter and tie them up (Gurvich 1977). 6. In the region of the Num-To Lake, the Nenets of the Forest go to the pasture every day to give their reindeer dried fish to eat. This practice has no other goal than to maintain the human-reindeer relationship and prevent the reindeer from moving away (Marine Martin, personal communication 2012). 7. The use of the relief as a natural barrier was observed among the Tofa in the 1920s (Petri 1927). 8. Detailed data collected by Rastsvetaev in 1927 among Even reindeer herders of the Tompo basin, who had a way of life similar to the Tozhu, seem to confirm the link between the size of the herd and the simplicity of the route (Rastsvetaev 1933). While households with small herds (around fifteen heads) accomplish circular routes between different fishing and hunting places, richer households (from 30–100 reindeer) tend to be bound to a single route, in one main valley. These data will be analyzed in more detail in another publication.

Rennesson, Grimaud, and Césard: Insect Magnetism This work is licensed under the Creative Commons | © Stéphane Rennesson, Emmanuel Grimaud, and Nicolas Césard. Attribution-NonCommercial-NoDerivs 3.0 Unported. ISSN 2049-1115 (Online) Insect Magnetism: The communication circuits of Rhinoceros beetle fighting in Thailand Stéphane Rennesson, Emmanuel Grimaud, Nicolas Césard, Centre National de la Recherche Scientifique The Rhinoceros beetle fighting scene in northern Thailand exposes a puzzling technique of bringing together human and animal action. We intend to show through the study of this game that some cases invite us to specify our understanding of the notion of communication. What can the breeder-players share with their beetles? The answer to this question is far from being self-evident because the partners of the game do not share at all the same perceptive and cognitive abilities. Amateurs agree on the fact that the beetles cannot be really tamed, but since they are sensitive to vibration one can, however, try to enhance their fighting potentialities and to guide them by tactile means during the fight. Drawing on radical questionings of the notions of signal and noise, we shall try to determine to what extent beetle fighting can help us to reconsider the debate on the possibility of presignaletic forms of communication. Keywords: Human-animal communication, vibration, signal and noise, beetle, Thailand All that is not information, not redundancy, not form and not restraints—is noise, the only possible source of new patterns. Gregory Bateson, Steps to an ecology of mind, 1972 Among the many popular games in Thailand whose principal actors are animals (insects, fish, birds, cattle), “Rhinoceros Beetle” fighting (Xylotrupes mniszechi tonkinensis, locally called kwaang) involves an unusual form of partnership between man and animal that will enable the notion of communication to be considered under a new light. Elsewhere we have shown that what is interesting about kwaang fighting is the difficulty for humans to control their beetles during fights in a purely mechanical way (Rennesson, Grimaud, and Césard 2011).1 The beetles are said by the amateurs to be capricious and unstable, at least considering their fighting mood. The players and the bettors know that their animals can at any moment exit the structure to which they are trying to confine them. One can even wonder if the entire playing device is not primarily devoted to fostering the ambiguity of the relation of control.2 As we shall later examine in more detail, one challenge for the player is to succeed in establishing and maintaining contact with his insect. It is by means of vibration—by creating with the insect the conditions for a “tactile and vibratory” form of communication—that this interaction acquires all of its substance. By so doing, the players follow a well-documented ability for vibratory communication among arthropods (see figure 1). Beetles, notably, can produce various kinds of stridulation depending on the different species (for example, by scraping their protothorax against their mesothorax as with the Rhinoceros Beetle). At the beginning of the twentieth century, Slovenian biologist Ivan Regen—who studied field crickets’ behavior—was the first to recognize that stridulation was a means of communication between insects.3 On this account, kwaang fighting deserves its place in an alternative history of communication systems that is not based on an overly strict definition of communication, but instead draws from more diverse cases, including interspecific ones. In the long history of man-animal communication, the feats performed by the well-known “Clever Hans”—the horse that knew how to count—have become a kind of turn-of-the-twentieth-century case study.4 Although it was proven that Hans was neither telepathically guided, nor gifted with conceptual intelligence, he was still able to give correct answers by means of visual clues that his questioners unwittingly provided as they performed micromovements, to which horses are particularly responsive. On the basis of a kind of analogic communication,5 the importance of which among mammals has been highlighted by Gregory Bateson (1972), the questioners were in fact suggesting the solutions to Hans without knowing it, as the variation of their muscular tension signaled to the horse that his counting was approaching the correct answer. Figure 1: One of the two players meditates on the effects of the vibration he produces on the beetles by way of rolling his notched stylus on the log. Among other things, the “Clever Hans” case teaches us to recognize the complexity of communication, in both its conscious and unconscious aspects. While it is true that experts were at first victims of deceit and self-deception while developing the experimental apparatus, we are more interested in Vinciane Despret’s view (2004) that the horse had taught its breeder how to ask questions at least as much as the latter had trained the former to answer. Moreover, it is probably no coincidence that the collection of essays edited by Thomas A. Sebeok and Robert Rosenthal (1981)—on the subject of communication between humans and animals, specifically what this enables us to understand about interaction between humans—depends on case studies that only involve mammals (big apes, cetaceans, and horses). Needless to say, this is the part of the phylogenetic tree where we find animals that share the greatest propinquity with human beings in terms of their “subjective universe,” or Umwelt (Jakob von Uexküll [1934] 2010). It is there that we have the opportunity to observe the workings of logical continuities and discontinuities in the handling of symbols, clues, and icons.6 We would like to shift the question by examining a case that is even more puzzling, where the kind of analogic communication highlighted by Gregory Bateson (1972) among mammals seems impossible.7 In beetle fights, obviously, the cooperating partners do not share the same perceptional and cognitive abilities at all. Here the possibility of analogic communication is less spontaneously conceivable than it is between mammals, and it cannot be taken for granted that a shared framework of action exists between a human being and an insect. The game presupposes the participation of two species—beetles and men—whose perceptual worlds are very remote from one another. How can they communicate? With both humans and animals in mind, Bateson advocates a position opposite to Claude Shannon’s (1948) telegraphic and linear view of communication. Steps to an ecology of mind can be read as a reaction to Shannon’s reduction of interhuman communication to an exchange of messages between a pole transmitter and a pole receiver (Bateson 1972). Conversely, Bateson describes rather complex cybernetic circuits composed of numerous feedback loops in which signals circulate endlessly. But, as we shall see later, beetles make it quite tricky to decide exactly what constitutes a signal and whether or not there is any communication of that sort. Is there a communication before any signal is produced or is signal production a condition needed to establish a communication? “All that is not information, not redundancy, not form and not restraints—is noise, the only possible source of new patterns,” writes Bateson in “Cybernetic explanation” (1972). We know that a living creature, be it a man or an amoeba, does not need to be conscious to emit and receive a signal, to modulate physical magnitudes, and to interpret it. But Gilbert Simondon suggests another idea about the ontogeny of signal that enables us to go beyond the rather cryptic concluding sentence of Bateson’s article.8 As Simondon says, it would be restrictive not to see any communication in the simplest organized life forms responding to their environment and modifying it by doing so, even if there is apparently no exchange of signals as such. Organisms endowed with a central nervous system like human beings (and beetles, too, actually!) can combine this “ecologic communication” with a more elaborated mode, that Simondon calls “ethologic communication” (2010: 92–93). Considering acoustic and vibratory communication, Simondon argues that signals involved with ethologic communication do not constitute in themselves a mere conquest of intelligence over the basic needs dealt with at the ecological level. These signals are not merely selected by a more complex organism in its surroundings amid the background noise made up of a myriad of stimuli.9 As Simondon argues, one can observe a real continuity from one mode of communication to the other. Instead of being an obstacle to “ethologic communication,” the background noise can prepare individuals to receive signals (ibid.: 91). Simondon thus defines the “vibratory field” as a kind of primitive support for communication (ibid.: 93-96). Background noise functions as a sustaining energy whose modulations carry “motivation germs” and patterns of action. Contrary to what happens in communication through technical channels, Simondon suggests that background noise can be regarded as a potential source of stimulation and synchronization. Indebted to Bateson and Simondon for their radical questioning of the notions of signal and noise, we will see here to what extent beetle fighting can help us to reconsider the debate on the possibility of presignaletic forms of communication. The kwaang fight set-up can appear rudimentary at first glance. Two male beetles are placed on a wooden log that serves as the combat area. This log has two small holes that contain females and their pheromone is expected to excite the two males. The two players position themselves at either end of the log. They can then influence the behavior of their insects in three ways: (a) Through direct contact (players can touch their beetles with their fingers), (b) by means of a notched stylus that can be operated in two different ways (players can stimulate the beetle by touching it directly with the stylus or press the stylus against the log and nimbly roll it between their thumb and middle finger so as to produce vibration—see video 1 and video 2), and (c) by turning the log around its longitudinal axis (the player’s goal is to help his beetle find the best position in which to grab its opponent or escape its opponent’s grip). Video 1: Rhinoceros beetles are quite responsive to direct contact.11 Video 2: The log as a communication medium. We could limit ourselves to this simple technical description, but it is not without disadvantages. It confines the humans and insects to roles that poorly convey the subtlety of their interaction. Looking at the situation from the beetles’ perspective, we could say that they force humans to submit to the requirements of their Umwelt (Von Uexküll [1934] 2010). Beetles are sensitive to vibrations, so it is through vibrations that players communicate with their insects. Here we will show how the vibratory processes to which men are forced to conform can be seen as a communication test that takes the player to the limits of the possibility of pooling communication resources.10 The mechanistic temptation Let us first take a look at the direct contacts players establish with their insects. To grip a kwaang, one applies downward pressure with the finger on the posterior edge of the abdomen. This process enables the very efficient claws to be unhooked from the legs at the front and then from those in the middle and back if the gesture is prolonged. By lightly pinching the two upper legs with the thumb and forefinger, one can make the coleopteron let go completely. This technique is used to withdraw one’s beetle from the log between combat rounds. Among the forms of control available to the player, this action is certainly one of the most directive. The control relationship here is similar to that which one can have with an electric device by means of a switch. It is thus possible to place the beetle in a “hook/unhook” position. Yet, it would be a mistake to think that beetles can be controlled so mechanically while active, during a fight. This is only done outside of the action of the fight itself (see figure 2). Figure 2: In order to untangle the two beetles, one of the two players passes its stylus between his animal’s legs. By doing so, he intends to anesthetize the insect’s nervous system. Another kind of contact, perhaps even more authoritarian than the preceding one, is established when players grab their beetles during fights to give their insect a second chance in a fight that has started badly. The player catches the beetle by the upper horn—a continuation of its thorax and therefore linked to the animal’s body—between the thumb and forefinger and shakes it vigorously. Players say that this has the advantage of erasing its short-term memory. This gesture is used when one of the beetles runs away or exits the game and the player needs to bring it back into play. Released from the inhibiting burden of past traumas, the kwaang that got off to a bad start is now able to launch a fresh attack against its opponent. Even if it is difficult to evaluate the shake’s true impact on the beetle’s nervous system, this local theory on coleopteran “psychology” suggests an interesting avenue: to some extent, a kwaang could be reset just as one can press the reset button on a computer to unfreeze the system, which is then available for a new task (see video 3). With this action, one moves away from a stop/start control relationship and enters into another type of relationship. The beetle could be influenced by this way of reprogramming (that is, by refreshing its conditions of perception). Usually, after briefly shaking his kwaang, the player almost immediately repositions it on the log and makes it turn around two or three times directly above one of the two females inserted in the log. Then he stimulates it with the tip of the stylus, guiding it toward its opponent. The player is conscious that he is dealing with a complex, sensitive nervous structure that he can influence through something of a human short-circuit when the beetle is inhibited by the power relationship with its opponent. The shake extricates the beetle from the narrow circuit in which it has been inserted, to draw it into more specifically human kinetics. In the player’s hands, it experiences an agitation to which its nervous system must seldom be subjected in the wild. Video 3:Resetting one’s beetle. These moments of manipulation, which suggest that beetles can be influenced mechanically, nevertheless appear very sporadically. It is a matter of withdrawing one’s coleopteron from the game at moments of weakness, either between two rounds within a fight or between two fights. The rest of the time, the interactions that players develop with their insects are much more ambiguous in terms of control, and in the course of these, they clearly have no need for theories or a firm grasp of the beetles’ ability to process information and interpret the signals they receive. At no point can kwaang fighting be reduced to a mechanical system, to a remote control or string-puppet apparatus (see figure 3). Figure 3: The player starts to establish a connection with his kwaang by rolling his notched stylus on the sugar cane stick the beetle is tied to. Establishing a connection The following scene will enable a very close examination of how contact is established between a player and a beetle, as well as a better understanding of what is unique about their relationship in terms of control. We are in the home of a player who is preparing to train one of his beetles. The insects have been placed on sugar cane sticks on which they can feed. The player looks around, searching for a beetle that would be worth testing. After momentarily hesitating, he lifts one of the sugar cane sticks from the floor. He raises the gleaming, brown coleopteron to eye level and observes its silhouette closely. The animal does not react. Now its master rotates it in all directions, allowing it to be seen from several angles. With its head riveted to its support, the beetle continues patiently chewing the cane that the player peeled for this purpose. The player lightly taps the end of the insect’s lower horn, which drops almost immediately. He once again lifts the beetle to his eyes and sees that the horns have already closed again. The man has to make two attempts before his coleopteron is prepared to keep its horns apart. This position allows the player to assess the gap between its horns, on which much of his gripping ability depends. From its upper horn, he unties the cotton string that keeps it on its sugar cane stick. Although freed, the animal returns to its meal. At the moment, it seems oblivious to the vibrations in the sugar cane stick produced by the notched stylus, which the player is rolling between his thumb and forefinger. After half a minute, the man decides to tap the cane several times with the stylus as if to rouse his insect before once again rolling the instrument on the skin of the sugar cane. The contact between the cane and the stylus produces a soft, continuous hum. The player stops for a few minutes to observe the beetle’s behavior more attentively. It has stopped eating and has lifted its head but is still not moving. It has lost interest in the sugar cane, but the player is losing patience with the indifference that has greeted his attempts at remote stimulation. So he decides to touch the coleopteron directly with the stylus. He gives some sharp taps to both sides of the base of the upper horn that extends the animal’s thorax; it immediately leans to one side, then the other. Its movements are abrupt and look like reflexes more than anything. The beetle does not seem particularly inclined to cooperate. And yet the player, undeterred by his insect’s resistance, says it is reacting well. He moves the stylus away from the kwaang to vibrate it against the cane a few centimeters away. After a few seconds, the coleopteron finally starts moving its front legs, and then the legs in the middle and back follow that movement. It advances a few millimeters. It stops again and then timidly resumes its crawl. It takes a few forward steps. The player stops rolling the stylus, just long enough for the beetle to calmly arrive at the end of the sugar cane stick. The player rolls his stylus on the animal’s left legs and the animal starts turning to the left. He makes the beetle do a complete rotation before doing the same thing on the right legs, making it perform a half-turn. Now he places the stylus between the horns. With a tap on its head, the insect vigorously pushes away the stylus, and then quickly makes its way to the other end of the cane stick. The player withdraws the stylus in order to roll it again. The animal quickly reaches the other end of the cane. Clearly intrigued by the source of the vibrations, the kwaang attempts to move toward it but its front legs slip on the hard skin of the cane. It explores the surrounding area with its front claws for a few seconds then changes its mind and makes a half-turn. It advances a few centimeters and then stops again. Although the player redoubles his efforts—amplifying the rhythm of the vibrations by rolling the stylus against the fingernail on his thumb—the coleopteron is no longer reacting. The man therefore decides to resort to direct contact again. He makes the insect turn around a few times in each direction before placing his stylus between the horns to make it move forward, with success. When the kwaang starts to move, the player places the stylus back under its lower horn, resting the tip against the cane to block the insect’s path. The beetle tries to force its way through, backs up a little, then braces itself on its six legs and tries to get under the stylus in order to lift it. But the player cleverly follows the animal’s movements, keeping the stylus under the lower horn. By repeating the process six times, the man succeeds in making the beetle reverse a few centimeters and then he withdraws his stylus abruptly. The insect has been launched. The player follows its forward movement by rolling the stylus delicately between its horns. The coleopteron speeds up while giving taps of the head to an opponent that he fails to grab. The insect now appears to be obeying fully; now the player’s stylus almost never leaves it. The man makes it walk along the cane from end to end. He seems to be able to make it turn around, and to open and close its horns at will (see figure 4). Figure 4: The player now tries a direct contact to stimulate his kwaang by giving some sharp taps to both sides of the base of the upper horn that extends the animal’s thorax. It has taken the player several minutes to warm up his kwaang. Now he considers it sufficiently receptive to stimulation to test it against several opponents. These will not be real fights, but training situations in which he will attempt to gauge the specimen’s fighting potential and martial qualities. The player decides to lower it onto the wooden log that serves as the ring. He does this by tilting the cane stick, placing the end just above the female beetle that has been inserted in the log. Upside down, not far from the female, the beautiful brown beetle does not budge, utterly ignoring her. The player once again rolls the stylus against the cane hoping to get the insect moving again. It does not hesitate long. Once it arrives at the end of the stick it reaches out its front legs and ends up securing itself to the softwood log in order to land above the female. It goes no farther. While its antennae beat furiously above the captive, the player carefully sets the cane stick down behind him to find an opponent (see video 4). He quickly sets his sights on a beetle that is darker, slightly larger, and sturdier, but has a shorter posterior horn. Without showing any concern for the brown beetle whose full attention seems to have been captured by the female, the player now tries to awaken the fighting spirit in the future opponent. Video 4: Having one’s beetle get down on the log above a captive female. Left to fend for itself, the brown beetle abandons the female after a few minutes and moves several centimeters away from her. It stops momentarily before turning right. It then passes under the log and reappears on the other side before making a half-turn and disappearing again, only to end up returning directly above the female. Its antennae flutter for a few seconds and then it moves away two centimeters and freezes. The player is in the process of warming up its opponent but he notices the first beetle’s inactivity. So he tickles it with the stylus between its horns to get it moving again. After bringing it back to its position over the female, he leaves it again to concentrate on its opponent, which he lowers onto the log above the other female. Once both beetles are on the log, the player places his hand between the two insects to keep them from seeing one another. He makes one of them turn around and the other move forward, successively leading them to the females and then making them back away. As long as he is unsuccessful in making them simultaneously active, he keeps switching from one to the other, dealing with each of them for a few seconds at a time. When they show signs activity simultaneously, when they are walking or waving their antennae above a female, the player vibrates his stylus against the log. After two minutes of this, he decides to withdraw his hand, allowing the beetles to enter into conflict (see figure 5). Figure 5: Once the beetles are on the log, the players go on stimulating separately their beetle ahead of the fight. They alternate direct touches on the body with vibrations obtained by rolling their stylus on the log. By doing so, they try to give opportunity to their insect to smell the pheromones of the two captive females. Vibration as a medium This somewhat long description was necessary in order to show that players do not leave stimulation up to the presence of females and a rival male. We cannot limit ourselves to an etho-naturalistic understanding of the process, according to which males fight each other only for the right to cover the females. If this were the case, kwaang enthusiasts could bank solely on the insect’s ability to stimulate itself. But we know that in the wild, most meetings between two male beetles end in either avoidance or a relatively quick fight; in any case, the fight does not last long enough to produce a spectacle that kwaang players would consider worthy of that name. A monospecific setup that involved simply releasing two males onto a log would not provide the same quality of spectacle as an interspecific setup in which players pit their skills against one another, as is the present case, stimulating their protégés and sustaining their combativeness. Kwaang enthusiasts have instead chosen to move away from the configuration found in the wild through a clever cooperative system that allows them to prolong the fight, which can last as long as twenty minutes. The very subtle methods by which contact with the beetle is established, lost, and restored is the other significant point that is deducible from the preceding description. The challenge, of course, is to do everything possible not to lose it. It is through a set of very precise “microactions” that contact is established, even if the connection is by no means automatic. The player gets his coleopteron moving—not without difficulty—solely by producing vibrations with the stylus before the beetle is even placed in the presence of its opponent within a triangular configuration (two males + one female = a fight). He achieves this either through direct contact with the animal’s body or by relying on the conductive properties of the wood, which propagates the waves produced remotely through the rubbing of the stylus. The player infiltrates the animal triangle through the method of vibration, but this vibration is not reducible to that which another beetle would produce. One might think that it is only by imitating the presence of another male that a player could succeed in awakening his insect’s warrior instinct. But this ignores the beetle’s ability to distinguish between vibrations emitted by a stylus and those produced by another male. It would certainly be a mistake to bank on the presence of mental pictures, of representations of the male in the kwaang brain. A beetle’s cerebral structure is not sufficiently evolved. On the other hand, watching the preceding interaction, we notice that a vibration is not in itself perceived as a sign of the presence of another male. This vibration must accumulate and resonate with the actual presence of a male and females in order to produce its full effect. It is highly probable that the beetle never mistakes the rubbing of the stylus against the wood for an opponent’s imminent arrival on his path. Just as the fight configuration cannot be reduced to a mechanical control model, it would be overly simplistic to think that man’s role in the beetle’s world is that of a lure. Players know that kwaang respond to a call to action that does not necessarily have a stable motivation or meaning. If the meeting of two males either causes one or more of them to retreat or leads to a confrontation—quick though it may be—the player must persist for a long time before rousing the coleopteron’s interest, maintaining its activity, and keeping it from dwindling again. At this point in the demonstration, the image that probably best conveys what is at play is that of electricity, even if this will need some qualification. The beetle acts as something of a vibration accumulator. An intensity is produced between the player and the insect. But this intensity is not comparable to energy. Beetles do not stock energy like an electric accumulator. The player cannot bank on this energy the way he can count on the energy stored in the compressed spring of a watch. Keeping a beetle active cannot be likened to managing energy that has been transformed into mechanical energy. It is more a question of charging the animal with an intensity that it circulates by struggling against its opponent. And yet, in order to circulate intensity, what is needed is not just a generator but also a circuit composed of conductive materials. The conductivity of the wooden support is relatively weak, enough to cause signal loss. The player is never sure that the signals he sends are not losing intensity. When he enters into direct contact with the animal, either with his hands or through the stylus, he shortens the circuit and maximizes the opportunities for the intensity to circulate again. It is advisable to periodically “restart” the kwaang by touching it directly, reestablishing a circulation of intensity that can be remotely interrupted at any time. This explains the alternation between directly manipulating the animals and taking a step back from the action. The intensity can be maintained by vibrating the stylus against the log once the beetle has been sufficiently charged or until it discharges following a capture or a defeat during the fight itself. Signal and modulation It is quite clear that the player cannot control his beetle as one controls a puppet. The stylus does not produce a transmission of mechanical energy between them. Also, despite the precision this tool offers, the player is not in a relationship of the kind one can have with a remote-controlled robot or a programmed object requiring mastery of some form of coding (such that “under this condition, do this or that”). By establishing contact, a reflex arc is solicited from the insect, which carries out a quite precise operation, at least one that is more particularized than the mere forward walk produced by rolling the stylus against the wood. Next, through direct contact the player resets a smaller circuit with the animal that ensures minimal signal loss. But this resumption of contact between the player and his beetle only has value because the overall conductivity of the circuit must be reestablished at any price. To be persuaded of this, one need only take a closer look at how players release their beetles for the fight after the warming-up process or after restimulating them during the match. Knowing how to warm up one’s beetle is not enough; one should also know how to release it (that is, leave it to its own devices), or at least leave it to its opponent. The two opponents should be brought to this stage as directly as possible, in order to preserve the equity of the fight. If one of the animals presented the side of its body to the other, it would be getting off to a bad start. But this is not all. Players also make sure to break off contact with the animal at the right moment, sufficiently close to the opponent to give their insects as much speed as possible before they enter into conflict (see figure 6). Figure 6: The art of breaking off contact with one’s kwaang. Once the beetles have been released, the players are no longer allowed to touch their champions directly, except when the game has been stopped because at least one of the opponents is no longer in a combat position (having either backed away or left the combat area). When the two coleopterons confront each other, we enter into another phase of intensity production without necessarily abandoning the realm of contact. However, the nature of the contact changes. It is no longer interspecific contact between a stylus and a beetle or between a human hand and an insect, but monospecific contact between two animals that can no longer be considered simply accumulators (see video 5). But when the accumulation phase is successful, the conflict between the two produces spectacular reactions, with figures resembling karate holds. In these figures, it is not always easy to decipher which one has the advantage, and reversals of fortune are frequent. This is what makes the game interesting. At this stage, the players do not hesitate to vibrate their styli continuously. However, it would be simplistic to think that they do it only to support their own beetle. Of course, a player may vibrate the tip of his stylus a few centimeters from the ends of his insect’s legs to support its movements. Video 5: Releasing the beetles is a matter of switching kinesthetic contacts. But what is significant here is that players are never really sure which beetle benefits from the vibrations. The goal of the maneuver is to keep the circuit alive; it matters little whether or not one knows how connections are made. The beetles bathe in a single vibratory field, of which the players constitute the two poles. The log must be vibrated in order to sustain the circulation of a minimal intensity. The player can keep his stylus away from his insect or move it closer according to the situation. Many players choose to move their stylus closer when they get the impression that nothing is happening or, on the contrary, when something is happening. Modulating vibrations is necessary to this circuit, which is lengthened or shortened. It is by narrowing or extending the distance between one’s stylus and the two beetles that the insects are kept up and running (see figure 7). Figure 7: The two players sustain jointly the vibratory field. Having the vibratory upper hand Does the manner in which a player interrupts the hum of his stylus to tap the log stem from a different rationale? One is certainly changing frequency and producing a characteristic acoustic thump (“clack”) that breaks with all rubbing and humming. It would be easy to believe that this clack has the power to reset the beetle, similar to when the beetle is shaken in order to erase a traumatizing experience from its memory. But interviews conducted contradict this idea. One may wonder whether this thump is aimed principally at the beetles, at the players as a means of stimulating themselves, or even at the audience. In any case, one senses that this loud acoustic signal is not of the same nature as the pure vibration to which only beetles would be sensitive. It is not unusual to see small bells on the players’ styli, adding a pleasant rhythm to the spectacle. But no one can really say whether or not these have influence; they are more for appearances (see video 6). Video 6: Rolling one’s notched stylus with its bells on the log. On the other hand, the clack marks the upper limit of a vibratory spectrum, that which causes the beetles to perceive signals and to act. Everything seems to take place below the threshold of human aural perception. Beyond this frequency, no one can tell whether or not the beetles are receptive—to the audience’s shouts of encouragement for example. For players and spectators alike it is difficult not to participate in this game without imagining “mechanistic” responses. An immediate reaction is expected from the clack. From the player’s point of view, it gives the insect an incentive to strengthen its hold, to grip its opponent more forcefully between its horns. The fact that there is a link between the clack and the closing of pinchers does not necessarily make it a mere mechanistic action, from the point of view of its implications for the shared management of game intensity. As long as the pinchers have not closed—as the player has either anticipated or simply hoped—he will keep repeating his gesture until he obtains the desired result. Striking the log with the stylus brings the players into a different relationship with the action in terms of cooperation. The possibility of doing so at this point in the fight, of supporting the actions of one’s beetle and responding to them through a signal that will reinforce its impulse, opens a door to a whole set of strategic behaviors depending on whether the beetle is in an offensive or defensive situation. A player’s personality is very much linked to how he alternates rolling, striking, and clacking. It also very much depends on how he manipulates the log. In fact, players can directly act upon the log by turning it around its axis. Log manipulation is subject to a balance of power between the two players that merits closer examination (see figure 8, figure 9, and video 7). Figure 8: The players not only can use their styli to motivate their beetles; they can also help them in a more kinetic way by turning the log around its axis. Figure 9: When no advantage is recognized, manipulation of the log is subject to a balance of power between the two players. Video 7: When the stakes are high the game tends to come down to a balance of power between humans… Players tell us that by turning the log in the appropriate manner, they can give their beetle an advantage or correct its path, when at least one of the coleopterons is moving in the direction of the other. When the two beetles are above the axis, there is no point in manipulating the log. It is only when one of the two is not parallel to the log’s longitudinal axis that turning it becomes beneficial. One thereby seeks to position one’s beetle lower than its opponent so that its lower horn (the only one that is mobile) can pass under its opponent’s horn. When an advantage is recognized, the log can only be turned by the owner of the kwaang that is in a position of power (see video 8). This rule is ambiguous and should be discussed, since it has important implications. Is this because the advantages are not always discernible and one needs to facilitate recognition when they occur? How does one explain the withdrawal of the disadvantaged player’s influencing abilities? In light of what has been said above about vibration and the uncertainty surrounding the effects of its propagation, there is no guarantee that if the losing player continues rolling his stylus, he will not be compounding his disadvantage. Vibrations tend to favor the beetle in the dominant position. But to understand the log rule, it is also necessary to understand how its manipulation affects the circuit already patiently established by both players. This is the first time in the game that the player’s work is not merely a matter of emitting signals, but of supporting the action kinetically, by influencing the very support on which the interaction takes place. It is only after the intensity of the circuit has been established and is considered sufficiently stable that one may enter into this balance of power. Video 8: The beetle lifting its opponent is clearly recognized as getting the upper hand on it. The owner of the latter thus has to stop manipulating the log. At this stage, the two players transmit different forces to the log, which creaks and turns more or less from side to side. The position of the log results from the tension between two forces—between the system’s two human poles. It is no longer the shared management and joint production of a vibratory circuit; this has been installed. Now a pure balance of power is being expressed between two divergent human interests. This can be visible as such, but its expression may not be accepted in those terms for very long—as an expression of something human (that is, as merely the negotiation between the human components of the fight). Manipulating the log enables one of the two players to take the vibratory lead. The player whose beetle is regarded as having the advantage—that is, the one that has at least succeeded in passing its lower horn under that of its opponent so that it may grip it with its pincers—gains the exclusive ability to give the circuit a new injection of tension. Having the vibratory lead does not mean having greater ability to control one’s beetle. strictly speaking, it is the beetle itself that has the lead and has earned its player a vibratory advantage. And when the players detect a decrease in intensity they do not hesitate to take up their styli to reactivate the whole circuit (see figure 10). Figure 10: The player on the left is said to have the vibratory upper hand. Conclusion In kwaang fighting, communication between humans and beetles takes the form of a large circuit transmitting and circulating intensity, composed of five “loops”: a) two interspecific loops linking humans to insects, establishing contact by means of the stylus, b) two monospecific loops, one concerning exchanges between the two kwaang, and the other involving the two players in a balance-of-power situation and subjecting the players to changes in intensity, and c) an inclusive circuit represented by the wooden log, ensuring that the whole system reacts to the activity of the beetles and the two human competitors (see Appendix I). One might wonder if these loops are executed in series, in parallel, or even according to a third mode. Here we are no doubt reaching the limit of the electricity metaphor. If they were executed in series, player A’s vibration on the log would affect beetle A, which would transmit to beetle B, inducing an action from player B and so on. And yet this model—based on a chain of causes and effects with unvarying intensity from one end of the circuit to the other—does not work. In the case of parallel execution, the intensities produced by the different loops would have to add up or accumulate but this is not the case either. There is indeed a chronological chain of processes and methods for sustaining a vibratory field, but this field is highly random in its connections as well as in its intensity circulation modalities. It is absolutely essential for the interaction to be made longer than it would be in the wild, and this is why players need to experiment with different methods and vary the means of sustaining the system’s conductivity. In this context, at what point can it be said that players communicate with their kwaang? The beetle ultimately imposes its sensory universe while leaving humans with the possibility of appending code, meaning, and technique. The player is not attached to his insect in a purely mechanical way. Neither does he form one body with his coleopteron like a horseman and horse that share kinetic conventions and habits. The contact modalities of kwaang fighting do not allow themselves to be exhausted by the existing anthropological models ordinarily invoked to explain problems of cooperation at the boundary of the self (“distributed cognition” in the work of Edwin Hutchins [1996], the “cybernetic nature of the self” in that of Gregory Bateson [1972], or the “ecological psychology” of James J. Gibson [1977]). The gap between the motor and perceptual worlds of men and beetles is quite large and can only be narrowed by broadened communication theory that can more particularly build on a cognitive ethology applied to beetles. To cross the gulf separating them from their insects, players must submit to an unusual mode of communication. The kwaang impose their universe of low-frequency vibrations. In this context, beetle fighting is very much a theater of forced cooperation, but this cooperation is expressed in a circular way. The beetle is encouraged to cooperate with the man when the man is compelled by the insect to change his frame of reference and navigate an uncertain vibration field. The human player cannot rely on a simple rational understanding of the game. He is only allowed to shift from one level of communication to the other when the beetle is stuck in the sole modulation of physical magnitudes. Yet, between the two, there is no need for any sharing of mental images, representations, or joint attention frameworks. It is communication of a completely different kind, essentially based on establishing a vibratory flow, a vehicle for signals whose emergence, effects, and interpretability are never certain (see figure 12). Figure 12: A successful fight that is obviously stimulating both the players and the gamblers to frenetically negotiate odds and amounts! Appendix 1. Kwaang fighting device’s communications circuits diagrams.12 References Bateson, Gregory. 1972. Steps to an ecology of mind. London: Intertext Books. Benveniste, Emile. 1952. “Communication animale et langage humain.” Diogène 1: 1–8. Deleuze, Gilles. 1981. Cours à l’Université Paris 8 : La peinture et la question des concepts —Cours n°3 du 05/05/81. La voix de Gilles Deleuze en ligne. http://www2.univ-paris8.fr/deleuze/article.php3?id_article=83 ———. 2003. Francis Bacon: Logic of sensation. London: Continuum Books. Despret, Vinciane. 2004. Hans, le cheval qui savait compter. Paris: Les Empêcheurs de penser en rond / Le Seuil. Geertz, Clifford. 1973. “Deep play: Notes on the Balinese cockfight.” In The interpretation of cultures, 413–53. New York: Basic Books. Gibson, James J. 1977. “The theory of affordances.” In Perceiving, acting, and knowing: Toward an ecological psychology, edited by Robert Shaw and John Bransford, 67–82. Hillsdale, NJ: Lawrence Erlbaum. Hutchins, Edwin. 1996. Cognition in the wild. Cambridge, MA: MIT Press. Ingold, Tim. 1988. “The animal in the study of humanity.” In What is an animal? edited by Tim Ingold, 84–99. London: Routledge. Kojima, Wataru, Yukio Ishikawa, and Takuma Takanashi. 2012. “Pupal vibratory signals of a group-living beetle that deter larvae: Are they mimics of predator cues?” Communicative and Integrative Biology 5 (3): 262–64. Lestel, Dominique. 2002. “Langage et communications animales.” Langages, 36e année, n°146: 91–100. Pfungst, Oskar. (1911) 1965. Clever Hans (The horse of Mr. Von Osten). New York: Holt, Rinehart, and Winston. Rennesson, Stéphane, Emmanuel Grimaud, and Nicolas Césard. 2011. “Jeu d’espèces: Quand deux scarabées se rencontrent sur un ring.” In Humains non-humains: Comment repeupler les sciences sociales, edited by Sophie Houdart and Olivier Thiéry, 30–39. Paris: La Découverte. Sebeok, Thomas A., and Robert Rosenthal. 1981. The Clever Hans phenomenon: Communication with horses, whales, apes, and people. Annals of the New York Academy of Sciences. Shannon, Claude E. 1948. “A mathematical theory of communication.” Bell System Technical Journal 27: 379–423. Simondon, Gilbert. 2010. Communication et information. Paris: Editions de la Transparence. Von Frisch, Karl. 1950. Bees: Their vision, chemical sense and langage. Ithaca, NY: Cornell University Press. Von Uexküll, Jakob. (1934) 2010. A foray into the worlds of animals and humans. Minneapolis: University of Minnesota Press. Le scarabée magnétique : Les circuits de communication du jeu de kwaang en Thaïlande Résumé : Les combats de scarabées en Thaïlande constituent une sorte de laboratoire où on expérimente des formes de coopération pour le moins insolites entre humains et insectes. A partir de la description des modes de contacts entre les joueurs et leurs animaux, les auteurs montrent comment des cas d’interactions interspécifiques obligent à préciser ce qu’on entend par communication. Quand on songe que les capacités cognitives et perceptives des deux êtres sont fort différentes, on peut se demander ce que peuvent partager les amateurs et leurs coléoptères. Les amateurs s’accordent toutefois pour dire que leurs gladiateurs miniatures ne peuvent pas être dressés mais qu’en tirant profit de leur grande sensibilité aux vibrations, on peut maximiser leur potentiel martial. En s’appuyant sur un questionnement radical des notions de signal et de bruit, l’article montre jusqu’à quel point les scarabées relancent le débat autour de l’existence de formes de communication pré-signalétique. Stéphane Rennesson teaches anthropology at the Institut d’Etudes Politique in Paris. As the result of long participatory fieldwork in Thailand as a boxer, a trainer, and a promoter, he has written an in-depth ethnography of the world of Thai boxing in its country of origin. More recently, he has been working on numerous other games that are very much structured and that have met popular success in Thailand. He is mainly interested in competitions that require uncanny collaborations between humans and various animals, studying animal fights (beetles, cocks, fishes, buffaloes) and birds’ singing contests. Emmanuel Grimaud has worked on various subjects, including the first ethnography of the Bombay film studios (Bollywood film studio, CNRS Editions, 2004), a duplicate of Gandhi (Le sosie de Gandhi, CNRS Editions, 2007), Indian robotics and mechanical gods (Dieux et robots, L’Archange Minotaure, 2008), and Japanese humanoids (Le jour où les robots mangeront des pommes, Petra, 2012). More recently his research has been focusing on the anthropology of eye movements. He also founded the ARTMAP (www.artmap-research.com), an international collective of researchers and artists experimenting with technology between art and science. He received the bronze medal of the CNRS in 2011. Nicolas Césard has done ethnographic and ethnohistorical research mainly in Indonesia. His PhD thesis has shed light on the processes that, for a century, have led the northeastern nomadic groups of Borneo to transform their subsistence ways of life in favor of commercial strategies, a sedentary lifestyle, and the adoption of new social dispositions. For several years he has examined the interactions between societies and their environment, as well as the management of natural resources, through the field of ethnoentomology. Drawing on a comparative analysis of interactions between people, insects, and the natural environment at large, he takes interest in ecological, biological, and technological dimensions of people’s relations to nature, both on pragmatic and symbolic standpoints. ___________________ 1. This article is the result of collective fieldwork conducted by Stéphane Rennesson, Emmanuel Grimaud, and Nicolas Césard. Other forms of “species game,” such as fish fights or birdsong competitions, involve very different approaches to animal communication. This comparative perspective will be developed in future articles. 2. Nurturing the uncertainty of a communication device is far from being a distinctive feature of interspecific relations. It can also prove to be fruitful among interactions between humans as accurately showed in this themed section by Matthew Carey as he analyzes flirtatious relationships in Morocco. 3. Besides, as was shown by Japanese biologists, pupae and larvae of some kinds of beetles communicate through vibrations in the humus where they dwell. Their study gives credence to the sensibility of these animals to vibrations that propagates in a solid environment (Kojima, Ishikawal, and Takanashi 2012). 4. Bred and trained by Wilhelm von Osten, who was a mathematics teacher, Hans in fact displayed some bewildering proficiencies since he was able to solve mathematical problems, give the date, point out an individual in a crowd from a picture, and even spell words by means of a numeric code. He would answer by hitting the ground with his foot as many times as necessary. These equine feats were at the heart of a controversy that shook the Berlin scientific community for a few months from 1904 to 1905 (Pfungst [1911] 1965; Despret 2004). 5. Following Charles Sanders Peirce’s typology of signs as ways to denote an object (icon, index, and symbol), theories of communication tend to make a difference between analogic/analog forms of communication and digital forms. Analogic coding encompasses every avenue of communication that is not verbal: body movements, gestures, postures, facial expressions, etc. With regard to humans, it also includes paralinguistic elements such as voice inflections, rhythms, and intonations. Analogic communication essentially refers to social relations and is a matter of correspondences of magnitude. It is all about the modulation of intensities. On the other hand, digital coding, that is to say verbal and articulated language, enables to depict the state of the world and is operated through a purely arbitrary and conventional association between signs and what they stand for. For a discussion on the hybridization of digital and analogic codes, see Bateson about dolphins (1972) and also Deleuze’s comment on Bateson’s dolphins (1981) and on Francis Bacon (2003). 6. Researchers have also been looking for proof of symbol-handling by social birds and insects. Regarding the latter we could point to the famous controversy between Karl Von Frisch (1950) and Emile Benveniste (1952) over bee language, discussed by Tim Ingold (1988) and Dominique Lestel (2002). 7. The possibility of a communication between human beings and mammals is comprehensively discussed in this volume by Charles Stépanoff’s study of reindeer herding system. 8. In his courses at the Sorbonne in the 1970s, Gilbert Simondon called for a broader theory of communication not only concerned with humans but also including the way bacteria communicate with their environment, the various modes of communication of animals (vibratory, acoustic, etc.), as well as the latest communication networks produced by humans (2010). 9. On this problem, see Jakob von Uexküll ([1934] 2010) and James J. Gibson (1977), who approach motivations among animals in terms of qualities and affordances. Both of them invite us to consider these motivations as closely connected to the physical properties of the environment, paying attention to the way species are engaged in a selective scanning of the features offered to them by their surroundings, according to their cognitive and perceptual abilities. 10. As such, we assume a totally different approach from Clifford Geertz (1973). Instead of considering animals embedded in a human game and manipulated as symbols in a cultural text to which they are alien, we try to be sensible to the way the player is forced to adopt a certain mode of communication, becoming his own beetle’s extension. 11. All videos are available for viewing on the Hau: Journal of Ethnograhic Theory website. To watch a video, click the video screenshot. 12. A PDF of the Kwaang fighting device’s communications circuit diagrams is available online: http://wvww.haujournal.org/extras/kwaang_circuit_diagrams.pdf.

Wagner: The chess of kinship and the kinship of chess This work is licensed under the Creative Commons | © Tony Crook, Justin Shaffner. Attribution-NonCommercial-NoDerivs 3.0 Unported. ISSN 2049-1115 (Online) Preface Roy Wagner’s “Chess of kinship”: An opening gambit Tony Crook, University of St. Andrews Justin Shaffner, University of Mary Washington   Usually, the peer-review mechanism of a scholarly journal is designed to ensure that an article can stand for itself. In the unusual circumstances that the editors take the step of augmenting an author’s piece of writing with a short foreword, this is often intended as either an introduction, an explanation or a justification of the piece - that is, to exert some controlling effect on the meaning of the article. This brief text intends nothing of the kind, and for best results should not be read that way. We offer up a few ideas, whose usefulness is restricted, only to provide an opening to that which follows, and which is sure to clear these away. The publication of “The chess of kinship and the kinship of chess,” a chapter from a new manuscript entitled, The place of invention, falls on the heels of a visit this summer (2011) to Brazil, where in addition to participating in a series of seminars and lectures, Roy Wagner also engaged in a symmetric anthropological exchange with indigenous shamans and leaders—such as Davi Kopenawa Yanomami and Mauricio Yekuana in Rio de Janeiro, and Iginio Tenorio Tuyuka in Manaus and in nearby communities. The shamans there on the Rio Negro initiated Wagner into their own form of knowledge-practices, including the psychoactive kahpf, which has its own capacities to spin initiates’ heads, in ways not unlike the effects of Wagner’s writing. Although Wagner’s “The Chess of Kinship” adopts what may appear to be an unconventional format and strategy for making its point - for example, readers looking for an introduction or conclusion will be disappointed - there are no missing pieces and there is no larger puzzle picture. Readers will soon realize that they provide a component piece and have to play their part. Neither kinship nor chess amount to, or provide, an “end point” for the argument, but rather the interrelations set up the “means” by which the capacities of one can be depicted or figured through the other. Disconcertingly perhaps, many of the insights are created with the reader, and if unable to point to a point they will at least understand their own part in the pointing. That kinship is not a self-contained domain that can be modeled, but rather a set of relations which provide alternating views of what the relations, and the persons caught up in, or caused by them, can be made to do, is the simplistic point which provides the frame for the inter-actions here. The argument, much like the author’s favorite chess move the Knight-Fork (the Knight being the only piece on the board that must end a move with a shift from white to black, or black to white), develops in one frame or register before using that position as a platform to jump to another frame or register. Both kinship and chess appear, and disappear, and - in the illustrations, data, references and asides in between - it is always clear that both are nonetheless present, and that each gets illuminated, or forked, by the other in turn. The artifice of using each to mediate and to think about the other is acknowledged and taken seriously as a method. Perhaps in more familiar language, we might say that form and content have been productively fused. There are two methodological aspects here: the first concerns a refusal of any generalized relation between model and reality, the second concerns the fashioning of analysis after particular Melanesian knowledge-practices. Rather than adopting a theoretical method to frame or access a domain of social life called “kinship” then, the inter-play in Wagner’s article suggests an innovative method to get at what human kinship is all about. This point, and the language here, go back to David Schneider who might be said to have invented the “New Melanesian Ethnography” nearly fifty years ago—for in Joel Robbins’ terms “there is no doubt that what makes the New Melanesian Ethnography distinctive as a way of doing ethnography is its insistence that theory be made out of materials that one finds in the same place one finds one’s data” (2006: 172). In Marilyn Strathern’s terms, the challenge for anthropology is to attempt to “imagine what an indigenous ‘analysis’ might look like if we took seriously the idea that these islanders might be endorsing their own theory of social action” (1992: 150). In other words, mid-twentieth century anthropology’s scrupulous separation of theory and data has been methodically reformulated such that more weight is given to the theory of social action, of say, Papua New Guinea’s Mount Hageners than, say, Philosophy’s Martin Heidegger. This insistence on finding, making and relying on theory derived from the same place as the ethnographic data, has also led to an important point here: the refashioning of anthropological analysis after Melanesian forms so as to demonstrate or convey their effect rather than merely trying to describe it. The method here seeks to go beyond the deployment of vernacular terms in formal or informal academic models, and so develop analyses on the terms and in the forms of Melanesian creativity. That is, not to deploy ethnographic content to illustrate forms of academic creativity as an end in itself (Wagner 1981), but to develop the revelatory capacities of Melanesian forms as a means to describe how local creativities (Melanesian and academic) differently exploit what, for talk’s sake, could be said to be the (con)fusion between form and content. Schneider’s simplistic point was to frame anthropological “kinship” theorization as the projection of a hypothesis or model onto the materials under study - and the counterpart conceit of being taken in by, or muddled by, and then forgetting that the effect of a theory providing a measure of reality was self-induced. Schneider was puzzled why this turning back on itself did not appear as circularity in the argument. For Schneider, the British structure-functionalists and unilinear descent theorists of the 1940-50s had created a persuasive and powerful way of talking about kinship-theory which relied upon the device of making the theoretical figure or model and the ethnographic reality or ground appear to be the same thing—through the counterpart insistence that theory and data were clinically isolated. The trick required the development of a specialized language in order for the knowledge to be effective by, so-to-speak, turning back on itself whilst eclipsing or concealing the artifice. It became obvious to Schneider that people on Yap had created their own tricks and language for talking about, and doing, what we might call “kinship” (cf. Bashkow 1991). Schneider drew attention to the origin of ideas, to the means by which theory and data configure each other, and thus to theory sourced from the field and method sourced from relations. Wagner’s development of this point came in the following terms: What we need is not a model of how symbols interact with “reality,” but a model of how symbols interact with other symbols. And clearly, since a commitment to “reality” is so persistent and insidious among anthropologists as well as their subjects, such a model must account for the fact that some symbolic expressions are perceived as ‘reality’ whereas others are not (1978:21). Schneider’s pointed finger then, directed attention to what theory does rather than what theory is or thinks it is about. This simple idea could be said to have provided the opening move to the “New Melanesian Ethnography”—and yet also be responsible for the effect of how serious theoretical work can appear as a game whose proponents appear to be concerned with the ideal rather than the actuality (Josephides 1991). In the hands of Marilyn Strathern (1988), Hageners have taught anthropologists not to read “gender” from what something is (e.g. a woman or a man) but what something does (e.g. has an effect taken to be female or male, according to local idioms). Similarly, in Tony Crook’s work, Angkaiyakmin articulate a theory of knowledge whereby “meaning” derives from the effects of what it does—and shows why anthropologists with their own ideas about knowledge should have had such a hard time trying to pin down meaning (2007). Read with familiar thinking in mind, such writings as these and “Chess of Kinship,” can produce a disorientating experience, but one suggestive of an alternative to the muddle of models. Roy Wagner got and developed Schneider’s simplicity, and has been reworking the point ever since he received a copy of “Some muddles in the models” (1965, and reprinted in this volume) whilst he was in the field with the Daribi. The Curse of Souw (1967) turned descent and alliance theory inside-out, and contains a nascent symmetrical anthropology in equating and thus making the anthropologist’s models analogous to Daribi’s symbols (see also Corsín Jiménez this volume). This was developed through Habu (1972) and The invention of culture (1981), towards a fully-fledged “theory of symbolic obviation” in Lethal speech (1978). As for Schneider, kinship has been a consistent venue for Wagner’s re-workings, and the underlying critique of “models” in the present piece is evident in “Are there groups in the New Guinea Highlands?” (1974), and “Analogic kinship” (1977) which pursued the obviational theory of life-cycle exchanges before it was laid out in Symbols that stand for themselves (1986). Throughout this body of work, Wagner’s moves in anthropology were perhaps less obviously a prolonged experiment with our own forms of knowledge modeled after Melanesian ones, but there can be no mistaking this in the work that has followed. At this point too, the mutual influence of other Melanesianist scholars, especially Jadran Mimica (Intimations of infinity, 1988), Marilyn Strathern (The gender of the gift, 1988 and Partial connections, 2001), and James Weiner (The heart of the pearlshell, 1988) should not be under estimated. As Wagner writes in An Anthropology of the Subject “The fieldworkers who retrieved the data on holographic perspectives in Melanesia were surprised by it, and often, as in my case, it took them years to figure out what they were looking at” (2001: xxi). In the 1990s, then, the “New Melanesian Ethnography” took an important turn through the influences of ethnographic materials appearing to be holographic, of fractal mathematics as an analytical metaphor, and of experiments in re-fashioning theory after Melanesian pragmatics. Wagner himself claims (2011: 121) that what he latter termed “obviation” was originally taught to him by the Daribi, specifically his main myth informant, Yapenugiai, and in effect amounts to a kind of Daribi “language ideology” or theory of talk (“po” = “information,” “language,” “speech”) recast in terms of symbols (Wagner 1978), one that makes explicit that talk has the capacity to posit its own figure and ground, including their reversal. This allowed Wagner to re-perceive anthropology itself as a kind of po or talk—a particular technique of “description” as Marilyn Strathern puts it (1999: xi-xii)—and to also re-position it outside the problematic social constructionist paradigm of “one nature, many cultures.” “The Chess of Kinship” deploys a kind of Melanesian “power talk” whose pragmatic effect is obviation or figure-ground reversal. The Daribi call it porigi, (po begerama pusabo po) “the talk that turns back on itself as it is spoken” (Wagner 2011: 122). Regarding their own forms of revelation, the Barok of New Ireland told Wagner that “When you realize the secret of the pidik, you stand not at the end but at the beginning of knowledge.” The Foi call it “tree leaf talk” (irisae-medobora), “words that conceal their own base or grounds, as tree leaves hide from sight what goes on behind them” (Weiner 2001: 164). The Telefomin also employ the method of turning words and “spinning initiates heads” in order that their dizziness may eventually induce clear vision (Jorgensen 1990). Wagner’s recent work has revisited Schneider’s questioning of models in anthropology, and argues that, rather than achieving some perfect equivalence between parts and whole, Melanesian analogics suggest that holography works precisely by resisting anything but an imperfect or scale-model of itself. And if this were not reminder enough for anthropologists to remember their contingent role in measuring their own models via the life-worlds of other peoples, Wagner has since opted to explore the insights and exploit the effectiveness of Melanesian knowledge-practices, and has developed a dialogic or “chiasmatic” basis to his exposition (2001, 2010). Initially, this can feel to a reader like a Castanedan apprenticeship or cultic instruction session, and that is rather Wagner’s intent. But the incentive for the willing reader is that these Melanesian—or Mesoamerican—forms of exegesis might just pin down or knock the conventions of academic thought sufficiently off-balance to offer a future glimpse of human creativity and of an anthropological science game enough to match it. Wagner’s experiments in strategically deploying Melanesian informed techniques of anthropological writing constitute a kind of ethnographic theory whereby Melanesian forms and preoccupations are inflected within our own. Such controlled equivocations (Viveiros de Castro 2004) have inspired other variations of the Knight-Fork move elsewhere such as perspectivism—itself a figure-ground reversal of Euro-American multiculturalism—originally developed out of an encounter with Amerindian forms of thought (Viveiros de Castro 1998), and also Symmetric Anthropology (Abaete Manifesto). In Brazil, Wagner also met with indigenous students at the Universidade Federal de Amazonas in Manaus, who are conducting their own “reverse anthropology” researches inspired by The invention of culture, which was recently translated and published in Portuguese (2010), using analogies drawn from Amerindian, rather than Euro-American thought. And so the experiment with anthropology continues. In the “Chess of Kinship,” Wagner demonstrates and conveys one possible move in this game, a revelation or display of knowledge that requires a different kind of work and care on behalf of the reader whose involvement here is differently positioned and requires playing by different rules. References Abaeté Manifesto. https://sites.google.eom/a/abaetenet.net/nansi/abaetextos/manifesto-abaet%C3%A9 Bashkow, Ira. 1991. “The dynamics of rapport in a colonial situation: David Schneider’s fieldwork on the islands of Yap”. In Colonial situations: Essays on the contextualization of ethnographic knowledge, edited by George Stocking, 170–242. Madison: University of Wisconsin Press. Crook, Tony. 2007. Anthropological knowledge, secrecy and Bolivip, Papua New Guinea: Exchanging skin. Oxford and New York: Oxford University Press, British Academy Postdoctoral Fellowship Monograph series. Jorgensen, Dan. 1990. “Secrecy’s turns.” Canberra Anthropology 13(1): 40–47. Josephides, Lisette. 1991. “Metaphors, metathemes, and the construction of sociality: A critique of the New Melanesian Ethnography,” Man (n.s.), 26, 145–61. Mimica, Jadran. 1988. Intimations of infinity: The counting system and the concept of number among the Iqwaye. Oxford: Berg Publishers. Robbins, Joel. 2006. Book review of Leach, J., The Creative Land (2003), and Read, A., Papua New Guinea’s Last Place (2003). Contemporary Pacific 18 (1): 171–175. Schneider, David M. 1965. “Some muddles in the models, or, how the system really works.” In The relevance of models for social anthropology, edited by M. Banton (A.S.A. Monograph I). London: Tavistock. [Reprinted this volume] Strathern, Marilyn. 1988. The Gender of the gift: Problems with women and problems with society in Melanesia. Berkeley: University of California Press. ———. 1992. “Response.” Book Review Forum, Pacific Studies 15(1): 149–59. ———. 1999. Property, substance and effect. Anthropological essays on persons and things. London: The Athlone Press. ———. (1991) 2004. Partial connections. Updated edition. Walnut Creek: Altamira Press. Viveiros de Castro, Eduardo. 1998. “Cosmological deixis and Amerindian perspectivism.” The Journal of the Royal Anthropological Institute 4 (3): 469–488. ———. 2004. “Perspectival anthropology and the method of controlled equivocation.” Tipiti: Journal of the Society for the Anthropology of Lowland South America 2 (1). http://digitalcommons.trinity.edu/tipiti/vol2/iss1/1 Weiner, James F. 1988. The heart of the pearl shell: The mythological dimension of Foi sociality. Berkeley: University of California Press. ———. 2001. Tree leaf talk: A Heideggerian anthropology. Oxford ; New York: Berg. Wagner, Roy. 1967. The curse of Souw: Principles of Daribi clan definition and alliance in New Guinea. Chicago: University of Chicago Press. ———. 1972. Habu: The innovation of meaning in Daribi religion. Chicago: University of Chicago Press. ———. 1974. Are there social groups in the New Guinea Highlands? In Frontiers of anthropology. An introduction to anthropological thinking, edited by M. Leaf. New York: Nostrand Company. ———. (1975) 1981. The invention of culture. Chicago: University of Chicago Press. ———. 1977. “Analogic kinship: a Daribi example.” American Ethnologist 4: 623–642. ———. 1978. Lethal Speech: Daribi Myth as Symbolic Obviation. Symbol, Myth and Ritual Series. Ithaca: Cornell University Press. ———. 1986. Symbols that stand for themselves. Chicago: University of Chicago Press. ———. 2001. An anthropology of the subject: Holographic worldview in New Guinea and its meaning and significance for the world of anthropology. Berkeley: University of California Press. ———. 2010. Coyote anthropology. Lincoln: University of Nebraska Press. ———. 2011. The place of invention. Unpublished manuscript. This work is licensed under the Creative Commons | © Roy Wagner. Attribution-NonCommercial-NoDerivs 3.0 Unported. ISSN 2049-1115 (Online) The chess of kinship and the kinship of chess* Roy Wagner, University of Virginia The real comparison between the anthropological study of kinship and the game of chess is not immediately apparent from their formal properties, and only becomes relevant when they are viewed as strategies, or patterns of events occurring in time. The single “proportion” that both share in common is a kind of cross-comparison between dualistic variables called a chiasmus, illustrated in kinship by the classic cross-cousin relationship, and in chess by the asymmetric double-proportion between the king and queen, the only gendered pieces on the board, and the moves and tokens of the other pieces in the game. The difference may be summed up in the word “mating.” Chess may be described as the “kinship” of kinship. Failure to understand the chiasmatic, or doubleproportional essence of both has resulted in many dysfunctional models of cross-cousin marriage, and many very quick games of chess. Keywords: Kinship, chess, humour, knowledge practices, cross-cousin marriage, strange attractor, Melanesia The standard anthropological representation of “kinship,” first presented by Lewis Henry Morgan in 1871, was a static pattern, useful for strictly comparative purposes, and it set the discipline on a self-inertial or “structural functional” trajectory, in which pattern, consistency and especially relativity became key points of reference. As such it stood in sharp contrast to the way in which human lives are actually lived and thought upon, the patterning of events and the design of strategies. Both the field of play and the cast of players are part of the overall design, an arbitrary framework for the working out of fortunes. In kinship you mate at the beginning of the game, in chess you mate at the end. The word “mate” has a very different etymology in each case, and a very differentmeaning, but it is the same sound used in much the same way, and it is very strategic. In chess you start out with all your personnel there at once, ranked and ordered in a very specific way, and with some exceptions you proceed to diminish their numbers as the game progresses. In kinship you start out conceptually with very few personnel, and proceed to multiply their numbers as timeprogresses. Then you proceed to rank and order them according to very specific categories—generally genealogies and lineages rather than knight, bishop, rook, and pawn. (You can always get more pawns at a pawnshop.) In chess there is a single strategy, known only to the game itself, and each player tries to figure out what it is. When one of them gets it—it gets the other one. In kinship there are a great many potential strategies, but everyone, including the anthropologist, thinks they know the one that counts. Or at least pretends to, because you don’t want the other players to see your cards, and most of them are bluffing in any case. And the one who dies with the most strategies, wins. Chess is a very high profile intellectual game, but in the end it is not really intellectual at all—more of a Jedi mind-trick, or like the role of a grafter—the kind of con-man who preys on other con-men in Redford and Newman’s movie The sting. The grifter holds the middle ground between kinship and chess, and that middle ground is called strategy. Whose strategy? That’s it, you just got it in one! Kinship and chess are parallel worlds, defining highly specific social, cultural, and physical contexts, which are only slightly overlapping. Nonetheless, there is at least one way in which they are one and the same thing. And, though “strategy” pretty much covers it—this being the main reason anyone bothers with kinship, apart from sterile classificatory games, and the main reason anyone bothers with chess, apart from of course winning—let us go on. Of course, nobody ever bothers about winning and losing in kinship, no sirree! Just a few highly disreputable individuals, some of them called “men” and others “women.” Oh, and I almost forget, “children.” There are many things about kinship that are not true of chess, and many things about chess that are not true of kinship. But there is one thing about both that most compels us, a thing that makes comparisons paradoxical, and paradoxes therefore comparative. This is the chiasmatic, or double-proportional comparison, the thing that Tony Crook (1997), who discovered it at Bolivip, Papua New Guinea, while decrypting the secretive “Mother House” complex, calls “changing the subject in mid-sentence”—a sort of syntactical “cross-cousin marriage” if the comparison be allowed. The Daribi people, also of Papua New Guinea, call it porigi, and describe it as po begerama pusabo poin their language, “the talk that turns back on itself as it is spoken.” When asked: “What makes a man a big-man, is it having a lot of wives, or pigs?” Daribi will respond: “A man who can talk porigi effectively gets all the wives and pigs he wants.” Before I go on to demonstrate the pivotal role that chiasmus plays in both chess and kinship, it might be helpful to understand just exactly what kind of strategy it involves. For example, the ancient Greek philosopher Heraclitus made his reputation by uttering cryptic statements like: “We live the gods’ deaths and they live ours.” Does this mean that Heraclitus knew some secrets about human beings and gods, or their strange relations, that others did not know? Hardly, Heraclitus was as entirely innocent of this kind of knowledge as you or I; he just knew how to use the ergative well, and use it in a chiasmatic strategy. An ergative expression is one in which a conventionally active action or verb is displaced into a passive role, with an exponential gain in power and emphasis. Note that “living someone else’s death” is an exemplary ergative of this sort, rather like “dying on the job” in a worker’s paradise, and, when used chiasmatically in a double proportional comparison, it conjures a powerful ironic effect, as in the late Soviet joke: “We pretend to work, and the State pretends to pay us.” In other words the best way to keep—or even invent—a secret is to make it a function of the form (like the porigi) rather than the content. In this way the formal strategy shared between chess and kinship is more like a coup in topology and mathematics than it resembles the “relational” or emotional gambits favored by so-called “humanists.” Let me illustrate: stated both symmetrically and asymmetrically at once (in the way that Heraclitus configured his prophetic utterances), the secret that there is no secret becomes a sort of half-truth about itself, and therefore a double- truth about anything else—more or less what the fractal mathematicians call a “strange attractor.”1 (The Elders of the Mother House at Telefolip, P.N.G., actually use this device as a major teaching strategy for their initiates (cf. Jorgensen 1981), and in 2000 Mike Wesch and I caught one of them in the act of trying to use it—called in that case “The Two Dolls”—on us).2 So there is a problem with this double-jeapardy illusionism after all; what did the Elder do to Mike and I that he had not already done to himself? Did the strange attractor called The Two Dolls not control him as well? And how did his self-conceit in this way differ from that of the grifter, the sleight-of-hand magician, the chess grandmaster, or for that matter the kinship-expert? It was just a way of talking, to be sure, but then Freud called his psychoanalysis “the talking cure.” And the game of chess is likewise said to be very educative. Where do we find the double-proportional chiasmus in the chess game? The layout of chess is a study in contrastive symmetries; there are two sides (or players), white and black squares (8 x 8) arranged in a totally symmetrical format, and each player begins with a symmetrical layout of pieces (bishops, knights, rooks, and pawns, traditionally called “men,” but not explicitly gendered). These are the “four arms,” priesthood, cavalry, fortification, and infantry, in the military regimen of ancient India, where the game originated. And then there is the other proportion, and that is the one dictated by the only explictly gendered pieces on the board, which, according to the rules, must face each other across the board—an asymmeùy—with the white square on the right of the player chosen by that color. These are the Queen and King, the most important elements in play, the ones that have their traditional courtly roles reversed (part of the same asymmetry). Normally, in real life, it is the Queen that holds the social positioning of the realm, whereas the King “kicks ass” and is the commander-in-chief. But in chess these roles are reversed; the queen is the most effective warrior of all, and the King, by position, holds the value of the game. A strange attractor, called “The Two Dolls,” for in fact it energizes (e.g. ergatizess the game and makes it much more than a mere game, turns it into a metaphor of royal statecraft. In strategy, that is, the two royals are played against each other incounterpoint to the maneuvers of the two “armies.” Basically, it is all about mating. Check it out. Like two Barbie-dolls, each trying to out-Barbie the other. One is tempted to say we have the same contradistinction in kinship, between the so-called genealogical framework that ranks and orders the scheme as a whole, and the interplay of affinal kin, those related (“by alliance”) through explicit gender-interaction. But that is deceptive in that genealogy is as much a function of gendered interaction as affinity, having the same source, and affinity is as much a function of genealogy as it is of marriage. Hence “descent theory” and “alliance theory” are, as the Norse say, “two horns on the head of the same goat,” and it is not a doubly proportioned one. The key to the chiasmus was given by Claude Lévi-Strauss in The elementary structures of kinship. We might well do a little proportion-shifting of our own and call it “The elementary kinship of structures,” so well did it turn the tables and advance a positive or proactive approach to the subject. To complete the atom of kinship, according to Lévi-Strauss, the negative marriage rule configured by Morganian genealogical reckoning must be counterbalanced by some explicitly stated positive counterpart, a known kin strategy for countering the distributional scattering of the generations. Just as there is an incest taboo, so must there be an outcest taboo; marrying in, consolidating one’s lineal gains, is just as important as marrying out, regardless of other considerations. So far so good, for we have a double-proportional counterpart to the rank-and-order versus gendered role-reversal schema found in chess, in that the face-to-face gendered relation of man and wife is counterparted, in Lévi-Strauss’s “atom of kinship” schema, by the “strange attractor”—the “back-to-back” kinship engendered by the man’s sister and the wife’s brother and their respective progeny—the so-called “cross cousins” in the standard kinship repertoire. The explicit proviso of Lévi-Strauss’s “cross-cousin” argument, that “the mother might just as well have been somebody’s father, and the father might just as well have been somebody’s mother,” was made, at different times and in different contexts, by Radcliffe-Brown (1940) and by my Barok congenors in New Ireland, each unaware of the other’s existence. It suggests an imaginal counterplay of purely metaphorical reproduction going on behind the scenes of Morganian kin protocols—like the metaphoric intrigues of the King and Queen in chess vis à vis the straightforward maneuvering of their “armies.” Thus the actual marriage of cross-cousins, however classificatory, is an “easy answer” to, and a quick fix for, the dilemma posed in Lévi-Strauss’s argument, something of an overcommitment to the premise. A kin relation motivated by a strange attractor has no more a certain or predictable structural outcome than a gambit in chess. Both are stochastic, determined as much by their own presence as by the other factors in play. The Daribi, who call their cross cousins hai, say that they are “exactly the same as siblings,” but with an important difference. Since they belong to different wealth-sharing groups, male hai’ must exchange continual payments of wealth to redeem the leviratic claims they share in the inheritance of each others’s wives. I found only one instance of real cross-cousin marriage at Karimui; this was at Hagani, a place where I resided. A man of Sora’ pressed his claims to an unwed Hagani woman. He would hang around outside the longhouse for days crying: “She is my cross-cousin, why can’t I marry her?”“Finally we got so tired of this we just let him have her.”“Aren’t you supposed to beat them both up and give them a stiff lecture about how they are bad people?”“Well, sure, ideally; but by that time our relations with Sora’ had gotten so dicey we just decided to give it up.” Among the “matrilineal” Usen Barak of New Ireland the situation is more complicated; they call this kind of marriage “marriage with the tau (real father’s real sister) or gogup (cross cousin).” Residents of the two northern villages, a subdialect group, put it this way: “The ancestors would never have tolerated anything but strict adherence to the rule of marriage with the tau or gogup; with the erosion of moral values in modern times, however, there is much laxness, particularly among the three southern villages.” Among the three southern villages, another subdialect group and the one where I resided, they countered: “The ancestors would never have tolerated anything so incestuous as marriage with the tau or gogup; now that moral standards have relaxed, however, the people of Belik and Lulubo are free to follow their base desires. This is particularly true of the hamlet of Lulubo called “Giligin,” where everybody marries their tau or gogup.” Since I had some good friends at “Giligin’s Island,” as I called it, I decided to check things out. Fortunately one of them was not only fluent in the English language, but also literate; with his help I collected the complete genealogical record of Giligin to a depth of five generations, and examined each of the marriages carefully. Even making allowance for the so-called “classificatory” or categorial kin reckoning, I could find no instance of marriage - with the tau or gogup in the whole set. When I finished I said to my confrere: “Now I can see that every single marriage at Giligin has been with the tau or gogup” “Yes,” he replied, beaming; “as I told you, we here at Giligin are a strictly moral people.” Speaking of strange attractors, the only instance of direct cross-cousin marriage that I found in the whole Usen area was in my own village of residence, Bakan, in the area that categorically denied the practice, and it involved one of my best friends, the man who lived in the house next to mine. When I asked him to account for it, he said “It was a matter of pure chance, I had nothing to do with it.” Then, straightening himself up to his full height, which was cotsiderable, he explained: “I am known as one of the most moral men in this whole area.” Chess is a game in which there is a single dyad, that of the two players, who take turns in making moves, assuming the roles of one of six optational and function specific pieces, like occupational assignments in a military caste-system. Kinship is not a game, it is life for those caught up in it, and serious work for those who study it. Another big difference is that in kinship, although ideally arranged in dyads, each participant is involved in a great many different relationships at all times. And although that involvement is simultaneous, from the moment of birth and before, the kin participant must learn to differentiate what amounts to a single, diffuse, and all encompassing mode of relating, and adapt their action to the specifics of each culturally determined relationship role. There is no direct analogue to that in chess, which is by contrast digital in the play-mode. Both Radcliffe-Brown and Bateson have pointed out that the adaptation of relating in kinship is limited to three generic modes of analogic imitation, each one of them a variation on the single theme of relating appropriately. There are 1) respect (deferential) relationships, like those of worship, in which the obligations between junior and senior are conspicuously exaggerated; 2) avoidance relationships, in which the pointed avoidance or absence of interaction between the parties constitutes the substance of the relationship itself, and 3) joking relationships, in which the performance of inappropriate behavior, speech, or both offers the alter the option of either acceptance or rejection, and therefore affirmation or denial of the proferred bonding. These three modalities divide the play of kin roles among them, and the necessity of differentiating among them is one of the first things a child learns. Joking, the pretended unseriousness of relating as opposed to the feigned overseriousness of respect or deference, brings us to the vexed question of the overall or long-term purpose or design of kinship. Answering it is no easy objective (questioning it is worse), for it is neither structural nor functional, and it is no surprise to find that it hinges on the very same double-proportional paradox—the strange attractor—that bedevils its entanglement in everyday affairs. Kinship is “connections established among the living on behalf of the dead” and at the same time “connections among the dead made on behalf of the living,” (Wagner 2001: 104) and thus neither “life” nor “death” is going to supply us with any but a dismissive answer. “In a riddle whose answer is ‘chess’,” wrote Jorge Luis Borges, “what is the only prohibited word?” (Borges [1941] 1998: 126) Instead of the solution, in other words, we might as well start guessing at what the riddle itself might be. Nor is the riddle an easy one, though one is reminded of the closing lines of a sonnet written by Edna Saint Vincent Millay (1934) about the ancient Egyptians: Their will was law, their will was not to die.And so they had their way; or nearly so. Still, we get some clues from the work of Richard Huntington among the Bara of Madagascar, and from that of Gregory Bateson among the Iatmul of the Sepik River in Papua New Guinea. Death, for the Bara, is a matter of the crystallization of life’s statuses and relationships—the very fact or matter of death threatens the living with a kind of “deep freeze” peril, a contagious absorption of life’s spontaneity into an ageless matrix of crystalline perfection. (“The perfect,” as the saying goes, “is the enemy of the good.”) Faced with the presence of death, the Bara conventionally do everything possible to reassert the spark of life; they stampede the cattle through the village, create havoc, drink themselves silly on rum, and the nubile teenagers gather in groups in the forest, chant obscene songs to one another, and couple promiscuously. Had Millay written a sonnet about the Bara practices, she might have called them “death warmed over.” Nonetheless, what the Bara have to tell us about kin relationships has little to do with humor, seriousness, or abject avoidance; faced with the desolation of mortality and what we have learned to call “survivor’s guilt,” they pretend—pretend with (real) violence, hilarity, alcoholic rush, and the sublime ecstasy of (“Oh, my mama should see me now”) illegitimate intercourse (“Burgundy adultery,” as a friend of mine used to call it—a good name for a starship). And that is one of the keys—the “might have been” or “philosophy of as if”—to the riddle we are supposed to be concerned with; there is no kin relationship on earth that is not to some degree a matter of pretending (though in chess it is all dead serious), a subtle art that every child learns at an early age. There is no such thing as non-fictive kinship; when the pretense breaks down, so does the kinship, and we have something of an iron law of kinship: real kinship is not the thing that is going on in any pretext that might otherwise be confused with kin relationships. Nor is death itself the thing one thought it was going to be, and this is the lesson taught in Gregory Bateson’s Naven. Though it is a countereffect of the integral duality of Iatmul life, something that Bateson singled out and called schismogenesis, the Iatmul honorific mortuary rite is a case in point of obviation. This is a highly counterintuitive alternative to the usual sense of completion or consummation of a human life, a positive negation, in which the end result is neither the subject (“life”) nor its antithesis (“death”). The obviate human being (“Tod und Verklarung” “Death and Transfiguration” as Richard Strauss called it) is neither living nor dead. In Bateson’s account, all the symbolic accoutrements of the deceased’s life and achievements are assembled together in the form of a token human figure, an effigy representing the deceased. This figure was set up by members of the initiatory moiety of which the deceased was a member. It was a boast of the greatness of their moiety. And when the figure was completed all the men of both moieties crowded round it. The members of the opposite moiety came forward one by one to claim equivalent feats. One man said: “I have a wound here on my hip, where (the people of) Kararau speared me. I take that spear,” and took the spear set against the figure’s hip. Another said: “I killed so-and-so. I take that spear,” and so on till all the emblems of prowess had been removed (Bateson 1958: 155). The ritual Iatmul mortuary practices not only answer the riddle of life-in death and death-in-life; they obviate it. Obviation is the destiny of symbols, as natural to them as death is to human beings. The word “obviate” means not only “to render obvious that which was heretofore obscure,” but also, according to its dictionary definition, “to anticipate and dispose of.” To anticipate death in life is to dispose of life in death. As to Bateson’s role in this, as the anthropologist, one might well conclude that “the historian tells the story, the literature interprets the story, but the anthropologist obviates the story,” renders it innocuous as though it had never been in the first place. Kinship is obviated not in the way we understand it, but in the way that it understands us. Though “underdetermine” might be a better word than “understand,” as in Millay’s sonnet. This puts a whole new spin on our subject, for we have nothing in our whole epistemological repertoire to suggest that something as inert and abstract as “kinship” might have powers of comprehension at all, nor that anyone might, as Ingmar Bergman has it in the movie The Seventh Seal, “play chess with death” (shakspielen med do den)—though that is just exactly what the Iatmul do in their mortuary rite. The Iatmul are “understood” so well in that rite that they have nothing left to live for, or even die for. More generally, insofar as kinship is concerned, not only are our very thought-patterns embodied in the things we think about (events, circumstances, objects), but they also run the danger of being “understood” either too well or not well enough by our parents—the closest one can come to kinship incarnate—and thus growing up absurd. (A real anthropologist’s offspring always run some danger of growing up absurd—but then, consider the source.) Moreover, it is just precisely this inversion of subject and object, the ergative, or strange attractor, that we have seen to bring both chess and kinship “out of the woodwork” and into the world of lived reality, especially when one considers not simply what constitutes them, but what empowers them. Metaphors talk, to you; they have agency, and minds of their own (though admittedly a tad bit schizophrenic). Chess makes your hands move with a patience and skill that not even a lover would tolerate. As for the chess pieces themselves: “it is not the hand of God that moves us, but the god of hand.” The kinship that “understands” one better than one can understand it and the chess that compels the player into “unnatural” moods of intense concentration are parts of an encompassing double-proportional “feedback loop” that extends far beyond the limits of “bush sociology” and grandmaster tournaments. It is one that involves the meaningful properties of language as well—the ways in which language relates to itself as well as the ways in which the speakers of language relate to one another—and that intertwining is as much a part of our heritage as the double-helical infrastructure of DNA. There has always been an unspoken assumption among students of kinship that their subject is somehow related to the need for human solidarity—families, bonds, groups, and that sort of thing. And although that is admittedly a “functionalist” idea, somehow related to the admittedly “structuralist” notion that metaphor or trope is the font and sole purveyor of meaning in language, the connections necessary to bring these two together extend beyond the limits of disciplinary boundaftes and require “thinking outside of the box.” So we might as well start from scratch. We have never found a human assemblage that did not possess both a spoken language and a mode of relating expressed through kin terminology. Thus we can conclude that both of these are somehow necessary to the existence and composition of the species Homo sapiens, the species, be it noted, that was responsible for the concept of species in the first place. (We are the great classifiers of the world, and we classify ourselves as such.) But meaning is the “wild card,” so to speak—a “box” that must necessarily think outside of itself—re-classify itself beyond the ability to classify what it is doing in the process of doing so. No one can say just exactly what a novel or innovative metaphor—a re-invention of language as it were- -means, until or unless it has become “tired” and has classified itself among the familiar and conventional signifiers that make up the lexical properties of language. That is a very long-winded way of putting it, and it has been said many times before. But the long and short of it is that trope or metaphor is always brought into being by making unlikely and unconventional cross- connections in the lexicon-creating an identity between two unlikely parts of language, just as a novel strategy in chess evolves by making unlikely combinations of pieces and moves (a knight-fork, for instance—one of my favorites, along with, of course Burgundy adultery). To put it succinctly, metaphor is the mating and meaning strategy of language, the necessary way in which language relates to itself. Likewise, kin terminology, the necessary reference-coding of kin relationship (without which it would not know itself for what it pretends to be) is the way in which speakers of language relate to one another. Both are part and parcel of the reproduction of language through its speakers, and the reproduction of the speakers themselves by means of language. One of Gregory Bateson’s best and most famous adages was “one cannot not relate.” That is undeniably true, but unfortunately it leads to the fallacy of assuming the reality of naive or spontaneous “relationships,” a kind of collateral damage left over from the “psychiatry” era of the 1970’s, when one could actually get money from the government for pretending that sort of thing. Even for chess players, the need for face-to-face relationships has been subverted by the Internet. Relating, which means “putting the sides together,” is both basic and essential, and defines the human condition both on and off the chess board. There is a big difference here. Shall we psychoanalyze the knight to find out what it “feels like” to move two squares over and one to the side? I don’t think so. Chess throws that much-abused term, “relationship” into high relief. Seen from the top, the knight’s potential moves describe an octagon, but unfortunately a real knight can only access one of these positions at a time. The bishop’s moves describe an angular lattice, the rook’s a Cartesian coordinate system, but only the board itself describes all of these at once. It is not necessarily the foursquare (actually 8 x 8) two dimensional Aubix cube puzzle it appears to be, for it is equally conceivable on a diagonal format, and can also be visualized as a series of internested knightly octagons. Each player, or “side,” faces a mirror-perspective of their strategic layout, “consults” the chess-relational mirror, with the singular exception of the two gendered pieces, the Queen and King. There are no easy answers to the question of why human beings consult mirrors in any case; perhaps it is nature’s very own form of counterintelligence. For the one you see in the mirror has both sides reversed, as well as front and back, and no one else will ever see you that way. That has distinct advantages in mating of course, in both senses of the term, and all things considered it is the chess view of yourself: “move and mate in two.” (I check myself in the mirror before I go out, and so does my date; and though the date itself may come to nothing, the two of them have a great time together—almost as if they were playing a game.) But who are this mysterious them that just stole our evening from us? Let us continue. The real difference between one “side” and the other, or between the game and reality, comes with the realignment of gender with respect to laterality. Normally in the game of life an individual has only one gender and two sides, a right and a left. In chess, however, each player plays only one side, but has two genders, a King and a Queen, one on the left and the other on the right. In the context of the game, as opposed to real life, the players are not really human beings at all, but are playing the roles of what I have called the Antitwins (Wagner 2001, Ch. 4), a cross-purposed subvariant of the human form that is somehow necessary to our existence. They do all the things we cannot, and we do all the things they cannot (They throw ourdice, we throw theirs, cf. Wagner 2001, Ch. 4), or, in the words of the Mary Tyler Moore theme song, they “can take a nothing date and make it seem worthwhile.” Normally there is no such thing as proof in chess; there are rules, gambits, and plenty of execution. (Imagine Robespierre trying to substitute a piece called “The Guillotine” for the King and Queen.) But the proof of the Antitwins in chess is that the King and Queen have exchanged their normative roles—the Queen has taken on the role normally attributed to the King in real life: making cool moves and working strategy, and the king exercises the right of position and social status—for the king by position holds the value of the game itself. The game of chess, of course, has nothing to do with human matings, powerplays, or domestic arrangements, for even in its own fantasy world it belongs to the top of the political food-chain. Full of the imagery of royal houses and their powerplays, it is all about power, and what must be done to safeguard power while controlling the moves of others. Kinship, of course, is nothing like that...or is it? By the time it becomes an anthropological object of study it is already an almost mathematical abstraction, a generator of events that your average kinsperson would scarcely recognize (“Who, me? A cross-cousin? Not on your life. Ahlm a kissin’ cousin; jes ask my Mama, Auntie-Twin”). Even so, like even the most mundane domestic crisis, it trades on the profit margin of control and credibility. Thinking about it, our usual reaction to a generator of events has almost nothing to do with what it is and how it works. The twelfth century Japanese sage Dogen wrote: “What is happening here and now is obstructed by happening itself; it has sprung free from the brains of happening” (Dogen 2008: 7). One morning in 1989 I was busy with what I thought was a huge discovery at the time: that incest is not the tabooed object or form of misbehavior we had thought it was, the opposite of kinship, but rather its perfect appositive. It is the formless content of all kin relations as against the contentless form of the way they have been described and studied. Transported, I began to sketch out the first group portrait of the Antitwins, which I labeled “the twincest,” “the icon of incest,” and “the mirror-gender symmetries.” But time really flies when you’re having fun, and I realized I had forgotten to check my mail that morning. When I did I found a draft copy of Jadran Mimica’s “The incest passions” (Mimica 1991a; 1991b), the best and by far the most articulate study yet conceived on the subject of incest as a sui generis phenomenality. But that discovery, as Mimica was well aware, is beside the point that the actual, regular, and even compulsive practice of downright incestuous relations in all modern societies, within the closest familial relationships, exceeds all reasonable expectations. To understand what this means and why it continues, especially among highly educated people in modern industrial societies, one would need not only an incest taboo but an outcest taboo as well. And if that outcest taboo worked just as poorly as the (barely understood) incest taboo seems to, then all the rationalizations or irrationalizations made to support it would likewise go for nothing, for thinking is not what family behavior is all about. Control is what it is all about. Kinship, like history, natural process, and the strategy of the machine, presupposes a logic of consequentiality, or cause and effect, in the happening of things. Relationship, however, betokens something altogether different, more like the antilogic of irony, wherein one is given the effect first, as in the opening scenario or “setup” of a joke, and then surprised with the unlikely causality of the punchline. Even to put it this way, though, is something of a singularity, or joke on itself, for jokes and relationships—basically monkey wrenches thrown into the machineries of thought—do not fall into a consistency of thinking things backwards (as though one had discovered the perfect system for nonsystematic thinking), but carry an inherent disqualification in the strategy of their telling or working out. They celebrate the uncanny. It is unlikely from this point of view that relationships, as well as the incest and outcest they depend on, ever had a beginning-one might as well search for the origin of the joke (cf. Wagner 2001: Chap. “The Story of Eve”), Relationships first came into focus later on as “kinship” when the cause-and-effect rationalizations were developed to turn them into lines of descent, genealogies, affinal relations, and so forth, for what good are relatives with nothing to relate? It is not only difficult but wellnigh impossible for an anthropologist to imagine what mating was like among a so-called precultural people without immediately thinking of “marriage,” if only to project a pattern that might conveniently be negated by an anti-term such as nonmarriage. inevitably, this leads the prehistorian, otherwise a sound and sane individual, to project one of those classic “natural man” strategies: “Before human beings as we know them had evolved the people coupled with one another through a primitive system that might be called “nonmarriage,” had inconceivable offspring through a form of what we now know as dissemination, which left everybody feeling empty and rootless. it is only now, on hindsight, that we can trace out their movements in the mud, using flint chips and stone tools that hardly scratch the surface.” Chess, of course, is purely “symbolic” (tell this to a grandmaster, and then duck), but the model of kinship I have been reviewing here is something more than that. It has no necessary origin or point of termination, if only for the fact that its meanings and its relationships, as I have shown, are spun mutually and retroactively out of the same generative matrix, which I have termed a strange attractor or double-proportional chiasmus. This would have to include as well the incest taboo and its outcest variant, the arbitrary “rule because there have to be rules” upon which Lévi-Strauss predicated the whole incest-cum-reciprocity argument of The Elementary structures of Kinship. But we have seen that proscriptions and reciprocities of this type produce both the prohibition and the thing prohibited (the practice of incest as well as its prohibition) out of the same central motif, and with the same generous facility with which the Daribi both affirm and deny the “siblingship” of cross-cousins, and the Barak both inadvertantly practice and proscribe direct cross-cousin marriage with commendable moral ardour and self-contradictory results in each case. of course, when dealing with a strange attractor the traditional exception that proves the rule turns very quickly into the rule that proves the exception, so of course there must be people (“somewhere”) that follow the rule of direct cross-cousin marriage just exactly as Lévi-Strauss had predicted. In 1964 I spent a few days with just such a people, the Yagaria speakers of Lagaiu Village in the eastern highlands of Papua New Guinea. Very thoroughly, in consultation with a group of elders, I managed to elicit their kin terminology, correlate it exhaustively with the genealogical record, and determine that they married according to a very strict regimen of bilateral cross-cousin marriage (they called the category of marriageable lines devo’a. Though I did not stick around long enough to see how the “system” worked in practice, I’ll bet it did, for I found them to be very sharp rationalizers. Where else in science can one find such piquant irrelevancies? our hero Gregory Bateson developed the double-bind theory of schizophrenia out of the double-proportional schismogenesis model he had discovered in his work among the Iatmul. Although it remained in vogue among psychiatrists for only a very short time and soon lost ground to other, more “clinically correct” therapies, it imitated schizophrenic symptoms like nothing else in the world (no one has ever managed to cure schizophrenia, or cross cousin marriage either). Often the best we can do is imitate. It is rumored, for instance, that our solar system (the sun, with its attendant satellitic bodies) developed by gravitic accretion out of a primordial disc shaped nebular cloud. In that case gravity would be the primum mobile, and most of the gravity in the solar system is invested in the sun itself. But wait a minute, there is another proportion to this schismogenesis, for most of the angular momentum (gravity’s necessary counterforce) in the system is invested in the planets, satellites, asteroids, and even the tenuous Oort cloud. Hence another “origin” for the whole is a distinct possibility, which is that the sun once had a companion star located in the orbital vicinity of Jupiter, one whose explosion redistributed the system’s angular momentum into the pattern we find today. Since neither hypothesis precludes the other, the question of which is the “correct” one is as trivial as that of how exactly the cross-cousin relation ought to be formulated, and as inconclusive, both being predicated on a strange attractor. The problem with thinking things this way, and of the strange attractor, if I may make so bold, is one of self-absorption and acute self-involvement; it is just simply that a system formulated in this way lacks the ability to step outside of itself and see itself for what it is. Bibliography Bateson, Gregory. (1936) 1958. Naven: A survey of the roblems suggested by a composite picture of the culture of a New Guinea tribe drawn from three points of view. Second revised edition. Berkley: University of Stanford Press. Bergman, Ingmar. 1957. The seventh seal. Stockholm: AB Svensk Filmindustri. Borges, Jorge Luis (1941) 1998. “The garden of forking paths.” In Collected fictions, 119–128. New York: Penguin Books. Crook, Tony. 1997. Growing knowledge: Exploring knowledge practices in Bolivip, Papua New Guinea. PhD Thesis. University of Cambridge. Dogen, Zenji (trans.) 2008. “Gudo Wafu Nishijima and Chodo Cross.” In Master Dogen’s Shobogenzo, Volume IV. Berkley: Numata Center for Buddhist Translation and Research. Gleick, James. 1987. Chaos: the making of a new science. New York: Viking. Jorgensen, Dan. 1981. Taro and arrows: Order, entropy and religion among the Telefolmin. PhD dissertation, University of British Columbia. Lévi-Strauss, Claude (1949) 1969. The elementary structures of kinship. London: Eyre and Spottiswode. Mimica, Jadran. 1991a. “The incest passions : An outline of the logic of Iqwaye social organization. Part 1.” Oceania 62 (1): 34–58. ———. 1991b. “The incest passions: An outline of the logic of the Iqwaye social organization. Part 2.” Oceania 62 (2): 81–113. Morgan, Lewis Henry. (1871) 1997. Systems of consanguinty and affinity of the human family. Lincoln: University of Nebraska Press. Radcliffe-Brown, Alfred. 1940. “On joking relationships.” Africa: Journal of the International African Institute. 13 (3): 195–210 Saint Vincent Millay, Edna. 1934 “Epitaph for the race of Man.” In Wine from these grapes. New York: Harper & Brothers. Wagner, Roy. 2001. An anthropology of the subject. Berkley: University of California Press. Ward, David S. 1973. The sting. New York: Universal Pictures. Les échecs de la parenté et la parenté des échecs Résumé : La vraie comparaison entre une étude anthropologique sur la parenté et le jeu d’échec n’apparait pas à première vue en en restant à leurs propriétés formelles. Elle ne fait sens que lorsqu’on les regarde en tant que stratégies, ou modèles d’événements inscrit dans une temporalité. La simple « proportion » que les deux ont en commun est une forme de comparaison croisée entre des variables dichotomiques que l’on désigne sous le terme de chiasme, illustrée pour la parenté par la relation classique entre cousins croisés, et pour les échecs par la double proportion asymétrique entre le roi et la reine — les deux seules pièces du plateau qui ont un genre — ainsi que par les déplacements et figures des autres pièces du jeu. La différence pourrait être récapitulée dans le mot « coupler ». Les échecs pourraient être ainsi décrits comme la « parenté » de la parenté. L’incapacité à saisir ce chiasme, ou l’essence de double-proportionnalité des deux pourrait être à l’origine de nombre de modèles dysfonctionnels de mariages entre cousins croisés, comme à de nombreuses parties d’échecs très vite terminées. Roy Wagner was born in Cleveland, Ohio, and received his A.B. degree in medieval history at Harvard, and his M.A. and Ph.D. degrees in anthropology at the University of Chicago. His theoretical interests include kinship, indigenous conceptual models, mathematical anthropology, and shamanism; his areal interest include Melanesia and Australia, and more lately, North and South America. he has taught at Southern Illinois University, Northwestern University, and the University of Virginia. He is the author of a comprehensive theory of myth and symbolization called “obviation” in Lethal Speech (1978), and a book called The Invention of Culture (1981), which is a great time-saver in that it may be read by title only. ___________________ * “The Chess of Kinship” is chapter taken from a book length manuscript entitled The place of invention. The chapter opened a two day seminar in Florianopolis, Brazil at the Universidade Federal de Sanata Catarina, entitled “Coyote Anthropology” (Antropologgia de raposa, pensando com Roy Wagner) organized by Jose Antonio Kelly, August 10-11 2011. 1. The conceptual device described in James Gleick’s Chaos: the making of a new science (1987) as a strange attractor is an allegedly causitive function that is neither predictable nor unpredictable, and neither random or ordered, but highly sensitive to the event of its happening at a particular time and place. A good example of what happens when ordinary scientific fantasies turn into fractal (mathematical) realities. 2. The ontological premise (e.g. thought experiment) of “The Two Dolls” may be regarded as the final spinoff-product of the Mother House complex at Telefolip, Papua New Guinea. It is based on a paradox involve the original of something (the Creatress Afek, who created both nature and culture at the beginning of things), and the perfect copy of that original that may not be distinguished from it in any way Afek’s nemesis Boben, who later performed all the miracles of Afek and made all of her journeys but “only to claim credit for what Afek had done.” The point of the exercise is to force one to conclude that if there is no difference between the two, no difference may be known to exist.

Gabail: Performing opacity This work is licensed under the Creative Commons | © Laurent Gabail. Attribution-NonCommercial-NoDerivs 3.0 Unported. ISSN 2049-1115 (Online) Performing opacity: Initiation and ritual interactions across the ages among the Bassari of Guinea Laurent Gabail, Université Paris Ouest La Défense The article aims to show how ritual efficacy may rely on an intentionally produced, constitutive uncertainty. Drawing on the Bassari male initiation ritual, I argue that the central feature of initiation is not so much the achievement of an ontological transformation, than it is the creation of an unsettling context in which the uninitiated cannot but recognize the complex character of the initiates’ identity. I describe two different devices whereby this is accomplished. The first relies on the donning of masks which enable initiated men to exhibit themselves in an opaque body associated with a highly predictable agency. The second is their presence as ritual tricksters characterized by an unproblematic body and disturbingly erratic behavior. In both cases, ordinary interactive scripts are perturbed. Initiation thus provides initiates with the means of mastering techniques that allow them to appear simultaneously as themselves and as not themselves. In doing so, they convey to the uninitiated the very idea of transformation itself. Keywords: Republic of Guinea, Bassari, initiation, ritual efficacy, masks, interactions Almost all ethnographers who have worked in the Upper Guinea Coast region of West Africa are indebted to Simmel’s seminal work on secrecy (Simmel 1950). The German sociologist was the first to have grasped the interactive dimension of secrecy—“a form of reciprocal action”—as pertaining to a general sociological scheme, independent from the variability of its content. It is this idea of the secret as a mode of relationship rather than a content, and not the use Simmel made of Africanist material, that researchers of the region met with enthusiasm.1 Accordingly, his theory on secrecy, while not having been profoundly revised, has been subjected to adjustments made possible through more reliable empirical data. Indeed, his understanding of African secret societies involving masked, theoretically anonymous figures, seems today somewhat outdated. As he saw it, because such masks were the exclusive property of a secret male group, their main function was to implement a radicalization of the differences between men and women by dissimulating from the latter the hidden truth that these “horrible apparitions were not ghosts but their own husbands” (Simmel 1950: 364). If a woman inadvertently pierced the men’s secret, Simmel concluded, secrecy was broken and the secret society would be transformed into a “harmless mummery” (ibid.). Simmel, somewhat contradicting himself in considering the content of the secret essential, exploited the literature at hand but in doing so went awry. As evinced by numerous examples,2 women are invariably aware of the human origin of the masked figures. This knowledge, however, does not jeopardize the effectiveness of the initiation ritual during which the “secret of the masks” is revealed, nor does it affect the actual existence of the secret group, as Simmel suggested. The same applies for the initiation practices of the Bassari I will discuss.3 Not only are women aware of the human presence behind the masks animated by the initiates, but in most cases they know exactly who the initiates are. Far from being a rare occurrence or the result of a failed hoax, the acknowledgement of a human presence in the initiatory simulations is essential to the effectiveness of Bassari masked performances. I argue that the ambivalence surrounding masked performances offers the means of proving the initiation’s effectiveness to give rise to complex identities. In line with the general theme of this volume, this paper shows that uncertainty might be intentionally produced—that it can be a ritual technique carefully designed for specific purposes. Indeed, the key feature of initiation is not so much of achieving an ontological transformation; rather it is to create a disturbing context from which the noninitiated cannot but recognize initiates’ identity in a complex way—that is, neither as an ordinary human nor as a true spirit. Although it is an uchronic speculation, I would be tempted to contend that if Simmel had had Bassari ethnographic material at hand rather than that of the Poro societies, he may not have been misled by the discourse of initiated men. He would have recognized the masks for what they are, that is, a ritual technology intended to produce “snares of thought”4 (Smith 1979, 1984) whose simulated nature, which fools nobody, is a constitutive aspect. By some cruel irony, the scenario envisaged by Simmel on women discovering the human presence behind the masks was, as it were, tested historically in Guinea. From the mid-1960s, the Sékou Touré regime originated a vast iconoclastic movement that affected all regions of Guinea. This “demystification” campaign aimed at reforming mentalities to create “new men” in accordance with the Maoist ideology that inspired the charismatic leader. One of its chosen targets was, of course, the initiation practices (Rivière 1969). Among the brutalities pursued by the party’s militia, one consisted in publicly unmasking initiated men in front of women and children. The intention was to quash initiatory secrets by revealing the human identities behind the masks. Following this public unmasking, the ritual artifacts were generally destroyed. At first sight, this iconoclast policy seems to have been a success since the repressive measures of the new State led, in most cases, to the complete disappearance of the institutions responsible for the transmission of these rituals. In other cases, the initiations resumed toward the end of the regime in 1984, after a more or less heeded prohibition. The Baga societies of the coastal Guinea illustrate the first situation. For them, the devastating effect of Sekou Touré’s iconoclastic agenda was even more effective as it achieved what successive waves of Peul Islamization had begun in this region from the early twentieth century. Yet, the dissolution of Baga initiations did not induce all the expected results. Hence, it did not cause a cessation of the distinction between elders and youths, which remains built on the model of an initiatory relationship despite the absence of a ritual devised to institute it (Berliner 2005, 2008), and even contributed to a certain extent to the present-day religious dynamism of the region (Sarró 2009). Certain societies of the Upper-Guinea forest region, in contrast, fall into the second category. Recent ethnographical studies conducted in this area bear witness to the importance of the strategies of resistance deployed by the populations to understand their singularity within the contemporary Guinean State (McGovern 2004); most of all, they demonstrate the astonishing robustness of the ideas conveyed by these rituals in spite of their prohibition (Højbjerg 2002, 2007). Without minimizing the impact of iconoclast policies in the North of the country where the Bassari live, my ethnographic data do not attest to an experience as tragic as that met with in other regions during this dark period. Certainly, the party’s militia were involved in all sorts of exactions5 and attempted relentlessly to prevent the organization of initiation ceremonies. Masked men were captured, unmasked before women, and beaten; drinking millet beer was prohibited. Nevertheless, these authoritarian measures failed, and it was only at the peak of the “Cultural Revolution,” toward the late 1960s, when some men were compelled to flee over the Senegalese border to be initiated. With this exception, and at the cost of strategies of dissimulation that inevitably complicated the ritual’s organization, all men were initiated in Guinea without interruption.6 Without really discussing the manifold reasons that may explain the stability of the Bassari’s initiatory institution, I provide arguments that partly account for the failure of the iconoclastic movement. As shall become apparent, unmasking men before women was in itself hardly sufficient to bring about the dismantling of the initiation ritual.7 Actually, when the iconoclast militia took away the masks to reveal their human identity to the women, they had little chance of destroying the foundations of the initiatory institution: their attempts were basically just a clumsy and brutal way to test that which the ritual had already anticipated. Unmasking people, and beating them up with sticks is already what the initiatory ritual was about in the first place. Bassari initiation differs little from the general pattern underlying any number of male initiations elsewhere in the world: having reached age fifteen, young boys are separated from their families, symbolically killed and reborn through the intervention of an initiatory entity whose material existence depends on a simulation by the initiators. This initiatory entity, called endaw, often exists solely as an acoustic manifestation (a powerful, guttural sound uttered by an initiate), or when it is made visible to the noninitiated, as a formless, leafy mass hiding the men that animate it. Endaw is thought of as a chameleon whose sculpted representation is often placed on the rooftop of the men’s house. Although Bassari initiation is a classical case of compulsory, collective initiation giving access to adulthood status,8 it is made more complex by the fact that there is more than one adult age. The young boys’ rite of passage through the initiatory institution marks the entrance to the first grade of a system that divides adulthood into a series of distinct stages, each one lasting six years. Depending on the village, these age sets pass through three to seven age grades. The minimal number of three is not accidental. It corresponds to the age sets belonging to the first three grades that actively participate in the initiation ritual by providing three types of initiators, each having a different ceremonial role. The criterion that distinguishes between these three classes of initiators is related to collective violence they may, or may not, use against the novices. The primary consequence of the multiplication of ages is thus a multiplication of the initiators’ positions. The second consequence, which results directly from the first, is a diversification of the interactive patterns that define the initiates’ public behavior toward the uninitiated. Far from producing ordinary men, initiation allegedly gives birth to men whose identity is profoundly changed for having undergone the ritual. However, attesting to the reality of this transformation raises the problem of making its effects tangible to those who, precisely because they are not initiated (young boys) or not initiable (women), are kept away from the secret workings intended to guarantee its effectiveness. There are, of course, a number of ways of demonstrating the transformation wrought by initiation to the uninitiated, such as by marking the body durably through tattooing or scarification, by giving the initiates a new name, and so forth. Bassari initiation leaves no durable trace on the initiate’s body. The beatings the novices endure in the initiation camp leave transient marks that soon disappear. Circumcision, performed some years before initiation, is barely ritualized and does not entail any particular change of status. The only manifest aspect of the definitive property of initiatory transformation would be the initiate’s change of name. As children, young boys are given a first name that indicates their birth order among siblings born of the same mother: all the firstborn sons of a woman are called tiara, secondborn tama, and so forth. Having being initiated, young men bear a new name they usually choose themselves, which, unlike the generic childhood name, is strictly individual because of the extremely unlikely possibility of homonymy.9 Once initiated, men only respond to this name and it is forbidden to call them by their childhood name or to even address them by a kinship term. This rule, however, does not last; after some months, the parents or the neighbors who had known the initiate as a child frequently revert to either a term of kinship or to the childhood name. In fact, if corporal markings or the acquisition of a new name, however significant, are not decisive in representing the initiatory transformation, it is because the break brought about by the ritual does not consist in the initiates being instilled with an essence that differentiates them from the uninitiated in a permanent and durable way. Rather, it consists in the initiates’ ability to incarnate periodically and momentarily the two principal categories of initiatory entities: masks on the one hand, and the koré on the other. The masks appear on various occasions; they are men whose identity is hidden under a costume of leaves and vegetable fibers. The koré are men with bared faces who assume the identity and kinship relations of another man. They behave contrarily to normal etiquette since they speak an obscure language, exhibit a repugnant gluttony, and act like bogeymen toward children. Men tell the uninitiated that the masks are biyil spirits, a generic term referring to invisible entities from bush spirits to ghosts. The koré, in contrast, are presented as humans whose strange behavior derives from the fact that they are temporarily inhabited by the eponymous initiatory principle, the koré. The French-speaking Bassari call these men with altered personalities “odd fellows” (types contraires) or sometimes “demented opposites” (fous contraires). The presence of uninitiated persons conditions the very existence of these two types of initiatory incarnations: when initiated men are among themselves without women or children, neither masks nor koré are present. The reason for these strange apparitions is to stir trouble during interaction in order to capture the attention of the uninitiated by suggesting the existence of a world that only initiates can find intelligible. Like the initiators’ different roles, masks and koré are present in the first three age grades. This theoretically implies six different types of situation. We shall see that the appearance and behavior that these simulated entities adopt in public before the uninitiated confirm the distinction made between initiators during the secret phases of the ritual according to whether they use violence or not. First, I discuss the male initiation cycle by briefly describing the main sequences of initiation from the novices’ point of view. Then, I examine how the men, once initiated, participate in their juniors’ subsequent initiations differently according to the age position they occupy. Finally, I describe the different public behavior, both of the mask and the koré, which the initiators’ different positions induce. If for each age grade, masks and koré enact and display complex identities that evoke the very idea of initiatory transformation, they differ in how they do so. More broadly, the issue underlying this article concerns the effectiveness of the initiation ritual and my intention is to provide descriptive elements indicating how the Bassari respond to a classic paradox of initiation already broached by Michael Houseman (1984): on the one hand, from an empirical standpoint, the newly initiated men are well and truly the same persons as the novices they had been before the ritual; on the other, the ritual is held to transform the initiates in such a radical way that they can no longer be identified with the novices they once were. The sons of the chameleon Initiatory proceedings alternate between public and secret phases during which the novices undergo physical and moral harassments. The ritual ceremony is mainly organized around two violent sequences that open and close a revelatory episode. The first is a semipublic fight between the novices and aggressive masks. In a location near the village, from a distance where women and children can only glimpse what is happening, each novice braves two masks in turn, spurred on by his initiatory godfather.10 The novices can hit the masks with a wooden saber and defend themselves by using a bow as a shield, whereas the masks have whips made of flexible wood in one hand to wound the novice and carry a long stick for protection in the other hand. The fight stops when one of the opponents touches the ground or if the violent blows reach their target. The candidates must demonstrate their courage by forcing the mask to draw back. Although the script is somewhat settled because the mask has instructions to retreat, the outcome of the fight remains unpredictable: the mask may fall over prompting laughs and applause from the male spectators who side with the novices in an ostentatious manner, or on the contrary, the mask may show its skill in outsmarting the attacks of a blundering novice and punish him with several whiplashes, drawing blood. Having fought against the masks, the novices triumphantly reenter the village where the noninitiates celebrate their courage. That same evening, they are taken some kilometers away from the village, to a camp that the noninitiated are strictly forbidden to approach. There, in a theatrical manner, is “revealed” to the novices what they already know but is now officially sanctioned: the masked men attack them for the last time and drop their masks at the very last moment, thus revealing their human origin. The novices are then led to an altar that the elders say is the tomb of a novice who died because he had spoken to women. There, they vow never to reveal the men’s secrets while ingesting a “special” beer made of bits and pieces of dried, ground chameleon. Apart from the masks’ humanity, the secret of initiation thus resides in the new meaning taken on by the connection between the chameleon and the initiatory entity endaw the initiatory chameleon is not the greedy entity told about to children, but a substance that the novices ingest. The consummation of initiatory beer, replete with highly evocative symbolism (the choice of a chameleon, an animal characterized by its ability to transform itself, seems hardly fortuitous), is not accompanied by an explanation of any sort, apart from the warning to keep it an absolute secret and with instructions about the way the novices should now behave: they are forbidden to speak to women and to acknowledge their parents in addressing them by their name. They must also be aggressive and attack the noninitiates, particularly young boys, with sticks and stones. The second violent sequence now begins. Over a few days, the novices spend their daylight hours in caves nearby the initiation camp where the initiators make them line up and beat them periodically with their bare hands or with the bark of trees fashioned into whips. Each novice receives exactly the same number of strokes as all of his peers. The purpose of these beatings is quite clear: it is to annihilate any kind of resistance in order to impose on them a state of undeniable passivity and of collective nondifferentiation. Men will recollect their memories of this event by saying: “You’re nothing. You don’t even have a name.” Each evening the novices are taken secretly to the village to spend the night at the men’s house. A large fire is lit, which the initiators stoke with branches of dried millet in order to create billows of suffocating smoke. This “limbo of statuslessness” (Turner 1977: 97) made of daily beatings and suffocating nights can last between two to three days. Houseman’s distinction (1993, 2002) between two types of secrecy (“avowed” and “concealed”), which he associates with two modes of inflicted violence (“heroic” and “persecutory” respectively), casts light on the difference between these two violent sequences. The first one, that is, the struggle with the mask, is clearly a case of avowed secrecy: the women are kept away from the actual scene of the fight while being informed of their exclusion.11 It is associated with a mode of heroic violence in which the aggression that the novice suffers and inflicts is a prestigious one. The second sequence, that of harassments in the initiation camp, consists in a case of persecutory violence taking place within the context of a concealed secret: the novices are submitted to unspeakable torment of which in theory the uninitiated should not even imagine the existence. At the end of their seclusion, the new initiates are publicly presented with the new name they have chosen for themselves. Being amnesic, they have to relearn the name of each woman and each child among which are evidently their parents and neighbors whom they know all too well. The ritual ends at this moment. Nevertheless, initiation is considered truly complete when nearly a year is up, at the end of a period that the Bassari designate by the specific term atyuwin—which is the conjugated form of the verb “to defecate,” and could be translated by “let them suffer the crap out of them.” In the months following the ritual, the young initiates remain under the iron rule of their elders who catch them out for each offense, whether real or invented, to inflict new punishments upon them. In the eyes of women and children, the new initiates seem endowed with a new identity that is a negation not only of their original condition but also of what the ritual is expected to produce. Mute and aggressive, they represent the opposite of normal sociability: they are neither the young boys they were before the ritual, nor the responsible adults they are supposed to have become. As the months slip by, their aggressiveness diminishes and they progressively recuperate the attributes of humanity the ritual had deprived them of. After a few weeks they begin to speak again, after a few months they relearn agricultural work; finally, they acknowledge their parents and call them by name. The mask-making is the only truly original knowledge initiation transmits to the initiates, the apprenticeship of which takes up most of their time in the bush. The end of this long, liminal period concurs with the time when they accede to the first age grade and are thereby considered to have become a true age set. Each age set is formed of several initiatory promotions that are incorporated to make up a single group whose progress through the age grade system does not depend upon initiation itself. All of the age sets thus shift together by one grade or degree at the end of a six-year period regardless of when initiations take place.12 Each age set climbs to reach the position that the superior age set had previously occupied. At the end of their initiation, young boys are thus integrated into the first age grade. When the six-year period is up, they automatically pass to the second grade with the other initiates that have become part of their age set. Six years later, they accede to the third grade, and so on, until they leave the system. In the region where I worked, a man leaves the system upon reaching the sixth grade, toward the age of fifty. He is then a-xark, which means “old.” However, ritual responsibilities lessen from the fourth grade onward (when men are getting on to thirty), that is, when they enter that part of the age grade system that is no longer automatically linked to initiation. The name of the fourth grade men, okwətek, which literally signifies “the finished,” illustrates this state quite well. The next grade, opidor, could be translated as “those who slouch” and the sixth, opešbinyan, which a circumlocution would give something like “those who plait mats with their bums on the ground,” but could more simply be called “those who sprawl.” In contrast, the first three grades designated by the terms lug, falug, and odyar cannot be translated. While attributed with no semantic range, their sociological significance, however, is completely transparent. These terms designate the three age sets of young men who live in the communal houses of the village center with girls belonging to corresponding age grades and who participate directly in the organization of the initiation ritual. In such an organization, elders only supervise from a distance an initiation that is essentially performed by youths, for youths. Although elders may be present at certain stages of the ritual, they are quick to hold the lower age grades responsible in case something goes wrong or things get out of hand. The ritual role of the first age grade is relatively limited. It is essentially reduced to helping those of more advanced age grades in the organization of the ceremony. They do play an active part in the “secretion” (Zempléni 1996) of the ritual: they are the ones who whirl the bullroarers at night, and who simulate the initiatory entity endaw by collectively covering themselves with a heap of leaves. Their attitude toward the novices being initiated is as kindly as an initiator’s could be. Their function is above all to help and encourage them: they are the ones who accompany the novices to the communal house every evening after the beatings, supporting them when the novices pretend they cannot walk by themselves. Some of them may also be quite pleased to see the novices bewildered, just as they themselves were some months or some years earlier. What is certain is that these young initiators do not take part in the violent ragging inside the initiation camp because the novices will ultimately become part of their age set. This attitude is coherent with the idea that members of the same age set are supposed to be joined in a relationship of interdependent camaraderie. The members of the second age grade are the ones who exert violence on the novices. The aggressive initiation masks are first and foremost chosen from this group whose members are also those who undertake the brutal bullying in the initiation camp. In a society that credits resistance to pain and scorns inflicting it, the violent role these initiators assume is disparaged. They thus frequently earn the nickname “dogs.” The older men deem this violence a necessary evil that cannot be curbed because the initiators of the second grade are only returning what they themselves had suffered. Men of the third age grade know the ritual best for having participated in it up to ten times. Their responsibility is to make sure everything goes smoothly. They are the ones who act as mediators between those belonging to the youngest age grades and the elders whose instructions they pass. Their position as “initiators of initiators,” that is, those having inflicted violence on those inflicting violence, enable them to restrain the initiators’ excesses of brutality. Reciprocally, the novices see them as those who made suffer those who now make themselves suffer and with whom, once the initiation is over, they enter into a joking relationship. In this way, three different types of initiators supervise the novices: those of the first grade advise them, those of the second grade intimidate and brutalize them, and those of the third grade protect them. Each man experiences initiation first as a novice, then successively in each of these positions, taking on “kind” role twice (first and third grades) and “nasty” role once (second grade). The result is a stable system of relationships where the connections between consecutive age sets are expressed in hierarchical terms, as if one had engendered the other—they answer to the designations father and son—and where the alternate classes are assimilated—they address each other with the single kin-term “grandparents/grandchildren”—and linked by a relationship of mutual connivance. For the moment, I have only mentioned the relationships between men, which are established for the most part within the initiation ritual itself. I will now discuss the external aspect of initiation by describing how these three classes of initiators behave in public toward the uninitiated. The ritual enables the display of complex identities under two different forms—masks and koré—at each age. I shall begin with the masks. Singing masks and initiatory masks The two main masks to be found in Guinea are masks made of leaves falling under two generic types called lukwuta and lenεr. Their costume, their voice, and comportment are quite different from one another, and generally speaking, the properties that distinguish them are stable throughout the entire region with only a few stylistic differences between villages. Within the same village, these masks are absolutely identical and only the physical traits of their wearers allows for any distinction between them. All initiates acquire the ordinary “savoir faire” of making masks. Although some are certainly more skillful than others, the actual technique is not the work of specialists. As artifacts, the masks have no value and only those parts that take time to prepare, such as the hood, are hidden from the uninitiated in the roof of the communal house; the other parts of the mask are usually abandoned on the site of the performance. The lukwuta mask is considered to be the elder brother of the lenεr mask. As such, they may be identified using common human names: tiara the first son, and tama the second. Lukwata, the elder brother, has two avatars. The one is an aggressive figure that women and children may not approach on any account. This is the mask that confronts the novices during the initiation fight and also is chosen for certain exceptional coercive measures (see figure 1).13 Lukwuta also appears in a “pacific” version, with a slightly different costume, whose main activity is to sing and dance with women (see figure 2). Children can easily approach him. The younger brother, lenεr, is a mask without a violent avatar that likewise sings and dances with women (see figure 3). These two types of masks share the annual ceremonial cycle more or less equally without ever meeting up. Lukwuta is usually taken out in the dry season, and exceptionally at certain annual female dances during the rainy season. Figure 1: Aggressive lukwuta masks before their fight with the novices. Figure 2: Women with lukwuta masks. Figure 3: Young initiate and lenεr mask. Lenεr appears with the early rainfalls and disappears soon after harvest. It accompanies the boys and girls of the first age sets during all agricultural work. The festivities called “the farewell” (ofelar) mark its disappearance, which precedes by a few days the departure of young people on their seasonal migrations toward the towns in Senegal and Guinea. Not being human, the masks do not belong to an age set. Yet, the men who animate them do. Hence, it comes as little surprise that the initiates of the first grade are associated with the young lenεr mask, whereas the violent initiators of the second grade animate the aggressive version of lukwuta, and the men of the third grade its pacific avatar. In this way, the initiators’ ritual behavior reciprocates the type of interaction between masks and noninitiates: the initiators who are secretly violent are also violent in public, but hidden beneath a mask. Although these frequently encountered masks are familiar to all, they remain strange entities and their apparition always creates a surprise. Even when expected, their irruption is always eventful. Their composite nature partly explains the arresting effect they have on the uninitiated. For one thing, their appearance is an assemblage of fibers, leaves, animal hair, and bared parts of the human body. Although the man covered in the costume lends an anthropomorphic shape to the entity, the face is fundamentally inhuman, being both anonymous and emotionless.14 Made up of elements associated with both hunting and the bush, but also with the world of humans (some of their ornaments are also those of men who dance with uncovered faces), the mask does not represent any known entity of this world. It is not like an animal, nor like a human. The costume’s hybrid character makes it highly productive symbolically, allowing for potentially rich inferences, but where it is impossible to determine exactly what they refer to. It is a “signifier without [a] univocal signified” (Moisseeff 1994: 30) comparable in this respect to many cult objects. It resembles nothing and represents nothing but itself. This may not be entirely true as men, when talking to women and, above all, to children, refer to the masks as biyil, that is, invisible spirits made visible. Yet, it should be noted that the simple fact of stating that the mask has as its referent an entity whose constituent property is invisibility—the most indeterminate form possible—indicates in itself the absence of a prototype. The biyil exist independently of their simulation by the initiates in the form of masks. Biyil, or rather a-yil in the singular, is a generic term for the invisible entities that share the same territory as the Bassari. Like the masks, the biyil are particularly fond of cool and humid places like wooded groves, streams, caves, or even the bottom of termites’ nests. There is no consensus about what the biyil actually are and opinions vary according to the context and from one individual to the next. They can intervene in everyday human activity as bush spirits who lead a hunter astray, as vindictive ancestors who decimate a descendant’s livestock, or as ghosts encountered under strange circumstances. Nevertheless, such happenings are accidental for, theoretically, humans and biyil are supposed to live in separate worlds and remain invisible to one another. While living in the same space, the biyil’s perspective of the world is inversed: what is mist for humans is the biyil’s bush fire, what is night for humans is their day, millet beer is their water, and so on. These worlds in the negative would normally be impervious to each other, but certain biyil strive to cheat humans and are cheated in turn by some men and women who have the innate capacity of “seeing at night” (awediax). The initiates draw their inspiration from this sphere of symbolic knowledge—where there are important individual differences yet a certain general coherence—in order to make the masks look like radical figures of otherness (they come from a world in the negative), bearing a profoundly indeterminate ontological status (are they ghosts, bush spirits, ancestors, or specters?) and whose visible presence evokes the idea of a mutual trickery. The mask wearer lends his body to animate the artifact and is hidden by it in return. He gives it a distinctive movement that distorts his own and personifies the character he embodies: masks do not run, they jump, they do not walk, they dance, and so forth. The wearer also lends a specific voice: the lenεr mask has a hoarse, throaty voice, whereas the lukwuta mask utters gurgling sounds that the Bassari call “water in the voice.” In both cases, it is a vocal technique that is relatively difficult to control, these modified voices not being obtained by the use of any additional instrument. From a strictly acoustic point of view, this vocal technique produces a sound that, in itself, is a good illustration of the mask’s ambivalent nature, the ethno-musicologists who have studied it having qualified it as diphonic. The masks’ anonymizing technique lies in the elaboration of a complex corporality, both from the point of view of the artifact making up the external envelope and from that of the wearer’s physical prowess. The resulting entity is a somewhat complex body, which is reducible neither to its external envelope nor to the human who animates it. Thus, the identity that emerges from this “recursive interlocking” (Fausto 2011) is named for what it is: each mask wearer has a personal mask name, which is neither the mask’s generic name (lukwuta or lenεr) nor that which he is known by as an unmasked man. The names of masks, like the names of initiated men, are constructed from verbal syntagmes, but using a simpler linguistic form: watəәne (“let’s see”), ɗatəәne (“let’s fall!”), korəәne (“provided we can!”). Somebody recounting the ceremony to people who did not attend can easily name the masks involved without revealing their human name: “What was the ceremony like?”—“A great success. The beer was plentiful and watəәne sang all night.” Masks, in spite of their inordinate temperament, are predictable. Interactions with them are always stipulated and extremely simple, like their psychology. When women are present, the manner of talking and acting with them is strictly regulated. Men only speak to them for simple reasons, with simple words. The masks would never understand irony as they are totally lacking in humor. Like foreigners who are slightly simple-minded and only understand the Bassari language in its most rudimentary form, it is with a simple vocabulary, using a respectful, somewhat paternalistic tone, that they are asked to do simple things: to retire to a hut in order to eat or drink, to go back to the dance area. For women, the interactive possibilities are even more limited, indeed reduced to a single alternative: flee or dance, depending if the mask is an aggressive avatar or not. This severe limitation on the freedom of action also applies to the complementary behavior adopted by the masks themselves: they can strike out or they can dance and sing. The highly constrained nature of the pragmatic framework within which the masks express themselves and the “formalized” actions (cf. Bloch 1974) that this framework allows, seems to leave no room for personal creativity. Yet, in spite of the underlying interactive rigidity, the masks’ performances give rise to a competitive context whereby each mask wearer vies with the others. This is particularly visible in the dance sequences of the first grade initiates. Their dance pace is steady and the men animating the masks need stamina in the course of these trying physical ordeals. Yet, still more, it is their voice that attracts all the attention. What counts is not only the intrinsic quality of the voice, whose volume is fairly moderate, but above all the musicality with which the singer varies the melody; the aesthetic ideal being to swallow words, to blend in resonant material to the point of making the semantic aspect of the intoned chant barely recognizable. This male competition is addressed to and brought about by women, and more particularly the young women of the mask wearers’ age grade and of the grade just above, with whom the men live in the communal houses, and with whom they maintain a relationship of formal friendship and of potential affinity. These women fuel the rivalry making judgments in the course of the dance, and by making the dancers carry on until exhaustion. If the hidden identity of the masked men may fool the younger children, the women always know exactly who they are. Their evaluation of the performances is crucial and partly determines the young men’s reputation and the success of their love life. For them, the masks have nothing strange about them except perhaps their appearance, and are only inhibiting in the way they are expected to behave toward them in public. Indeed, they know them intimately. The relationship between women and masks, established during the years of their youth through connivance and seduction, evolves throughout their lives. Even when women are no longer of an age to be dance-partners, they speak to the masks as they would to former lovers, actually with much more liberty than young women can. The paragon of this relationship is the association of elderly women and masks to protect an infant born after several miscarriages. The only notable rupture in the complicity that unites the masks and women takes place when the violent initiators of the second age grade don aggressive masks. The masks’ main targets are children; they are the ones most affected by the cognitive snare of identity that the masks present. The precocity with which young boys imitate masks is remarkable. In fact, as soon as they begin to walk, they are encouraged to do so by their elder brothers who fashion miniature mask attributes for them, and by the women who, when the child first tries a dance step, answer by clapping their hands in a rhythm characteristic of the masked dance (see figure 4 and figure 5). With the same enthusiasm the elders show for encouraging the young boys, they taunt the little girls who are tempted by the same mimetic game. The aggressive masks are no less fascinating for the young children. Every day, they are referred to as bogeymen that will take the children away if they are naughty. Adults may even enjoy superimposing the two types of masks and stimulate, with apparent benevolence, the children’s imitation of masked dancers, while making a thinly disguised allusion to the brutalities the aggressive masks will inflict on them and of which they are still unaware. Figure 4: Uninitiated boy dressing up a kid as a mask. Figure 5: Uninitiated boy imitating a mask By displaying opaque identities based on the elaboration of a complex body and on a simplified psychology generating highly formalized interactions, the masked men establish a relationship with women that is either expressed as a connivance with a hint of seduction, or becomes one of hostility and avoidance. They likewise trick children by exercising on them a fascination mixed with anxiety. Let us turn to the second strange initiatory figure, the koré, the procedure used to blur their identity and the type of relationship this creates with the uninitiated. To make the women smile and the children cry Unlike the masks, the humanity of the koré is unambiguous. Even from a linguistic perspective, the masks and the koré belong to different nominal classes: the masks fall within the class of borrowed words, whereas the koré belong to the same nominal class as that of humans (Gessain 1971: 159; Ferry 1991: 357). Nothing actually distinguishes them from ordinary men in their appearance, except for a band of white palm fiber tied around their arm or head for the men of the first age grade, and a skirt of red fiber for those of the third grade. The distinction is a bit more complex for the second grade men—I will return to this later—but it is made of the same vegetable fibers also used for the making of masks. The transformation of a man into a koré (or a koré into a man) is instantaneous and does not require a backstage, unlike the masks. It may take place in public, in front of the women and children: all that is needed is to attach the vegetable fiber to become a koré, and to untie it to become human again. Men usually indicate their changed state by emitting a powerful, shrill cry, which they repeat periodically as a reminder of their condition. The state of being a koré is described as a temporary modification of the mind, which the Bassari call ondèn, a term that may be translated in other contexts as temperament or character. The ondèn is located in the head: Ondèn-oŋ gəәr gaf ex (“the mind is in the head”). The koré are humans with an initiatory temperament.15 The capacity of becoming a koré is activated at the end of the year-long liminal period of initiation, when young initiates enter the first age set. They can take on the identity of a koré at any time but it is mainly during the initiation period, at the end of the dry season, when they appear. They also turn up after the harvest for the departure festivities of the lenεr mask and for the collective work organized once a year at the chief’s home. This last event, which is more frequent in Senegal than in Guinea, is most likely related to presence of chiefdoms in these two regions. Yet, their favored context of appearance remains that of initiation, when the village, deserted for the rest of the year, reaches a maximal concentration of people. A convenient way to describe their behavior is to label them a ritual buffoon or ritual clown. Yet, these terms do not really do justice to what is involved. Certainly, the koré behave in an extravagant manner intended to raise laughs, but this is not all. They are also profoundly disturbing apparitions. For children, the koré are without a doubt much more terrifying than masks. In describing them as “odd fellows” (types-contraires), the French-speaking Bassari indicate most appropriately that the koré render perceptible a figure of inversion and do not simply embody a burlesque personage. Their behavior reverses the most elementary of social conventions. Whereas the ordinary etiquette makes it shameful to ask somebody for food, the koré go begging from hut to hut, apparently without any embarrassment. Whereas one should eat moderately and in the presence of persons of the same sex, the koré sit down among women, or in the middle of a walkway on the ground, and greedily guzzle the most unlikely combinations of food gleaned from here and there. They usually keep this repugnant mixture of foodstuffs in the palm of their hands for hours on end, thus adding to their interlocutors’ disgust. In addition to their shameless gluttony, they talk in a barely comprehensible language. This particular idiom, although very close to ordinary language in its syntax, phonology, and grammar, is made almost unintelligible by the combination of two modifiers. First, the consonants are permuted. The linguist Marie-Paule Ferry (1981) has illustrated this with an example taken from the French where langue secrète (“secret language”) becomes ganle técresse (“tecres ganluage”). The second modifier consists of using antonyms as frequently as possible: when the koré say “night,” one should understand “day,” etcetera. By skillfully combining these two principles, the language becomes incomprehensible. It was just too difficult for me to follow, but the elder men assured me that they did not understand everything either, or sometimes did after a lapse of time. The young men of the first age grade excel in these language games and all aspire to utter endless tirades as quickly as possible. When their interlocutor is not perturbed enough to their taste, they may pull faces, get just a little too close for comfort, and stare fixedly. All this is flexible and depends on individual creativity and competence, some being naturally more talented than others. One would think this language is incomprehensible for women, but it is nothing of the sort. In fact, in everyone’s opinion, it is the young girls of the corresponding age grades who understand it best. Although they are not allowed to speak the language, they know it very well for having attended regular practice sessions every evening in the communal houses. This cryptic language, rather than excluding the young girls, offers a means of establishing a relationship with the initiates at the expense of the elderly who have lost its practice or do not understand its present-day version, and the children for whom its comprehension is completely out of reach. Although making use of different procedures, this mode of relationship is almost identical to that which links masks, women, and children, though in the latter case, the complicity with the women and the terrified fascination of the children occurs separately. With the koré, these two relationships are articulated simultaneously, within the same sequence. The typical behavior of the third grade koré illustrates this feature even more saliently. The cruel game they enact, and from which they derive a certain pleasure, consists in approaching a woman who carries a little boy in her arms and, like the younger koré, in engaging in a disturbing interaction with her. Just as the koré captures the woman’s attention and makes her smile, he brutally snatches the child from her arms. This is a set script and women know very well what to make of it. Some anticipate it with bemused sincerity, others with a slightly embarrassed protective compassion. The koré takes a few steps with the terrified child who immediately howls with fright, and after a few seconds, the abductor gives the child back to the woman. She, who could be his mother or grandmother, takes him in her arms and consoles him. To accentuate the effect of horror, some koré walk around with animal pelts, usually some small game killed at hunting, which they stuff crudely and hold in their arms like pets. Sometimes, they only need to point the dead animal in the child’s direction to make him howl. This behavior is absolutely unthinkable for men who are in majority fathers and, moreover, affectionate ones, in a matrilineal society where men leave women the responsibility of authority over children. Yet it is justified by the fact that the koré are not really themselves, but someone else. Each man possesses a koré name belonging to a living alter ego, usually a man of a higher age set. In this way, he does not act in his own name but in that of the alter ego whose identity he has temporarily taken on, acquiring at the same time the latter’s kinfolk as his own. The koré take advantage of this identity substitution by presenting themselves to the wife of their alter ego as a husband would. They then have the liberty, in tacit consent with her, to parody a demanding and dissatisfied husband. In short, the koré’s range of action is open and depends on their personal creativity. The stable characteristics of their behavior are the use of a modified language, the substitution of identity, and the establishment of a dysphoric interactional climate that produces embarrassment and turmoil. Although they employ means different from those of the masks, their objective of tricking the children in complicity with the women is altogether comparable. The young koré of the first grade rival each other in linguistic virtuosity in front of dumbfounded children and young girls who refrain from speaking a language they understand perfectly well, whereas the thirty-year-old koré of the third grade do their utmost to make older women smile in order to make them participate in a mock kidnapping. In many respects, the koré can be seen as antimasks. Whereas the opacity of the mask’s identity relies upon the elaboration of a complex “external” envelope, together with a rudimentary and highly predictable intentionality, that of the koré makes use of its symmetrical opposite: their unproblematic bodies contain unpredictable personalities. The connection between masks and koré is brought into clearer focus when analyzed in light of a final scenario I have not yet described, that which corresponds to the behavior of the koré of the second grade initiates. The dance of odd fellows During initiation, the koré behavior of violent initiators is a dance. Its analysis deserves a separate article, but I shall only examine one aspect relevant for the present argument: the way it contrasts with other dances. One and only one dance is associated with each male age grade; it belongs to the age grade’s corresponding age set whose members perform it each year in the course of the age grade’s six-year cycle. When men change age grade, they yield their dance to the next age set and receive that of the preceding age set. The koré dance belongs to the men of the second age grade. Similar in many respects to the other men’s dances of which I have given a more detailed description elsewhere (Gabail 2011), it is based on the same aesthetic principles of homokinesis and gestural unison (Beaudet 2001): all the dancers execute the same step at the same time according to an arrangement that tends to draw the eye not so much to the dancers’ bodies, but to the swaying movement of their ornaments (see figure 6). Two ornaments are characteristic features of this particular dance. The first is a skirt of black fibers that the dancers wear around their waists; it is this artifact that indicates their condition of koré. The second is a big scraper held in the hand. During the dance, men animate these two specific ornaments in a way so as to produce a visual effect (the fringe of the skirt sways with the cadence of the steps) combined with an auditory effect (the rumbling of the scraper beats out the rhythm). The combination of these two expressive traits is the dance’s signature. Other men’s dances employ the same procedure, but with different ornaments, the movement of a feathered crest associated with the sounds of a metal bell, for example. The first difference that stands out in the koré dance is the absence of women. Men’s dances are always executed together with the members of two female age sets: the one that includes women who are the men’s formal friends and who belong to the same age set as the men, and the other, of a higher age grade whose members are their potential wives.16 Moreover, the ornaments worn by the women in these other men’s dances accentuate the conventional relationships between the male and female dancers: they dance with a stick offered by their formal friend and display on their back a decorated baby carrier that evokes an affinal relationship. In excluding women and their ornaments from their dance, the koré dancers present themselves as a self-centered male group. Although they do not flaunt the negation of women as brutally as the aggressive masks, they certainly show an underlying hostility toward them. Accordingly, women keep aloof from the performance. They neither approach the aggressive masks who fight with the novices, nor the koré who dance without them. The second notable difference is the status of two emblematic ornaments used in this dance. Unlike the plumed crests and the iron bells of the ordinary male dances that are kept at the owner’s home, the fiber skirts and the scrapers, like the masks’ costumes, belong collectively to the age set and are kept out of sight in the roof of the communal house. In other words, these artifacts belong to the same category as the masks themselves. This comparison between the koré’s ornaments and the masks’ costumes thus adds a dimension to the opposition that has already been mentioned: the koré are not only antimasks, they are masks with bared faces. Figure 6: Dance of the initiators of the second grade. The raffia fiber the koré of other age grades tie around their arm or head indicates this status even more subtly. These are not actual parts of the masks’ costumes; they are ties used to affix part of the mask costume to the mask wearer’s body. Literally, they are mask straps. “To attach” (a-xap) is the expression used to designate the act of fixing the costume to the wearer. The same term is used to describe the act of tying on the raffia fibers that precipitates the transformation into a koré.17 These ties, moreover, are attached to the same body part that is used to attach the mask’s costume, so that when men transform into koré, they actually attach an absent costume. This observation connects the general assumption that masks are biyil spirits with the “contrary” behavior of the koré. Indeed, the koré’s behavior reflects the world envisaged from the inverted perspective of the biyil precisely because koré are masks without a costume, that is to say unmasked biyil. Yet, as opposed to real biyil, who remain invisible to humans or only appear to them in unforeseen circumstances, nobody is fooled by the simulated character of the masks and the koré. These are snares of thought and recognized as such. What is beneath the initiator’s mask is neither an ordinary man, nor a true biyil. It is a complex figure of a particular type: an invisible spirit rendered visible through human simulation, in other words, an initiate. When men appear in public, having assumed the form of a mask or koré, they never cease to be themselves. They mobilize techniques that alter their identity without annihilating it. In wearing the mask’s costume, they display the presence of the biyil outwardly; in attaching the raffia ties they evoke the koré’s presence inwardly. Yet, identifying the koré as the unseen aspect of the mask is of little help in defining what a biyil really is, or what a koré is either. In both cases, they are entities whose status remains profoundly indeterminate. The initiates’ complex identities derive precisely from this ambiguity. The invisibility of the true biyils and the blurred identity of the koré index this complexity. The paradox mentioned at the start of this article remains therefore unresolved and is actually largely exploited: initiation provides initiates with the means of mastering a technique enabling them to appear simultaneously as themselves and as not themselves, such that it becomes impossible to decide if that which separates initiates from noninitiates is a difference of nature or not. Indeed, the purpose of the initiation ritual is less to transform people (e.g., turn boys into men), than it is to provide the means whereby initiates are able to convey to the uninitiated the very idea of transformation itself. Therein lies what men can do and boys cannot. There is no more subtle—yet explicit—way of capturing this idea than choosing the chameleon as the ritual’s apical symbol. Conclusion As a consequence, it matters less to know precisely what these entities simulated by the initiates are—I would even argue that it matters not to know exactly what they are—than to know how to interact with them. In that regard, women may be said to know as much as men do, these initiatory entities being tailor-made for them. It is at this level that I am tempted to locate the grounds of the initiation ritual’s distinctive efficacy: the interactions these complex entities bring about and the stable patterns of relationships they establish with the uninitiated. Envisaged from the standpoint of these repeated interactions, initiation largely overflows, both upstream and downstream, the strict period of the ritual’s organization. More than the initiators’ bullying, it is the interaction that the already-initiated men have with the children who are not yet initiated and with the women who never will be, which provides the ritual with the conditions for its effectiveness and its replication. One would hardly exaggerate in considering that when the ritual itself takes place, everything is already more or less played out, the basis of the ritual’s efficiency being established well before the masks’ secret is officially revealed to the novices. Likewise, the ritual’s irreducible features, its violence and its esoteric dimensions, do not really make sense until long afterward, when the novices themselves become initiators. In such a context, initiation can hardly be contained within the temporality of the ceremony’s organization alone. It is a general logic of relationships between genders and ages, which everybody experiences according to their own sex and age, and which leaves its imprint in all sorts of contexts. These can be the highly formalized setting of the ritual itself, but also the more humdrum appearance of the masks and koré during collective festivities, and even the more intrusive daily interactions with young children. Reaching beyond the limits of the initiation ceremonies themselves has allowed me to show that techniques giving rise to dysphoric interactions between the initiates and the uninitiated act to signify the initiatory transformation itself. It is necessary, however, to distinguish between two very different ways of unsettling initiated/uninitiated relations. The one relies on the donning of masks which enable initiated men to exhibit themselves in a complex, opaque body; this leads the uninitiated to question the very nature of the entity they are facing (what am I interacting with?). The other is centered on the enigmatic figure of the koré, which stands in sharp contrast to the masked dancers: under an unproblematic exterior, their faces revealed, these unnerving pranksters act and speak in an obscure, completely unpredictable way, thus prompting the uninitiated to wonder who exactly they are interacting with. Dysphoria, in each case, relies on a different device: whereas the masks are characterized by an excess of predictability, the koré are distinguished by their complete lack thereof. Thus, while interacting with masks follows strict rules and entails an oversimplified communicative common ground, it is downright impossible to adjust to interactions with the koré. One can only be the assenting victim of their constant improvisations, with no hope of finding any common basis of everyday communication. In both cases, ordinary interactive scripts are perturbed, providing young boys and women with a direct experience of the initiatory transformation achieved by men. Acknowledgments This text greatly benefited from the comments of participants in Philippe Descola’s seminar (“Les raisons de la pratique”), where a first version of it was presented. I am also very grateful to the editors of this issue and to the anonymous reviewers for their help in improving this manuscript (not the less for accepting it in French in the first place). Saying that Heather Yaya, Isabel Yaya, and Michael Houseman translated the text could not quite capture what they did to improve it, while Grégory Delaplace has given a welcome hand for the last push: may they all be thanked warmly. References Bellman, Beryl L. 1984. The language of secrecy: Symbols and metaphors in Poro Ritual. New Brunswick, NJ: Rutgers University Press. Berliner, David. 2005. “An ‘impossible’ transmission: Youth religious memories in Guinea-Conakry.” American Ethnologist 32 (4): 576–92. ———. 2008. “Transmettre la peur et la fascination: Mémoires d’une interaction initiatique en pays bulongic (Guinée, Conakry).” Systèmes de pensées en Afrique Noire 18: 105–31. Bloch, Maurice. 1974. “Symbols, song, dance and features of articulation: Is religion an extreme form of traditional authority?” European Journal of Sociology 15: 55–81. Bonhomme, Julien. 2010. “Initiation.” In Dictionnaire des faits religieux, edited by Régine Azria and Danièle Hervieu-Léger, 541–48. Paris: PUF. Boyer, Pascal. 1988. Barricades mystérieuses et pièges à pensée: Introduction à l’analyse des épopées fang. Nanterre: Société d’ethnologie. Delacour, A. 1912. “Les Tenda, Koniagui, Bassari, Badyaranké de la Guinée Française.” Revue d’Ethnographie et de Sociologie: 287–96, 370–81. ———. 1913. “Les Tenda, Koniagui, Bassari, Badyaranké de la Guinée Française.” Revue d’Ethnographie et de Sociologie: 31–52, 105–20, 140–53. Fausto, Carlos. 2011. “Masques et trophées: De la visibilité des êtres invisibles en Amazonie.” In Masques des hommes, visages des dieux, edited by Jean-Pierre Goulard and Dimitri Karadimas, 229–54. Paris: CNRS Éditions. Ferme, Mariane C. 2001. The underneath of things: Violence, history, and the everyday in Sierra Leone. Berkeley: University of California Press. Ferry, Marie-Paule. 1977. “Les noms des hommes et des masques chez les Bassari du Sénégal oriental.” In Langage et cultures africaines, edited by Geneviève Calame-Griaule, 84–99. Paris: Maspéro. ———. 1981. “Les ‘ganles tecresses des Ndèta’ (les langues secrètes des Tenda).” Objets et Mondes 21: 173–76. ———. 1991. Thésaurus Tenda: Dictionnaire ethnolinguistique de langues sénégalo-guinéennes (Bassari, Bedik, Konyagi). Paris: Peeters. Gabail, Laurent. 2011. “Comment dansent les institutions: Classes d’âge et rapports de sexe chez les Bassari de Guinée.” Journal des Africanistes 81(1): 9–33. Gell, Alfred. 1999. The art of anthropology: Essays and diagrams. London: Athlone Press. Gessain, Monique. 1971. “Les classes d’âge chez les Bassari d’Etyolo (Sénégal Oriental).” In Classes et associations d’âge en Afrique de l’Ouest, edited by Denise Paulme, 157–84. Paris: Plon. ———. 2006. La femme et le masque. L’éloge de l’équilibre chez les Bassari. Paris: Sepia. Højbjerg, Christian K. 2002. “Inner iconoclasm: Forms of reflexivity in Loma rituals of sacrifice.” Social anthropology 10 (1): 57–75. ———. 2007. Resisting state iconoclasm among the Loma of Guinea. Durham, NC: Carolina Academic Press. Houseman, Michael. 1984. “Les artifices de la logique initiatique.” Journal des Africanistes 54 (1): 41–65. ———. 1993. “The interactive basis of ritual effectiveness in a male initiation rite.” In Cognitive aspects of religious symbolism, edited by Pascal Boyer, 207–24. Cambridge: Cambridge University Press. ———. 2002. “Dissimulation and simulation as forms of religious reflexivity.” Social anthropology 10 (1): 77–89. McGovern, Michael. 2004. Unmasking the state: Developing modern political subjectivities in twentieth century Guinea. Unpublished PhD thesis. Emory University. Moisseeff, Marika. 1994. “Les objets cultuels aborigènes.” Genèses 17: 8–32. Murphy, William P. 1980. “Secret knowledge as property and power in Kpelle society: Elders versus youth.” Africa 50 (2): 193–207. Rivière, Claude. 1969. “Fétichisme et démystification: L’exemple guinéen.” Afrique-Documents 102-3: 131–68. Sarró, Ramon. 2009. The politics of religious change on the Upper Guinea coast: Iconoclasm done and undone. Edinburgh: Edinburgh University Press. Simmel, Georg. 1950. The sociology of Georg Simmel, edited and translated by Kurt Wolff. Glencoe, IL: Free Press. Smith, Pierre. 1979. “Aspects de l’organisation des rites.” In La fonction symbolique: Essais d’anthropologie, edited by Pierre Smith and Michel Izard, 139–70. Paris: Gallimard. ———. 1984. “Le mystère et ses masques.” L’Homme 24 (3-4): 5–34. Turner, Victor W. 1977. The ritual process: Structure and anti-structure. Ithaca, NY: Cornell University Press. Tuzin, Donald F. 1997. The cassowary’s revenge: The life and death of masculinity in a New Guinea society. Chicago: University of Chicago Press. Zempléni, Andràs. 1993. “L’invisible et le dissimulé: Du statut religieux des entités initiatiques.” Gradhiva 14: 3–14. ———. 1996. “Savoir taire: Du secret et de l’intrusion ethnologique dans la vie des autres.” Gradhiva 20: 23–41. ———. 2000. “Initiation.” In Dictionnaire de l’ethnologie et de l’anthropologie, edited by Pierre Bonte and Michel Izard, 375–77. Paris: Quadrige P.U.F. Les masques des initiateurs : interactions initiatiques au fil des âges chez les Bassari de Guinée Résumé : L’article cherche à montrer comment une incertitude constitutive et intentionnellement produite peut être à l’origine de l’efficacité rituelle. En décrivant le rituel initiatique masculin des Bassari, j’argumente que le propos central de l’initiation consiste moins en une transformation ontologique qu’en la création d’un contexte à partir duquel les non-initiés sont amenés à reconnaître l’identité des initiés d’une manière inédite et complexe. Deux techniques sont décrites pour parvenir à cette fin : la première est celle du port des masques, qui permet aux initiés de se présenter sous la forme d’un corps opaque associé à une intentionnalité simplifiée et prévisible; la seconde est celle du clown rituel, dont le corps non problématique est au contraire couplé à une intentionnalité insaisissable. Dans les deux cas, le script des interactions ordinaires se trouve profondément modifié. L’initiation fournit donc aux initiés la maitrise d’une technique qui leur permet d’apparaître comme étant simultanément eux-mêmes et pas eux-mêmes, imposant ainsi publiquement aux non-initiés l’idée même de transformation. Laurent Gabail is a Fyssen Postdoctoral Fellow at the African Studies Center (Oxford University). In 2012 he completed a PhD in anthropology at Université Paris-Ouest on social morphology, ritual, and dancing among the Bassari of the Republic of Guinea (West Africa). His work focuses on kinship, age-class systems, male initiation rituals, and dancing as a generative process. Since 2005, he is a member of the research team TIP (Traitement Informatique de la Parenté, or Kinship and Computing), which provides analytical tools and methods for the empirical study of kinship networks. ___________________ 1. If we restrict ourselves to the literature of this region, which more or less covers the area of the poro male initiation institution, see notably Bellman (1984) and Murphy (1980) for the Kpelle, Ferme (2001) for the Mende, Højbjerg (2007) for the Loma. In extending to societies that are culturally close but geographically distant, see also Sarró (2009) and Berliner (2005) for the Baga, as well as Zempléni (1993, 1996) for the Senufo Nafara. 2. See notably Zempléni (1996: 36) and Bellman (1984: 144). 3. The Bassari, who call themselves ɓe-liyan (lit. “Those of the laterite”), belong to the Tenda linguistic group that also includes the Bedik, the Coniagui, the Badyaranké, and the Boïn. Traditionally, the Bassari are hunter-gatherers who practiced a complementary agriculture that has become the predominant subsistence activity today. They number about 15,000 and share with the seminomadic Peul pastoral tribes a hilly region on both sides of the border between Senegal and Guinea. Since 2006, I have spent some twenty months in Guinea, staying in a group of villages called Bokore (lit. “Those of Koré,” after the name for male initiation) that comprises one of five predominately endogamous groups with a close linguistics and ritual proximity. Bassari ethnography started in Guinea with Delacour (1912, 1913) and continued under Sékou Touré’s political regime in Senegal (Bane group) and gave rise to numerous publications since the early 1960s, principally by Monique Gessain (1971, 2006) and Marie-Paul Ferry (1991). 4. As a reminder, if the expression “snares of thought” has been widely diffused in anthropological texts (Boyer 1988; Gell 1999; Fausto 2011), it was initially intended to account for the ritual simulations of the Bedik, the Bassari’s nearby neighbors. Reverting to Smith’s argument, I use the term to convey the idea of ritual efficacy as an interactional process rather than the cognitive operation an artifact brings about. 5. In the early 1960s it was still common for men to wear only a penile sheath and to braid their hair as women did. Captured men suffered the following punishment: they were forced to have their heads shaved and were made to eat their penile sheath. 6. According to Bassari men, concealing the secret phases of the ritual from prying outsiders was no problem. What became more problematic was the preparation and consumption of millet beer that remains, still today, the main reason for a gathering of people who live dispersed during the greater part of the year. The women had to prepare the beer in different locations where it remained. The ritual attendants would drink the beer when passing by each of these sites, one after the other. If a government official presented himself, he was shown only one of the storage locations so that only a small portion of the production brewed was destroyed. 7. In this respect, other case studies in Guinea have proven similar. Sarró (2009) has clearly demonstrated that the simple fact of removing the masks in front of women could not explain the end of Baga initiations. As for the Loma, the coercive measures conducted against initiation practices resulted paradoxically in reinforcing the institution (Højbjerg 2007). For historical conditions that may provoke the collapse of a men’s cult, see also Tuzin’s classical study of the “death of the Tambaran” among the Arapesh-Ilahita (Tuzin 1997). 8. In other words, this is a case of what is usually called a “tribal” initiation, as opposed to the elective initiation of religious specialists or the voluntary initiations into secret societies, which are both generally optative and individual. See Zempléni (2000) and Bonhomme (2010) for a review of the limits of these ideal-types. 9. These names are verb phrases that either take the form of an imperative plural (Ketalin: “arrange yourselves in a group”) or a negative passive (Gaxwondemi: “nobody belongs to us anymore”; Gaungwandemi: “nobody has patience with us”). See Ferry (1977). 10. The initiatory godfather is a stand-in father chosen among the classificatory agnatic brothers of the novice’s father, while the masks are always chosen among the elder uterine brothers. I only mention in passing that one of the kinship-related concerns of the ritual is to redefine the novice’s network of relations by positioning agnates as protectors and uterine kin as aggressors which, in a matrilineal society, is an inversion of everyday norms. 11. Moreover, they participate actively through their lamentations heard by the novices from a distance and by showering praises on the candidates when they return. 12. Initiation schedule may depend on demographic or economic factors: e.g., Are there enough boys of an age to be initiated? Has last year’s harvest of millet been sufficient so that the boys’ fathers can produce enough beer? 13. For example, some years ago, the Bassari decided to prohibit the Coniagui women, neighbors to the Bassari, to sell hard liquor. The men charged some aggressive masks to chase them away. 14. The young lenεr mask is more minimalist. He does not wear a hood and hides his face with either a veil or a fly swatter. 15. To alleviate the terminology of this article, I employ a simplified acceptation of the term koré as do the French-speaking Bassari and the ethnographers of the region. In fact, koré corresponds to several related terms. In the Bassari language, koré designates initiation, o-koré the initiatory principle that modifies ondèn, and a-xoré (sing.) or ɓəә-xoré (plur.), the men whose mind has been modified by initiation. 16. Here, the expression “men’s dance” refers to the male quality that the Bassari accord to these performances even though these dances involve performers of both sex. In a symmetrical way, “women’s dances,” which involve women dancing with masked men, are qualified as female. The actual presence of members of both sexes in these dances is systematically neutralized by attributing an order of precedence to one sex or the other, this being often expressed in economic terms (the men or women provide the beer). This is expressed choreographically by the necessary participation of the male or female age set considered to own the dance in question. 17. The verb “to attach” is specific to the idea “to put on a mask.” These terms are different to those used for “to attach” in the sense “to tie” a bundle of sticks together (alan) or “to attach” to tie up an animal (a-yeb).

Gregory: Skinship This work is licensed under the Creative Commons | © Chris Gregory. Atribution-NonCommercial-NoDerivs 3.0 Unported. ISSN 2049-1115 (Online) Skinship Touchability as a virtue in East-Central India Chris Gregory, Australian National University and the University of Manchester In “What kinship is” Marshall Sahlins (2011) provides a new and provocative answer to an old and much-debated question, defining it as the “mutuality in being.” For the Halbi speakers of the Bastar Plateau in East-Central India kinship is defined by touch: juniors greet seniors with tactile gestures of familial respect that are reciprocated by tactile gestures of familial love. On certain ritual occasions these salutes are adorned with colorful flowers, tasty food, purifying water, sweet-smelling incense, nice-sounding words, and heartfelt sentiments. Non-kin, by contrast, are defined by non-tactile gestures of mutual respect. The general implication of this case for the study of kinship as “mutuality of sensible being,” to give Sahlins’ formulation a slight twist, involves a move away from the study of kinship as the abstract semantics of reference terminologies to a consideration of the pragmatics of face-to-face sensible relations between people. Little ethnographic research has been done on the latter; the Japanese word “skinship,” evoking as it does the coming together of touch and kinship, signifies a fresh approach to the analysis of kinship. Keywords: kinship, skinship, touchability, modes of address, senses, value, India The argument Skinship is a Japanese word most Japanese people think is an English word. It is not in the Oxford English Dictionary (OED) but one can understand why some Japanese might think it should be. The word skinship makes good sense to the English ear, evoking as it does the coming together of the sense of touch with the concept of kinship. It suggests a sensible approach to kinship analysis, one that gives primacy to tactile communication in the ordering of family relationship by drawing attention to the role of touchability as a non-verbal moral code in “face-to-face” relations between people. This is precisely the sense in which I will use the term here because I found, as I struggled to make sense of my fieldwork data on kinship usages of the Halbi speakers of Bastar District in Central India, that the conventional wisdom, which gives priority to the semantics of reference terminology, is abstract non-sense in the literal, non-pejorative sense of that term. Skinship is not only a word we need in the OED, it also signifies a much needed fresh approach to the analysis of kinship, one whose scope and limits I will strive to define in this essay. The notion of skinship, I will suggest, enables us to grasp the sensible principles of virtue that define the cultural specificities of an ethnographic case. Morality is the art of living together well and this presupposes some commonsense values that can transcend generational time and enable people to cope with vagaries of life and the joy and pain this necessarily brings. Familial love and respect are key values found everywhere but everywhere differently for they are shaped by deep historical relations of contiguity. In Bastar these values are polyvalent and embedded in extremely complex forms of verbal and non-verbal modes of address that are culturally specific to this region of India. My essay is divided into three parts. In the first part I briefly examine the question “What is skinship?” by looking at the Japanese origin of the word “skinship” in a cross-cultural context. In the second section I strive to develop the idea of skinship as a theoretical concept for use in ethnographic research. I do this by locating the idea in the context of existing theories of face-to-face sensible communication. In the third section I examine the implication of this concept of skinship for the study of kinship with illustrative data drawn from my own fieldwork in India. What is skinship? Skinship in Japan The Japanese sense of the word “skinship” is our starting point because it raises fundamental questions about the relationship between tactile communication and cultural difference. Wikipedia notes that it was initially used to describe “the closeness between a mother and her child due to the physical contact of their skin” while today “the word is generally used for bonding through physical contact, such as holding hands, hugging, or parents washing their child at a bath” (Wikipedia 2007). Another online source, Wordspy (2007), gives a somewhat broader definition that includes other senses. It defines skinship as “feelings of relatedness and affection between two people, particularly a mother and a child, caused by hugging, touching, and other forms of physical contact.” This is consistent with the only direct evidence I have on how the term is used in Japan. It comes from the seven-part video series called Childhood which examines child raising practices in five continents. In one sequence in the second part a Japanese mother called Mitiko argues the case for breastfeeding in the following terms: I was convinced mother’s milk gives strong immunities that will help the child grow strong. Then there is the “skinship” (sukinshippu) between mother and child, eye gazing into eye. Besides I think human milk is best for a human child. (Haines-Stiles and Montagnon 1991: II, 44 mins). The basic sense of the word skinship, then, comes from touch but it necessarily involves the other senses because it is impossible not to see, hear and smell people when you are close enough to touch them; the only sense not excited is taste. Different cultures handle this fact in different ways, but in Central India, as we shall see, familial touch is highly elaborated and is deliberately concerned to excite every sense, including taste. A notable feature of Japanese skinship is that the extremely close relationship between mother and young child is radically disrupted around the age of five or six when all tactile communication ceases. After this kinship becomes an “untouchable” relationship; henceforth familial love and respect is expressed vocally through the use of special terms of address, visually through bodily postures, bowing, seating arrangements, gustatorily through gifts of sweets and food, and olfactorily through gifts of perfumes and the like. So highly developed is the art non-tactile sensible communication in Japan that the casual observer of the culture, as I was for three months in 2004, gets the impression that intense feelings of familial love expressed in these non-tactile ways are a compensation for the proscriptions on skinship that begin around age five. Skinship in English-speaking countries Skinship in English-speaking countries could not be more different, a topic Montagu (1971) addresses in his classic work, Touching: the human significance of the skin. Skin, Montagu (1971: 1), notes, is “the mother of the senses,” our “first medium of communication.” This is first in the order of sensory development; the distant senses—sight and hearing—attain their full development later than the proximate senses of touch, taste and smell. The three most important senses of Homo sapiens develop in a definite sequence: (1) tactile, (2) auditory, and (3) visual. The order of precedence is reversed as the child approaches adolescence: (1) visual, (2) auditory, and (3) tactile. As soon as one has developed the know-how of being human, Montagu notes (1971: 236), “vision becomes by far the most important of the senses.” This biological ordering of the senses is everywhere given a cultural twist. Western civilization, Montagu notes, places comparatively more taboos on the proximate senses. “One of the great negative achievements of Christianity,” he argues (1971: 237), “has been to make a sin of tactile pleasure.” One of the consequences of this, he notes, is that babies spend a good part of their life alone, a marked contrast to Japan where an anthropological study found that co-sleeping was common and that sleeping alone “is a reluctant alternative most commonly occurring in the years between puberty and marriage” (1971: 249). The standard work on childcare in the US a century ago, Synnott (2005: 41) notes, was L.E. Holt’s The care and feeding of children (1929) which advised parents never to play with children under six months, and “the less of it at any time the better.” It also advised that infants “should be kissed, if at all, upon the cheek or forehead, but the less of this even the better.” J. B. Watson, the founder of Behaviorism, gave similar advice in his Psychological care of infant and child: “Never hug and kiss them, never let them sit on your lap. If you must, kiss them once on the forehead when they say goodnight. Shake hands with them in the morning” (cited in Synnott 2005: 41). Synnott also notes that the work of anthropologists such as Margaret Mead played an important role in challenging some of these dogmas. Today many of them have been turned upside down. Walls in maternity wards today are decorated with posters proclaiming the benefits of skin-to-skin contract such as placing the baby on the mother’s body after birth, co-sleeping and the like. Skinship in Bastar District Skinship among the Halbi speakers of Bastar District is very much like that in Japan when it comes to very young children but thereafter differs dramatically. Touchability is the basic defining characteristic of all familial relations in Bastar District: kin of all ages in Bastar are defined by those whom you touch. The untouchability for which India is infamous refers to inter-caste relations. Touch can defile and pollute the person touched and while this value can provide some insights into untouchability in India it is not the full story; of equal, if not more, importance are the positive values non-touch communicates. Respect is the basis of greeting gestures the world over; in India, as in many other parts of the world, this implies distance in the case of strangers but it can involve a touching greeting in the case of friends. The notion of respect, then, admits of many variations of which familial respect and its counterpart, familial love, are but two. These values define touchability in Bastar; the touching gesture of familial love quite literally makes sense of the notion “kindred.” Touchability is the biological condition for human existence but when it comes to the public expressions of touch between kin, friends, neighbors, and strangers different cultures draw the line between touchability and untouchability differently. The location of the line varies throughout India too but in Bastar, and Central India more generally, the complexity of formal greetings are such that the region is home to one of the more touchable cultures in the world. One shows respect to a foreigner or non-kin in Bastar by holding one’s palms pressed together at chest level and uttering “Johar” or “Ram Ram!” This is usually a simultaneous mutual action in the case of equals but asymmetrical in the case of unequals. When the respected person is a Brahman the respecter may raise the hands higher or, in some cases, even prostrate themselves on the ground. The general rule, which admits of exceptions in the case of ritual friends, is that if the other person is non-kin then one does not touch regardless of whether the person is of higher, equal or lower status. Kin, on the other hand, touch when greeting although weddings redefine relatives and turn joking relations into avoidance relations based on extreme mutual respect. There are many different ways to touch kin when greeting in Bastar; these tactile gestures are used along with a variety of other visual, audio-vocal, olfactory and gustatory signs to order kin in terms of closeness through the expression of various degrees of respect and affection. The basis of everything is the familial kiss. This is a form of tactile communication that presupposes reciprocal recognition between two people who give and receive familial respect and love. The familial kiss defines and redefines kin relations on a daily, weekly and yearly basis. Estrangement, anger and hate are facts of life in Bastar as everywhere; when domestic violence replaces domestic harmony the clenched fist replaces the familial kiss as the dominant mode of expression. Needless to say familial love and domestic harmony are valued positively as virtues, hate and violence as vices. The Halbi for kiss is cumto, a word I first heard in the context of a greeting where one person touches the feet of the other and receives a touch on the chin.1 The action did not match my conception of the English word “kiss.” As the word “cumto” is of Sanskrit origin and always translates as “to kiss” I reconciled the difference away as a quaint tactile euphemism of a culture where public displays of affection, and especially kissing, are taboo. Further ethnographic research on salutations in Bastar revealed that my understanding of the Bastar kiss was constrained by my unexamined Eurocentric conception of the kiss, a problem that has its origins in the English dictionary definition of the kiss. The OED definition—to press or touch with the lips (at the same time compressing and then separating them), in token of affection or greeting, or as an act of reverence—captures the basic sense of the English kiss. It is something that you do with your lips be it to your lover, your mother, the other or the Pope. Even non-tactile kisses are made with the lips: the sonic kiss is the smacking sound we make with our lips; the visual kiss is the pouting gesture we make with our lips. This definition of the kiss is perhaps not so much Euro-centric as lip-centric as a closer look at the Bastar kiss reveals. The untrained outsider’s eye observes what appears to be a simultaneous gesture when one person receives a touch on the chin and whilst touching the other’s foot. The indigenous interpretation is different. The Bastar familial salutation consists of two integral parts: an action and a delayed reaction. A junior person initiates the greeting by touching the feet of the senior. This, the people say, is an expression of respect (man). The senior person then touches the chin of the junior as an expression of love (maya). Only the second greeting is called a kiss (cuma). The first action is called pay parto, literally “feet push,” an act of respectful humility and obeisance. The familial kiss in Bastar, then, is a reaction by a senior to an initial action of a junior; the bodily instrument of this kiss is the right hand and not the lips. The Bastar familial kiss, then, is a solicited salutation; it a non-verbal expression of familial love, a reciprocal recognition of familiarity and a feeling in the literal sense of that term. Furthermore, it is something that only a senior has the right to convey, and something that he or she will only do if they are shown respect in the first place. In Europe it is the other way around: the public kiss is an expression of profound respect. The Pope, and his like, hold out their hand to be kissed; they do not do the kissing except when greeting a higher divine authority, be it Mother Earth or Jesus Christ. The exception, and a highly significant one as we shall see, is the Christian wedding where the groom is enjoined to kiss the bride at the end of the wedding ceremony. This one-way ritual male kiss serves to define the couple as the family unit in Europe. In Bastar, by contrast, the familial kiss defines the kindred, the members of the brotherhood, dadabhai, and the familial “otherhood,” saga; at the wedding the unity of the bridal couple is defined by a mutual feeding ritual, not a one-way kiss. The Bastar familial kiss has many features that define it as culturally specific to Bastar District, features that simultaneously reveal it to be a recognisable variation on a general all-India, non-lip kiss theme. Hindi-speaking migrants in Bastar, I have noticed, do not use the word Hindi word kiss (cuma) to describe a senior’s reaction to a gesture of respect and nor do they make the hand-to-chin gesture. Rather they place their hands on, or over, the head of the junior in a reaction that is called a blessing (ashirbad dena). Like the Halbi hand-to-chin kiss, this is an expression of familial affection. Another form is the so-called “sniff kiss” which seems to be widespread throughout Asia as a whole. The basic sense of the Indian hand-to-chin kiss, then, is a complex multi-modal one that could not be more different to the English lip kiss. Like kisses everywhere, the public expression of a kiss in India is subject to taboos and unwritten moral codes that become apparent only when violated. Thus when the actor Richard Gere embraced and lip-kissed Bollywood actress Shilpa Shetty on the cheek at an AIDS awareness rally in New Delhi on 15 April 2007 it was an act that could not have been better planned to offend local sensibilities. The right-wing Hindu nationalist group Shiv Sena viewed it as an obscene act, burned effigies of Gere and set fire to pictures of Shetty. Court charges have been brought against Gere. Such is the case too with Bollywood star Aishwarya Rai who was ordered to appear in an Indian court after a screen kiss with co-star Hrithik Roshan in movie Dhoom-II was deemed “obscene” even though Rai kissed the actress lightly on the cheek. Another source of cross-cultural misunderstanding here is the many different meanings of the word “kiss.” The Halbi word for kiss, cumto, has two distinct meanings as a form of touch. It can refer to an expression of sexual attraction between lovers or to an expression of familial love between kin. The physical expression of the sentiment, too, is quite distinct: the former is an action of a very private kind that should not be performed in public; the latter a familial gesture of a public kind that is, on certain ritual occasions, obligatory. The English word “kiss,” too, has this dual meaning and although there are variations in the modes of their expression which reflect the different cultural contexts. The lover’s kiss is a private act that should not be performed in public, but the violation of this norm does not excite the strong emotion it does in India. The familial kiss is a public gesture in the same way that a Halbi kiss is but it is usually a mutual rather than a one-way action. Furthermore, in the Australian culture that I am familiar with, it is something that men do with female family members, not with males. Thus I give my aunty a mutual closed lip-to-lip kiss, but I shake my uncle’s hand, an action I also perform with non-kin. Skinship as a theoretical concept The preceding discussion has examined the history of the word “skinship” and some of its possible referents in different cultures. The task now is to define the concept in a theoretical way that is useful for ethnographic research. It is obvious that the scope of the idea of skinship is vast and that a keyword like “kiss” varies greatly in semantic range and cultural significance both across cultures and within them. Not all kisses are gestures of familial love and not all gestures of familial love are called kisses. Nevertheless it is clear that a touching familial gesture, be it called a kiss or something else, is an elementary form of face-to-face interaction within families everywhere. This much is true by definition: family members are familiar with one another in an intimate, sensible way. Familiarity as a value informs the expression of touching familial gestures but always in the context of respect, a value that has even greater generality. But respect is a value that creates distance between people in face-to-face interactions whilst familiarity draws them together. A general concept of skinship must capture this inherent tension in the micro-space and micro-moments of face-to-face interaction. Skinship as sensible face-to-face communication At the most general level skinship can be defined as sensible face-to-face communication. The word skinship privileges touch but the intimacy of touch means that it excites all the senses. As a sensible mode of communication it necessarily involves both verbal and non-verbal communication. The study of the referential semantics of kinship terms, while obviously an important branch of the study of kinship, abstracts from nonverbal forms of communication and is quite literally non-sense in the literal, nonpejorative sense of the term. Skinship, as the concrete study of sensible communication, must precede the study of semantics; it is a complement to the study of kinship semantics not a critique of it. As a face-to-face mode of communication skinship is concerned with both verbal and non-verbal modes of address. A reference term such as “father” can function as a mode of address but with a totally different significance. The use of “father” as a reference term raises the question of the synchronic relationship between ego as propositus and alter-ego as referent; it also raises the question of the logical relationship of the term “father” to reciprocal terms such as “son,” “daughter,” and “child.” The use of “father” as an address term, by contrast, raises the question of the inter-temporal social relationship between addressor and addressee; it also raises the question of the affective relationship of the term “father” to alternatives such as “dad” and “daddy,” and to familiar reciprocal usages involving names, nicknames, and terms of endearment. Address terms, as the linguist Jakobson (1960) notes, raises the question of the emotive and conative functions of language, not the referential. Understanding modes of address, then, is a very different theoretical agenda. As a form of communication, skinship is concerned with the differences between public and private modes of communication with emphasis on the former because these public “interaction rituals,” as Goffman (1967) has termed them, provide the key to understanding the unwritten moral codes and values that inform face-to-face interactions. The definition of skinship as a sensible mode of face-to-face communication is very broad and the critic might well ask how this differs from the classic approaches of Hall (1968) and of Goffman (1967) to the study of face-to-face communication. This is a question that must be answered with some precision because it is on the foundations laid by these theorists of sensible communication that the idea of skinship as a theoretical concept must be built. Hall’s (1968) theory of sensible communication is the most general. He is concerned with the defining characteristics of human communication as a species of sensible communication in general. Goffman (1967) for his part is concerned with the defining characteristics of one specific type of human face-to-face communication, that which occurs in the hierarchical institution. He conducted fieldwork in a hospital and developed a general theory of “interaction ritual” from this particular case study. His central concern was to show how respect as a value expressed itself in face-to-face rituals of deference and demeanor. The family is a hierarchical institution and as such his work provides a general conceptual framework for analyzing respect. But familiarity is a key value in the household and Goffman’s conceptual framework needs to be extended to allow for this value but at the same time restricted to a consideration of familial respect. Skinship thus becomes the analysis of sensible modes of face-to-face familial communication. “Familial” in this instance can be defined as the relations of contiguity, consanguinity and affinity. Kinship, as the title of Morgan’s (1871) pioneering study illustrates, is classically defined as the study of consanguinity and affinity. Contiguity is sidelined; it is deemed to have no determining role in the definition of kindred. The study of skinship as face-to-face familial communication inverts this classic definition and assigns pride of place to contiguity. But what does “contiguity” mean in this context? Enter Hall and Goffman. Their theories of face-to-face sensible communication are based on a very precise conception of the idea of contiguity, one that recalls the Latin origin of the word “contingere,” to touch. Hall coined the term “proxemics” to refer to what was previously called the “social space of bio-communication” or the “micro-space of interpersonal encounters.” He was concerned with the human perception of and use of space in general. He defines four zones of interpersonal encounter—the intimate, the personal, the social, and the public—which begin at distances of 0 feet, 1½ feet, 4 feet and 10 feet respectively. These old measures of distance evoke the idea of contiguity as touch in an evocative bodily way that metric measures do not. When the feet of two people are intertwined in the centre of the intimate zone tactile communication is at its maximal skin-to-skin limit and all senses are excited; as the feet draw apart gustatory communication drops out and tactile communication becomes restricted to the reach of the hand as we move out of the intimate zone and into the personal; touch begins to get very difficult as the beginning of social zone is reached at 4 feet; thereafter tactile communication becomes impossible and this allows the visual and auditory-vocal modes of communication to reign supreme. Skinship, as a social relationship of contiguity, occupies those zones where touch is possible, the intimate and personal; but it also appropriates some untouchable space in the social and public zones where it consigns estranged kin and those kin who must be avoided. This kiss lends itself to an analysis of this type (see Table 1). Neither Hall nor Goffman were concerned with gestures of this kind but it is useful to illustrate their concept of contiguity in terms of an analysis of the kiss to identify both the importance and the limits of their theories. Not all kisses are gestures of familial love but they are a useful starting point for trying to get some idea of the concept. The mouth-to-mouth or lover’s kiss is literally the most sensible and most communicative because this is the only form of kiss that excites all the senses. It is the most intimate mode of communication and the one with the highest degree of multi-modality. When we kiss we taste with our tongue, we touch with our skin, we pick up aromas with our nose, sounds with our ears, and the sights with our eyes if we choose to keep them open. The paradoxical fact about this most supreme form of communication is that we cannot speak. The other important fact, of course, is that the kiss is the site of human sexual reproduction and is associated with the most intense of all human emotions: those “animal passions” of ecstasy, anxiety, jealousy, anger, sorrow, shame and hope. The lover’s kiss admits of a great number of sub-types. Chapter 3 of the Kamasutra (Doniger and Kakar 2002) deals with the question of kissing. In keeping with the Indian passion for classification, it distinguishes sixteen types of kiss and eleven places on the body where a kiss can be placed. The sixteen kinds of kiss fall into two broad classes: the first is the lover’s kiss of which there are 15 types; and the second is the so-called “transferred kiss” which gets only a passing mention in the Kamasutra. The latter is the key to understanding the familial kiss but to grasp this notion it is necessary, firstly, to consider the lover’s kiss. Table 1. The communicative dimensions of different types of kiss The very first type of lover’s kiss is called the “nominal kiss.” This is when “a girl only touches the mouth of her lover with her own, but does not herself do anything.” Next comes the “throbbing kiss” when “a girl, setting aside her bashfulness a little, wishes to touch the lip that is pressed into her mouth.” This leads logically on to the third type of kiss, called the “touching kiss.” This is when “a girl touches her lover’s lip with her tongue, and having shut her eyes, places her hands on those of her lover.” And so the classifying goes on until we reach a type the ultimate lover’s kiss called “fighting with the tongue.” The fundamental principle informing the lover’s kiss, the author of the Kamasutra argues, is the norm of reciprocity. This is expressed in a verse which goes as follows: Respond to an action with a counter-action, to a blow with a counter-blow, and by this same logic, to a kiss with a counter-kiss. (Doniger and Kakar 2002: 45) This observation is of anthropological interest because the “battle of the tongues” marks the limits of reciprocity for here two people become one as they engage in the mutual consumption of shared bodily substances through the mouth and the mutual inhalation of breath through the nose. The symbolic potential of this idea of two-becoming-one is something that theologians and poets of all ages have exploited, and something too the Freudians have long noticed. A highly significant fact about the tongue kiss, Phillips (1993) remarks in his Freudian essay on the kiss, is that we cannot do it to ourselves. We can hit ourselves, stroke ourselves but not kiss ourselves. The tongue kiss is a communicative act, but is one that we cannot perform until we stop talking. The lover’s tongue kiss, he adds (Phillips 1993: 103), is about mutuality not domination: “When we kiss we devour the object by caressing it; we eat it, in a sense, but sustain its presence.” Kissing on the mouth, notes Phillips, can have a mutuality that blurs the distinction between giving and taking, a point he illustrates with a quote from Shakespeare: “In kissing do you render or receive?” asks Cressida. “Both take and give,” comes the reply. The lover’s kiss, then, is a gustatory form of human communication where the distinctions between the obligations to give, receive and repay are blurred as giver and receiver become one. All the senses are involved except speech but taste, or rather simultaneous mutual tasting, is its key defining characteristic along with simultaneous mutual inhalation. This is the most sensible of all forms of human communication. After the lover’s kiss comes four types I have arbitrarily labeled the skin, sniff, sonic, and visual kiss. These can be distinguished by transcultural sensible criteria summed up by the degree of multi-modality shown in the final column of Table 2. The skin-to-skin kiss, which admits of many sub-types, occurs in the intimate zone at a distance of less than 1½ feet; all senses except taste are excited by interactions of this type. Next come the non-touch kisses: the sniff-kiss that relies on the nose assisted by the ear and eye; the more distance sonic kiss that relies on the ear and eye; and finally the visual kiss, the most distant of all that relies on the eye only. The Kamasutra (Doniger and Kakar 2002: 45) groups these four types under the generic heading of the “transferred kiss.” This is a particularly apt expression because the “transferred kiss” takes us from the private world of erotic love to the public world of familial love and from simultaneous mutuality to unilateral giving of which a mother’s loving kiss is the supreme example. Skinship in the Japanese sense of the word begins here. A mother’s loving kiss to the head of her breast-fed baby is part of human nature in the sense that it is not an interaction ritual of the type that poses difficult problems of cultural interpretation. The mother’s kiss is the most intimate of the transferred kisses in the intimate zone. More distant kisses in this zone are doubly-transferred gestures of affection and familiarity; they take us from the realm of motherly love to interaction rituals of a type that do pose problems of cultural interpretation. The gestures of affection and familiarity found in this slightly more distant zones admit of a multitude of culturally specific types, not all of which, we have seen, are called “kisses.” The Bastar familial kiss is one such example. Its form is obviously based on the mother’s kiss in that it is an asymmetrical gesture of affection given by a senior to a junior. Halbi speakers use the word “kiss,” cuma, to describe both but there is an important conceptual difference: the mother’s kiss is an unsolicited one-way act whereas the familial hand-to-chin-touch gesture is a solicited ritual interaction. A child solicits this chin-touch salute by showing respect (man) to an elder by touching their feet. The elder person will then touch the chin of the younger and may utter “tch tch.” This is an expression of familial love (maya). This familial kiss is a classic example of culture as learned behavior. Children are taught to show respect with a foot-touch salute from about the age of three or four years; they begin to return the chin-touch kiss to younger siblings around the ages or nine or ten. The inter-temporal nature of the solicited gesture of familiarity means that the senior has the option of not conferring kinship upon the junior, something that can happen if kin become estranged and familial hate prevails. Skinship in the anthropological sense that I want to give the term begins with doubly-transferred tactile gestures of familial love and affection and includes other non-tactile kisses of the sniff, sonic and visual types. These are all interaction rituals of a symbolic type that pose problems of cultural interpretation. Take the so-called “sniff kiss” for example. In some pastoral societies this form of familial kiss is modeled on the behavior of animals towards their young. The Vedic poets, Sanskrit scholars (Hopkins 1907: 121) tell us, had no word for kiss; they employed the word meaning “sniff” or “smell” gghra) to describe an action that was explicitly modeled on the behavior of their animals. Just as a cow recognizes its calf by means of smelling, so must the father recognize his new born child with a “thrice sniff at the head” (1907: 121). Other texts describe how a father when returning from a journey must sniff at the head of his children and low like a cow (Müller 1879: par. 11). In Fiji, by way of contrast, the familial “sniff kiss” is mutual rather than asymmetrical. When mother and daughter greet they touch cheeks and make a sniffing noise. However, this is an outsider’s perspective. Closer observation and discussion reveals that, conceptually speaking, the gesture is similar to the Bastar familial salute in that it consists of an inter-temporal action and reaction: the junior makes the first move as a show of respect; the mother responds with her gesture of affection a fraction of a second later. The sniffing action poses an additional problem of interpretation here because while the nasal gesture sounds like a sniff Fijians do not perceive it as such. No Fijian I have spoken to has been able to interpret the gesture for me but agreed that “mutual inhalation” rather than “mutual sniff” was a more accurate description. The “mutual inhalation” interpretation suggests itself from a triply-transferred kiss of the type that takes us from the realm of kinship to that of religion. Oneness is a key religious idea and the symbolic potential of the lover’s kiss has been exploited by many a theologian. The early Christian theologians, Perella (1969) notes in his scholarly book on the Christian kiss, used the mouth-to-mouth kiss as the symbolic key to unlocking the mysteries of Christian ritual symbolism right from the beginning. The medieval mystic St Bernard, to take but one example, argued that when the priest asks the groom to kiss the bride she is receiving an infusion of the Holy Spirit because the groom’s kiss transmits the breath of Christ. In other words, the groom is Christ for this sacred moment. Furthermore, when the Gospels say “I and the father are one” this is to be interpreted as the Mouth-to-Mouth Kiss of Father and Son. The mutual breathing brings about an indwelling of the Father in the Son and the Son in the Father. And so the idea is developed year after year, century after century and in country after country in Europe. Christianity, then, transformed the erotic lover’s kiss into a religious salutation; they got rid of the sensual dangers that lurked in kisses, Perella (1969: 29) notes, by defining as impure that kiss which someone does for a second time because they found it enjoyable. The Fijian mutual-inhalation familial kiss poses the question of the relationship of kinship and religion. Hocart (1952: 198) who conducted fieldwork in both Fiji and South Asia among other places, argued that “nine-tenths of all kinship systems are religious in origin” and that the key to understanding them is to be found in ideas of re-incarnation. Bastar skinship, as we shall see, gives tactile expression to this idea. However one might want to interpret the Fijian familial kiss it is an “interaction ritual” in Goffman’s sense. But what does this mean? Goffmam on ritual interaction Goffman (1967: 57) informs us that he uses the term “ritual” to describe face-to-face interactions “because this activity, however informal and secular, represents a way in which the individual must guard and design the symbolic implications of his acts while in the presence of an object that has special value to him” (emphasis added). Values inform the symbolic actions and respect is the supreme value for Goffman. Respect for him means sentiments of regard such as obeisance, submission, and propitiation; affection and belongingness; politeness and honor. By “interaction” he means face-to-face relations between two people for a moment in time and a limited extension in space: the glances, gestures, positioning and verbal statements that people consciously and unconsciously exchange on these occasions. These moments of time are “necessarily evanescent, created by arrivals and killed by departures” (1967: 2); “the intimate spaces are where there is ‘a close meshing with the ritual properties of persons and with the egocentric forms of territoriality” (1967: 1). Goffman’s analysis is concerned in the main with the analysis of the actor and the patient in a ceremonial context. He takes the ego-perspective of the actor in his analysis of the two components of ceremonial activity, deference and demeanor. “Deference” refers to the respect for others that the agent displays whilst “demeanor” refers to the agent’s self-respect through deportment, dress and bearing. He notes, of course, that the patient sees things differently. “Rules of conduct,” he notes (1967: 49), “impinge on an individual in two general ways: directly, as obligations, establishing how he is morally constrained to conduct himself; indirectly, as expectations, establishing how others are morally bound to act in regard to him.” “What is one man’s obligation,” he adds, “will often be another’s expectation.” Goffman’s general analysis of respect behavior in a institutional context such as a hospital can, as mentioned above, be extended to the familial context of a household by (a) restricting the notion of respect to familial respect and (b) introducing the question of familiarity as a value. The latter introduces the re-actor into the picture; this is the valuer who literally makes sense of familiarity as a value. Thus in its most elementary form skinship is an interaction ritual that involves not just rituals of deference and respect but crucially also solicited rituals of familial love and affection. These loving gestures quite literally make sense of kinship but only for people involved in face-to-face relationships in historically specific moments of time in intimate geographical spaces. These interaction rituals, in turn, are embedded in ever-larger historical times and ever-wider geographical places which the ethnographer must consider. Kinship as skinship is a fragile state of reciprocal recognition that must be continually reproduced because of the ever-present danger that feelings of familial love will turn into its opposite, familial hate. Gesture in this negative a micro-space is important too and there could be not better illustration of it than the line in Othello: “I understand the fury in your words. But not the words.” Implications of the concept of skinship for the ethnographic study of kinship The Hall/Goffman notion of “contiguity” enables us to define skinship in general terms as face-to-face relations of familial communication of a touching kind. But what are the implications of this for the study of kinship? The answer to this question, as I see it, is that it takes us from the study of kinship as the semantics of reference terms to the study of kinship as the pragmatics of modes of address. This takes us from a concern with kinship as a relation of consanguinity and affinity to kinship as a relation of contiguity. Trautmann’s classic analysis of Dravidian kinship exemplifies the importance of the semantic analysis of reference terms for the development of human understanding. It reveals how the anthropological method of abstraction and comparison can yield transcultural generalizations of great explanatory power. Interestingly, while the relations of consanguinity and affinity were his primary focus, his work represents a departure from previous work in that he stressed the importance of historical relations of contiguity. This notion of contiguity, however, bears no relationship to the notion contiguity developed above. Space for him is not the micro-space and micro-time of Goffman but the cultural geography of Indian kinship that a deep history of migration has produced. He finds that the three great waves of migration into India that began about 3-4000 years ago—the Dravidian and Indo-Aryan speakers from the west and the Munda speakers from the east—have defined three distinct zones of kinship. His comparative study of reference terms has identified these regions. Cross-cousin marriage characterizes the southern Dravidian system, the absence of cousin marriage combined with the superiority of wife-takers characterizes the northern Indo-Aryan system, while cousin marriage and alternating generations distinguishes the “frontier zone” in central India. Trautmann’s analysis is abstract in that the study of modes of address is beyond the scope of his analysis. In this respect his approach is conventional. He deals with terminological data from over twenty different dialect areas of India but never once considers the problems posed by address terms. The fact is that he could not have considered them even if he wanted to because the ethnographic data is simply not available for the most part. Bean’s Symbolic and pragmatic semantics: a Kannada system of address (1978), is the only book I am aware of that deals with the pragmatics of address in an Indian culture and even then her account is restricted to verbal modes of address. One searches long and hard to unearth the few articles on the subject that do exist (See, for example, Das 1968, Mehrotra 1977, Vatuk 1969). Of course, it is a perfectly legitimate exercise to abstract from verbal and nonverbal modes of address when the problem under investigation is the semantics of reference; nay more, the method of semantic analysis requires it. If this mode of analysis has its shortcomings then it is in its very success. The analysis of reference terms has been a central concern of anthropology since the days of Morgan and there are few comparative problems left to solve. An important recent collection of essays edited by Trautmann and others (Godelier, Trautmann and Tjon Sie Fat 1998) has tried to revitalize the study of kinship but, paradoxically, has failed to precisely because it has successfully resolved most of the interesting comparative questions that remain. What then does a move from the study of the semantics of kinship terms to the pragmatics of modes of address entail? The conceptual framework that I consider necessary for such a task has been spelt out above. It general terms it involves an ethnographic investigation of the role of the values of familial respect and familial love as expressed in face-to-face ritual interactions. Of particular importance is the manner in which the tension between the closeness and distance of kin is resolved (or not resolved). However, when one contemplates the implications of this for ethnographic research it becomes easy to see why the analysis of reference terms has dominated the agenda. Semantic analysis has the advantage that it is relatively easy. Kin lexicons are relatively small data sets, with most systems having between eighteen and thirty-five distinct reference terms (Kroeber 1909). This fact, combined with the abstraction from sensible modes of expression, simplifies the analytical task. Of course, analyzing these terms can become very complicated as one struggles to comprehend relations within and between systems and their diachronic transformations. Ethnographic data on modes of address is as complicated as reference lexicons are simple. Whereas a reference lexicon consists of a list of 20-30 terms, data sets on modes of address are of seemingly unlimited size. Halbi, for example, has fifty-six reference terms of which around thirty also function as modes of terminological address. In addition there are countless personal names and nicknames as well as four distinct call sounds, e na! e go! e O! and e ho!, that are used in highly specific ways. They also have a complicated system of non-verbal modes of communication of which the familial salutes are the most important. These consist of eleven gestures of respect and nine gestures of familiarity which, when combined in different ways, are capable of generating a large number of salutations that define many shades of grey between the “black” of extreme respect and the “white” of extreme familiarity. To complicate matters further, usages depend on whether the occasion is private or public and whether a public occasion is an everyday interaction or a life-cycle ritual. Usage also varies over the life cycle of a person. Finally, the values of respect and familiarity that inform the ritual interactions are difficult to access, influenced as they are by local religious beliefs and traditions that have become “natural” to insiders. Nevertheless, it is important not to fetishize the complexity for just as a tree has its origin in a single seed, so too is the tree that is skinship in Bastar simply complex. To grasp the simplicity in the complexity one needs to focus on the trunk and its main branches, not the leaves that arise and pass away. In the context of Bastar we can do this by focusing on the salutations because the branches they define stem from the two root values of familial respect and familial love. Words of address, as the instruments of fine discrimination, define the smaller branches and leaves. This tree image has its analogue in the notion of a semantic tree of reference terms that semanticists sometimes use. The two trees are different but related in the way that a real tree has to a simplified sketch of a tree I now use the Bastar case to illustrate, if only sketchily, what is involved in a skinship approach to the study of familial relations. I do this by examining how the Bastar salutations vary over the life-cycle of person. The changing form of the Bastar familial salutation over a person’s life-cycle The following data, which is drawn mainly from that used by the Maraar community in north Bastar, charts the changing form of the familial salutation over the life-cycle of a male. Domestic space in Bastar is defined by men in relation to women as brothers then husbands. We must locate ourselves in this domestic male space if we want to grasp the female point of view in face-to-face relations because, as we shall see, cross-sex interactions are the source of most of the complications in the system. Birth The life (jiv) of a new-born is said to begin around the fifth month of pregnancy when the mother feels her baby kicking. Motherhood begins when the mother starts breastfeeding her new-born baby. Breast milk is an obvious savory definer of the mother-child relation in all cultures but it seems to be especially important in Bastar. For example Gurumai Sukhdai told me a tale about the son of a king who was separated from his mother at birth. His father’s jealous co-wives were responsible. They told the mother that she had given birth to a cat and that they had to throw it away. The baby was buried in a pile of manure but survived. He was rescued by a dog and lived in its stomach for a while until he was cared for in a cow’s stomach. The boy grew up in this truly fantastic way unaware that he had a mother; meanwhile she lived on never knowing that she had given birth to a son. After some two decades of extraordinary happenings the moment of reciprocal recognition comes at a public gathering where the mother and son come face-to-face and relate their stories. When the truth of the matter dawns on the mother she takes her now a 20 year old son, sits him on her lap, and breastfeeds him in front of the assembled crowd. The jealous co-wives are then tied to a horse’s tail which was galloped off with them at high speed. This tale describes how a breast-feeding ritual converts the grown man into a son. It is not a ritual I have ever seen, or am ever likely to see, but the tale is a classic illustration of the cultural importance of the most intimate sense, taste, in the construction of Bastar skinship. Mother’s milk not only gives a child strong immunities and helps it grow strong, as Mitiko the Japanese mother quoted above said, but it is symbolic of a common “milk line” that unites not just mother and child but also brother and sister as “milk siblings.” In Bastar the milk line is reproduced down the female line from mother to daughter and from daughter to granddaughter. This fact, as we shall see, gives the brother a moral right to reclaim his share in his mother’s milk by requesting his sister’s daughter as a bride for his son. Fatherhood is ritually defined at the naming ceremony held a few days after birth. What anthropologists call “the equivalence of alternating generations” is, for the people of Bastar, a way of talking about rebirth beliefs. A man is reborn as his classificatory son’s son which means that the birth of a man’s son is the rebirth of his classificatory father. This belief is ritually enacted at the naming ceremony when the father’s sister (bubu) of the new-born is asked to comfort the crying baby. She refuses to do to so until she is given a sari. She takes the baby boy, comforts him, and addresses him as “father” (baba). Her brother, the baby’s father, also addresses him as baba. The baby’s mother addresses him with a term of endearment or by name but never as baba. The same ritual takes place when a girl is born but on this occasion the mother alone will address her daughter as “mother” (aya) because her daughter is ritually addressed as if she was her reborn mother. Other rituals are performed that strive to establish the exact identity of the reborn ancestor because an individual’s rebirth is never a simple mechanical process. For example if a person consistently and wantonly violates the moral codes of society, such as those that respectful modes of address help to define, then they may be reborn as ant or some other lower form of life. The new-born infant is the recipient of familial love for the first few years of its life but as it grows up it and becomes more independent the love ceases to be so freely given; the young child must learn how to solicit gestures of familial love by learning how to show respect to elders. Childhood As a boy grows up and learns to talk the first kinship language he learns are modes of address. When interacting with his father, for example, he learns to address him as baba; he also learns to respond to the different modes of address his father uses when trying to attract his son’s attention. In addition to baba his father may the special call sign e na! Or his name or nickname to attract his attention. As he gets older he will learn how to show respect to his father and all other relatives senior to him by imitating the hand-to-foot touch salute he sees other youngsters doing. He will expect his salute to be reciprocated by a hand-to-chin familial kiss, a familial gesture he will begin to use himself around the age of nine or ten when younger children give him a respectful greeting. By this time he will have learned how to address a wide range of kin and how to response to different address terms, call sounds and nicknames. Reference terms are the last thing he will learn to master. Understanding the difference between reference and address is a matter of cognitive skill development but learning the fifty-six Halbi terms of reference is another matter. The reference terms used for senior kin are a relatively simple matter because in Halbi, as in most other languages, the same word is usually used for both address and reference. Referring to members of his own generation poses few problems too for in Bastar everyone is either a brother (bhai) or sister (bahin); the latter are divided into marriageable sisters (maina bahin), with whom he has a joking relationship, and unmarriageable sisters (an unmarked category of bahin). Learning how to refer to members of the first descending generation will be the last thing he learns as he addresses these youngsters by name. I often come across young men in their twenties who still get confused when asked to explain this part of their own reference terminology. By the time a young man and his unmarried siblings have reached their maturity they will have learned the modes of address for about fifty different types of face-to-face relationships. I list just one here, the father/son dyad, to give some idea of the complex way the elementary values of respect and familiarly are expressed. A young man will refer to his father as “my baba” and be referred to by his father as “my beta.” These reference terms are part of a semantic structure that makes four semantic contrasts: age, sex, generation and crossness. The reference terms “baba” and “beta” belong to the same general semantic field in that they are male, and parallel, II; this field is bisected by generation with “baba” belonging to G+1 and “beta” belonging to G-1. These semantic fields, it must be stressed, are not expressions of respect and familiarity as values. The modes of address do this and it ways that strike the outsider as counter intuitive and contradictory. Table 2. Reference terms and modes of address for the father/son dyad The familial salute conforms to what an outsider would expect: the son touches the feet of the father as a gesture of familial respect and receives a touch on the chin as a gesture of familial love. This is a ritual interaction in that it is used to mark the beginning or end of some occasion be it the arrival or departure of the son after an absence or on some formal ritual occasion such as a wedding. The fact that a father addresses his son by name whilst a son cannot is also an obvious gesture of familiarity; the son must show his respect by using the term “baba.” The bipolar usage of the term “baba,” and the asymmetrical usage of the call sound “e na! By the father when addressing his son, is counter-intuitive for the outsider but they make good sense to the insider because they are part of the unwritten moral code that governs skinship in Bastar. The rebirth beliefs imply the theological oneness of alternate generations which means that mutual respect is the right conduct between members of alternate generations. The reference lexicon contains no evidence of alternation, it appears only in address. We can express this “equation” formally as: baba/baba :: G+1/G-1 :: father/son. This is a formal way of expressing an interaction ritual: a father in generation G+1 addresses his son in generation G-1 as baba and the son replies by using the same term of address. There are many other examples of this. For example when a man interacts with his maternal grandfather they address each as aja. The equation in this case takes the form: aja/aja :: G+2/ G-2 :: mother’s father/daughter’s son. When a woman interacts with her daughter they address each other as aya. Thus: aya/aya :: G+1/ G-1 :: mother/daughter. When a woman interacts with her brother’s son they both use the call sound “e na!” Thus: e na!/e na! :: G+1/ G-1 :: father sister/brother’s son. And so on. The father’s asymmetrical usage of the call sound “e na!” must be understood in the context of other usages of call sounds. It suffices to note here that it is a sign of respect certain men in the first ascending generation G+1 must show to certain men in the first descending generation G-1. This is an inversion of the expected order of things but the tactile sense of the inversion will, I hope, become clear when we come to the end of our young man’s life cycle. The table of modes of address above, then, seems contradictory: the salutes and naming practices followed by father and son in their interactions follows the “natural” order of respect and familiarity but the usages of address terms and call sounds do not. How do people cope with these contradictions? If morality is the art of living together well, as Mauss (2007: 156) noted, then “contradictory” modes of address of this kind are the tools that people use in face-to-face situations when practicing the pragmatic art of trying to live together well. The alternative modes of address are, of course, only contradictory from an outsider’s point of view. From an insider’s perspective they are just different variations on the underlying values of familial respect and familial love. The “web of kinship” is extremely complex and face-to-face meetings with different kin at different times and in different places requires one to act in different ways that are appropriate to the occasion in question. Living together well with kin is an art that must be practiced with a skill learned over time; it is not a simple matter of following rules. Right conduct is a pragmatic question that inter-actors must decide upon given the circumstances they are confronted with. The general principle of all kin interactions, however, is embodied in the common sense of the familial salute. The basic values that inform this mode of address are the gold standard of all face-to-face relations. It is the only mode of address that is common to all sixty-four dyads that can be distinguished. For the young unmarried man or woman this familial salute is what defines their kindred group. Non-kin do not participate in this touching interaction ritual; they are untouchable and are treated with distanced mutual respect or, in the case, of friends incorporated as quasi kin with gestures of convivial mutual respect which may or may not involve touch. In Bastar, and in central India more generally, there is a special category of friend formed by the performance of a religious rite. There are many categories of these “ritual friends” as they are called in the literature (Skoda 2004). Some become “like brothers” but the supreme ritual friends are those who are deemed to become one; they are said to be “identical” as distinct from “equal” or “similar,” an ideology that is enacted in their ritual greetings when they unite in one embrace (Pfeffer 2001: 113). Table 3 recapitulates in visual form my argument about the need to see the Bastar familial salutation as an interaction ritual consisting of two distinct component parts. First comes the junior kinsperson’s feet-touch gesture of familial respect. Then, a fraction of a second latter, comes the senior’s chin-touch gesture of familial love. The final column captures the compound action and re-action which appears, to the outsider, as a simultaneous interaction. This familial salute, then, defines kin as touchable at the most general level of classification. This is the trunk of the skinship tree with its two root values, familial respect and familial love. The verbal modes of address provide the means of distinguishing one type of kinsmen from another. Unlike the reference terms that group and classify, the address terms divide, specify, and individualize. For example, the salute will identify the interactors as kin in general while the verbal modes of address will identify them precisely as, say, “Tom” and his “uncle Harry,” Tom’s father’s youngest brother. Table 3. The two component parts of the Bastar familial salutation For the young unmarried man and his unmarried siblings the touchability of kin is about to be redefined as he and his siblings reach marriageable age. By this time he will have attended countless weddings but with the marriage of his own siblings “closeness” and “distance” acquire a new tactile sense. These new relations can be labeled “affines” but the consanguine/affine distinction is not strictly relevant in Bastar because all kin are “consanguines.” it is more accurate to speak of “generic kin” and “specific kin.” Generic kin are defined, first and foremost, by the familial salute, i.e. by touch. Marriage requires that some of these generic kin be redefined as highly specific kin. These kin are “descriptive kin,” to use Morgan’s language, whilst generic kin are “classificatory kin.” Marriage The wedding ritual The claim by linguist David Crystal that there “seems to be little active role for the olfactory and gustatory modes in human communication” (1987: 399) could not be further from the truth in Hindu India. Hindu ritual, as Eck (1981: 11) has noted, [I]s sensuous in that it makes full use of the senses—seeing, touching, smelling, tasting, and hearing. Ones ‘sees’ the image of the deity (darsan). One ‘touches’ it with one’s hands (sparsa), and one ‘touches’ the limbs of one’s own body to establish the presence of various deities (nyasa).One hears the sacred sound of the mantras (sravana). The ringing of bells, the offering of oil lamps, the presentation of flowers, the pouring of water and milk, the sipping of sanctified liquid offerings, the eating of consecrated food—these are all the basic constituents of Hindu worship, puja. For its famous otherworldliness, India is a culture that has also celebrated the life of this world and the realm of the senses. The Bastar wedding is a Hindu ritual that not only involves “the ringing of bells, the offering of oil lamps, the presentation of flowers, the pouring of water and milk, the sipping of sanctified liquid offerings, the eating of consecrated food” of which Eck speaks, but it is also a celebration of the communicative power of touch. The ritual, which can last for up to a week, is for the most part a joyous occasion involving much dancing, laughing and horseplay with colored water and mud, joking that can at times lead to fights when it goes too far; for the girl’s parents it is a sad occasion involving much open weeping. It is also a serious religious occasion where the bride and groom are anointed with holy oil (dev tel) and turmeric to render them auspicious and pure. Relatives spend one day rubbing turmeric-oil (haldi tel)on the bride and groom in an upwards direction, and another day applying it in a downward direction. These rituals change the ritual status of bride and groom but other tactile rituals bring about a radical transformation in the relationship between different relatives of the bride and groom. Marriage of a man’s elder brother When a man’s brother gets married a new woman enters the household. Because men in Bastar marry “marriageable sisters” (maina bahin) with whom they have a joking relationship, it follows that brothers of the groom also had a joking relationship with the new bride prior to the wedding. This very close relationship must now be redefined using new terms of reference, new terms of address, and new salutes. When a man’s elder brother gets married the relationship with the new bride becomes even closer and is marked by a special salute which can be called the devar salute, literally the “second husband salute.” The word “devar”is the reference term for the bride’s husband’s younger brother. She addresses him as “babu,” a term of endearment for young kinsmen in general, while he addresses her as bohu. This is a relationship of the most extreme form of familiarity which the term “second husband salute” evokes. The younger brother is deemed to be junior to his elder brother’s wife and as such he is obliged to greet her with a foot-touch gesture of familial respect. As he attempts to do this she grabs his hands to prevent them touching her feet; she then reciprocates this attempted show of respect by a cheek-touch gesture of familial love that involves holding his face between her two hands (see Table 4). Because gestures of familial love are solicited reactions there are no salutes of mutual familiarity in Bastar but this one is very close to it. Table 4. The component parts of the second-husband salute Marriage of a man’s younger brother When a man’s younger brother gets married a redefinition of the polar opposite type occurs. A touchable joking relationship is redefined as a non-tactile relation of extreme mutual familial respect. Again this is marked strongly in the modes of address. The man refers to his younger brother’s wife as bohari and also addresses her as such. She refers to him, her husband’s elder brother, susara, and addresses him as “respected person” (bare man) or uses the call sign “e ho! But again it is the salute that captures the essence of the relationship. It can be called a “familial salute with touching words.” The new bride, the junior partner in the dyadic relationship, pays her respect to her husband’s elder brother by standing at a distance, covering her head, bending down, touching the ground in front of him, and saying “I touch your feet.” He replies by making an upward gesture with the palm of his right hand and says “Please arise” (utha) using the polite form of the verb (see Table 5). Table 5. The component parts of the asymmetrical verbal foot-touch salute The transformation of this relationship from a tactile joking relationship to an untouchable avoidance relationship, albeit in one that uses touching words, is created during the wedding in a special ritual called susara manto, “respecting the groom’s elder brother.” This ritual occurs at the house of the husband when the bridal party returns home. When the welcoming ritual is finished the bride sits on the lap of her husband’s elder brother. Village people from all castes come to offer them wedding presents. When this is finished, they tell the groom’s elder brother to cover the head of the bride. Then the groom’s elder brother covers the bride with a new sari. From that day on the bride regards him as susara, an avoidance relation, and covers her head in his presence. The ritual ends when the groom’s mother takes the bride’s hand and leads her inside her new house. This ritual marks the liminal state between the pre-marital use of the generic familial salute and the post-marital use of this asymmetrical verbal foot-touch salute. The message of the ritual is obvious and MS Mali confirmed it. “Look everybody,” the ritual says, “these people could joke with each other up to this very minute but not any longer. They should avoid each other from now on.” Behavior that was virtuous in the past has now become a vice. The following diagram (fig. 1) depicts this transformation. Figure 1. Ritual transformation of the generic familial salute into the distanced verbal foot-touch salute used by the elder brother’s wife. A remarkable feature of this most-extreme-respect salute is that the respect is inherited by the husband’s elder brother’s children. In other words, the younger brother’s wife shows extreme respect not just to the husband’s elder brother but also to his children. She maintains an avoidance relationship with them: she must not use their names and cannot touch them. The children likewise treat her with mutual extreme respect. The salutes they give each other are of the verbal foot-touch kind. Marriage of a man’s sister The marriage of a man’s sister is a major turning point in the life of a woman. She has to leave the household home and become someone’s wife in another home in another village. No new salutes are introduced at her wedding for there is little need for them. She moves away and daily contact with her brothers and their wives comes to an end. The brother continues to use the pre-marriage modes of address for her relatives. The out-marrying sister, has to learn new salutes as the incoming bride, the ones we have just discussed above. One’s own marriage Public displays of mutual affection between husband and wife are regarded as bad conduct in Bastar but the wedding ritual, a classic liminal rite, provides the one and only time when husband and wife can not only violate this norm but are ritually obliged to do so. This happens in a game-playing ritual called citi pasa khelto which is performed after the tying ritual and before the final feast on the last day of the wedding. A mat is spread out and the newlyweds sit equally apart on the mat. Rice is then is poured from a bowl and the bride and groom compete to scoop as much as they can to their side. When the gathering up is finished the respective rice piles are counted with a small pot. The bride will demand more rice from the groom’s parents. “Ale mother of babu,” she says, “give me rice. Ale father of babu, please give me some rice.” When this is finished they play another game with cowries or coins. When that is finished they take a leaf and fill it with cooked rice. They each take a leaf of food and try to force-feed each other. This is accompanied by much laughter and joking. This mutual feeding ritual is obviously a “transferred lover’s kiss” in the sense outlined in the second section of this essay. Mutual feeding, like the lover’s kiss, involves mutual tasting. Mutual tasting is the most intimate of all forms of mutual sensory excitation. It is also the supreme symbol of oneness, an idea that is fully developed in Hindu ideas about marriage. As Inden and Nicholas (1977: 26) note, “A husband is said to give his wife a new life by making her into his ‘half-body.’” When a man marries it brings him into closer association with his wife’s elder sisters. These relations are generic joking relations. He maintains this relationship with his wife’s younger sister but the relationship with his wife’s elder sister becomes an avoidance relationship of symmetrical rather than an asymmetrical kind. Because a wife’s elder sister is deemed senior to the man he must make the opening gesture of extreme respect. This involves holding the palms of his hands together in a prayer position, bowing his head, and saying “I touch your feet.” She reciprocates not with a gesture of familiarity but with one of equally extreme respect. She covers her head with a sari, bows slightly, and also says “I touch your feet” (see Table 6). Table 6. The component parts of the mutual verbal foot-touch salute Once again a special ritual is held to bring this transformation about. On the morning after the last day of the wedding, when all the ceremonies have been completed and the groom’s party is about to depart, a ritual called sas luga deto, “giving a sari to the wife’s elder sister,” is held. As they are about to leave the courtyard the bride’s elder sister blocks the way, sings a song, and says, “Give me a sasluga.” The groom gives her a sari and from now he regards her as his sas. He treats her with respect and the verbal touch salute is one sign of this. She becomes an “untouchable” but a familiar untouchable because the touching gesture is transferred from the tactile sensory mode to the auditory-vocal mode. Parenthood When a man and his wife become parents they become respected elders and they acquire the power to confer kinship with gestures of familial love. Age, then, means that people spend less time showing respect to elders and more time responding to gestures of respect from juniors. We move to the other side of the salute so to speak. One’s perspective changes but not the forms of the modes of address. No new modes of address are used until one’s children marry. When looking for a bride for his son a man will first consult his sister to ask for her daughter as a bride for his son. This is a birth-ascribed moral right. His sister and her daughter are in his mother’s milk line and he has a right to claim the return of the milk. This is called “returning the milk” (dudh lautna). Because women are reborn as their classificatory daughter’s daughter this is tantamount to the return of the reborn mother. The modes of address express this: a man’s sister’s daughter may be addressed as aya, mother. If the man is successful in securing his sister’s daughter as a bride for his son then cousin marriage of the patrilateral cross-cousin kind will have occurred. My data shows that this happens in less than 5% of all cases. Nevertheless, the interactions rituals between kin are performed as if it has occurred in all cases. For example, a father will address his son’s wife as aya even though she does not actually come from his mother’s milk line. This means that in 95% of cases marriage creates a consanguineal relation where none existed before. In other words, cross cousin marriage is an effect of 95% of marriages not their cause. The instrument that brings about this creation of consanguinity is a special kind of familial salute which is performed at weddings. The ritual is the most spectacular of all the familial salutes for all senses are excited. The ritual is called samdhi bhent and it marks the arrival of the groom’s side at the bride’s house where the wedding ritual will be performed. The parents of the bride greet the parents of the groom outside the front gate of their house. The fathers refer to each other as samdhi and address each other as such; the mothers refer to and address each other as samdhin. The greeting ritual lasts about fifteen minutes during which time the same-sex pairs pay extreme respect to each other in a series of simultaneous, or near simultaneous, ritual interactions. It begins with mutual feet washing during which they show their extreme respect for each other by drinking the dirty water. They then place auspicious red marks (tika) on each other’s foreheads, place flowers behind each other’s ears, and feed each other betel nut. Mutual hugging ends the ritual which may include other mutual exchanges. The hugging between men is sometimes very vigorous. They clash their bodies together in a manner that borders on the aggressive. Informants say that it is done out of mutual respect (man), not mutual love, and concede that an element of hostility is involved. The samdhin end their interaction ritual by familiarizing the classic palm-joined Hindu gesture of gesture (anjali) with a touch. The samdhi use this gesture in their subsequent future interactions. The opposite sex interactions between the parents is not ritualized. They are assumed to be brothers and sisters and give each other peremptory familial salutes of the generic kind. Table 7 captures in visual form the tactile component of this ritual. Table 7. Close and very-close mutual familial respect salutes This egalitarian nature of this mutual respect ritual presents a stark contrast to what happens in north Indian weddings. On these occasions the father of the bride, the bride-giver, defers to the father of the groom. For example, he will touch the feet of the groom’s father who does not react. There are many north Indian migrants in Bastar and the people of Bastar are familiar with these rituals. As such one cannot help but think that the highly elaborated form of the samdhi bhent ritual has been developed in opposition to these rituals from the north as a statement of Bastarian cultural identity and distinctiveness. Death In the normal course of events seniors die before juniors. For a man the deaths of his father and his mother’s brother are liminal states of great importance in Bastar. The father’s modes of addressing his son have been considered above from the son’s perspective so there is no need to rehearse these again. This leaves the mother’s brother/sister’s son relation which comes to the fore in death rituals. Just as a man’s sister’s daughter is a reborn mother who should be addressed as aya (mother) it follows that a sister’s son, bhaca, is a reborn mother’s brother, mama. This ideology is expressed in the form of an inverted familial salute. Bastar’s unwritten moral code obliges the mother’s brother to show his respect to his sister’s son by touching his feet. In its most highly elaborated form, which occurs at the mother’s brother’s funeral, the sister’s son receives extreme respect without reacting. In other words, the first part of the generic familial salute is severed from its usual context and turned into a stand-alone gesture of extreme asymmetrical familial respect. Of course, the mother’s brother is not around at the time of his funeral to carry out this ritual but his elder and younger brothers, as classificatory mama, may be and if so will do it. In some communities of Bastar a mother’s brother will show his sister’s son this foot-touching respect in everyday contexts throughout his life. The relation between a father’s sister and her brother’s son is analogous. In other words, cross-parents show extreme respect to cross nephews by means of the unreciprocated foot-touch gesture of respect (see Table 8). Table 8. Foot-touch component of the familial salute as gesture of extreme respect for the cross-nephew These implications of this salute are two. Firstly, the fact that it is not reciprocated in its most elaborated ritual form means that only parallel parents in the first ascending generation have the power to confer kinship on a junior with their familial touch. Secondly, it means that the sister’s son, bhaca, becomes the most respected of all kin. People say that he is “just like a Brahman.” But why is the cross nephew treated with extreme respect “just like a Brahmin?” Such are the questions that arise from a study of Bastar skinship but readers interested in the answer will have to await the publication of my forthcoming book because the answer requires a much more detailed investigation of the basic data than I have been able to give here. Estrangement The wonderful order the Bastar familial salute creates is extremely fragile. The above discussion presupposes that the gesture of familial respect always solicits a gesture of familial love. This is the key gesture because, by definition, it confers kinship by either affirming a pre-existing relationship of consanguinity or creating one anew as in the case of the samdhi bhent ritual at the marriage of one’s children. A touching gesture of familiarity presupposes familial love but if familial hate prevails the gesture may not be reciprocated. If the familial hate is mutual the initial gesture of familial respect may not be forthcoming, a sign of familial estrangement. This is a moral sentiment of great force and importance in the world of skinship. Familial hate in its most extreme form is expressed by the swing of a sword, a transcultural fact evidenced by the data on the large percentage of homicides within the family. Familial estrangement can be asymmetrical or mutual. The father who refuses to recognize the love-child of an affair as his child is an example of the former; a bitter dispute between brothers over the inheritance of their father’s property can provide fertile grounds for the development of the latter. Mutual familial estrangement is a form of skinship because it involves the creation of new gestures of avoidance that strive to negate sensible communication. When the brothers are forced to live in the same intimate space as neighbors in a divided house this becomes a highly developed and difficult art. Tactile interaction is relatively easy to avoid but great care has to be taken to avoid eye contact; smells and sounds are almost impossible to control, especially those bad smells and loud sounds sent deliberately to annoy. Then there is the problem of the wives and children of the quarrelling brothers, who are also enjoined not to communicate with each other as they roam the neighborhood playing, attend school together and bump into each other at the weekly market. My fieldnotes are full of examples of cases of this kind but this is something that is not peculiar to Bastar. Only a few moments reflection is necessary to unearth examples in one’s own family history and in those of friends from whatever country. What distinguishes Bastar is the existence of a special salute for ending the period of estrangement. The salute presupposes that one party has wronged the other and that they seek forgiveness. The guilty party lies prostrate on the ground in front of the wrong party touching his or her feet. Forgiveness is begged and usually always received if go-betweens have arranged the event properly (see Table 9). Table 9. Foot-touch component of the familial salute as a gesture of ultra-extreme humility Salutes for non-kin Salutes to non-kin assume a variety of types, many of which are distanced forms of a familial salute. Familial salutes are within touching distance, non-familial salutes are beyond it. In other words, relations of contiguity are the grounds upon which face-to-face salutes of different sensible kinds are located: the tongue and the hand are the primary means of expression in the intimate and personal zones of the familial salute; the voice, ears and eyes in the social and public zones of the non-familial salute. Table 10 illustrates just three of the many types of “untouchable” salutes in Bastar. It is obvious from visual inspection that these are distanced versions of some the familial salutes considered above. Table 10. Some distanced salutes used for greeting non-kin Competing values A common feature of these salutes, familial or otherwise, is the role of respect as the central value. In the intimate and personal zones familial respect is complemented by familial love but as the boundaries of skinship are reached, then breached, respect emerges as the supreme value. But here it faces competition from other values. Untouchability, from the perspective of skinship, is an expression of respect of some kind, be it mutual or ultra extreme. Other values can inform untouchability and ritual pollution is the classic example used when talking about India. For some indigenous Halbi-speaking people of Bastar this value is a temporary affliction that affects people such as menstruating women, but not whole communities in some permanent way. One woman I know was deeply insulted when a guest, who perceived herself to be of higher status, refused to accept water from her hand. The guest did it because of the “impure” status of her host. The host, for her part, interpreted the guest’s action as one of extreme disrespect, the height of bad manners, and felt that her dignity as a person was insulted. Respect is for her the dominant value when engaging in face-to-face sensible actions of a familial or convivial kind. Market-places relations, by contrast, are governed by different values which inform her actions as she tries to bargain down a seller’s price. The pursuit of market values does not compete with respect for her; indeed, good manners requires it to be done in a respectful, good humored way (unlike, it could be added, the lending practices of certain finance companies in the global economy today). Respect, then, is only one value among many but for some women in Bastar it is the dominant value. For moral philosophers like Kant (1797) respect, too, is a supreme value; nay more, it is the only moral sentiment his moral philosophy admits. For Kant respect is what makes us human. “The respect which I bear others or which another can claim from me,” says Kant (1797: 127) “is the acknowledgement of the dignity of another man, i.e., a worth which has no price, no equivalent for which the object of valuation could be exchanged. Judging something to have no worth is contempt.” This brings us back to Hall and Goffman on the defining characteristics of human sensible communication but ahead to a student of Kant’s thought, Sarah Buss, who has noted in her essay, Appearing respectful: the moral significance of manners (1999), that experts on manners have a “strikingly similar drum beat” when it comes to discussions of respect and human dignity. Manners, she argues, plays an essential role in moral life. Virtue, she notes, is essential to good manners. The study of systems of manners, then, is the study of morality, a thesis she argues that will probably strike “people uncorrupted by philosophy” as obvious. But the obvious needs to be stated now and then. Values inform face-to-face ritual interactions but there is more than one value. While the study of abstract moral philosophies of people such as Kant are no doubt important, both philosophers and anthropologists still have much to learn from the moral philosophers of the concrete found in places like Bastar. Ethnography theory has its origins in the concrete study of concrete problems but kinship theory, as the study of reference terms, has overlooked the obvious point that modes of address of a sensible kind are primary. If anthropologists are unsure about what kinship is then the people of Bastar have no doubts. For them it is first and foremost a relationship of contiguity rather than a relationship of consanguinity and affinity, a “mutuality in sensible being” to give Sahlins’ (2011) formulation a slight twist. They measure kinship distance in feet, by how far apart they are on a common ground, not by the number of consanguineal and affinal steps on a genealogy. Touching gestures of familial love and respect in Bastar can affirm relations of consanguinity but in most cases they create it anew; salutes also have in them the potential to render consanguineal relations asunder. Reciprocal recognition of familial love and respect is the basis of the former, mutual familial hate the basis of the latter. Kinship as “mutuality in being” makes senses as skinship but not as an abstraction from it. Skinship is a reality for them; its being is literally felt. Acknowledgements Thoughtful suggestions from two anonymous reviewers have enabled me to make substantial improvements to this essay. I am also grateful to Professor Akio Tanabe of Kyoto University and the graduate students of my 2004 “Skinship” seminar for introducing me to the wonderful world of Japanese skinship during my short visit to Japan. My debts to the people of Bastar are too numerous to mention but as I rely primarily on data on Maraar kinship in this essay I must single out Mr M.S. Mali who first introduced me to the equally wonderful world of Halbi skinship. Bibliography Bean, Susan S. 1978. Symbolic and pragmatic semantics: a Kannada system of address. Chicago: Chicago University Press. Buckminster Fuller, Richard & Edward J. Applewhite 1975. Synergetics: explorations in the geometry of thinking. London: Macmillan. Buss, Sarah. 1999. “Appearing respectful: the moral significance of manners.” Ethics 109: 795–826. Crystal, David (1987) 1991. The Cambridge encyclopedia of language. Cambridge: Cambridge University Press. Das, Sisir. K. 1968. “Forms of address and reference in Bengali.” Anthropological Linguistics 10: 19–31. Doniger, Wendy & Sudhir Kakar 2002. Kamasutra. Oxford: Oxford University Press. Eck, Diana L. (1981) 1996. Darsan: seeing the divine image in India. New York: Columbia University Press. Godelier, Maurice, Thomas R. Trautmann & Franklin E. Tjon Sie Fat 1998. Transformations of kinship. Washington: Smithsonian institution Press. Goffman, Erving. 1967. Interaction ritual. Chicago: Aldine. Haines-Stiles, Geoff & Peter Montagnon 1991. Childhood (7 Videos). USA: EBC. Hall, Edward T. 1968. “Proxemics.” Current Anthropology 9: 83–108. Hocart, Arthur M. 1952. The life-giving myth and other essays. London: Methuen. Hopkins, E. Washburn 1907. “The sniff-kiss in ancient india.” Journal of the American Oriental Society 28: 120–34. Inden, Ronald B. & Ralph W. Nicholas 1977. Kinship in Bengali culture. Chicago: Chicago University Press. Jakobson, Roman. 1960. “Closing comments: linguistics and poetics.” In Style in language, edited by T. A. Sebeok, 350–377. Cambridge Mass: MIT. Kant, Immanuel (1797) 1964. The metaphysical principles of virtue. Part II of The metaphysics of morals. New York: The Bobbs-Merrill Company. Kroeber, Alfred L. 1909. “Classificatory systems of relationship.” The Journal of the Royal Anthropological Institute of Great Britain and Ireland 39: 77–84. Mauss, Marcel. 2007. Manual of ethnography. London Berghahn. Mehrotra, R. R. 1977. “Fluidity in kinship terms of address in Hindi.” Anthropological Linguistics 19: 123–25. Montagu, Ashley. 1971. Touching: the human significance of the skin. New York: Columbia University Press. Morgan, Lewis H. (1871) 1997. Systems of consanguinity and affinity of the human family. Oosterhout: Anthropological Publications. Müller, F. Max 1879. The Upanishads IV: the Kaushitaki-Brahmana-Upanishad. Delhi: Motilal Banarsidass. Perella, Nicolas J. 1969. The kiss sacred and profane: an interpretative history of kiss symbolism and related religio-erotic themes. Berkeley: University of California Press. Pfeffer, Georg. 2001. “A ritual of revival among the Gadaba of Koraput.” In Jagannath revisted: studying society, religion and the state in Orissa, edited by H. Kulke & B. Schnepel. New Delhi: Manohar Phillips, Adam. 1993. On kissing, tickling, and being bored: psychoanalytic essays on the unexamined life. London: Faber and Faber. Sahlins, Marshall. 2011. “What kinship is (part one and part two).” Journal of the Royal Anthropological Institute 17: 2–19; 227–242. Skoda, Uwe. 2004. “Ritual friendship in a converging tribal and caste society.” Journal of Social Sciences 8: 167–77. Synnott, Anthony. 2005. “Handling children: to touch or not to touch?” In The Book of Touch, edited by C. Classen. Oxford: Berg. Vatuk, Sylvia. 1969. “Reference, address, and fictive kinship in urban north India.” Ethnology 8: 255–272. Wikipedia. 2007. Skinship. http://www.reference.com/browse/wiki/Skinship (Accessed 8 May 2007). Wordspy. 2007. Skinship. http://www.wordspy.com/words/skinship.asp (Accessed 5 April 2007). Skinship: la touchabilité comme valeur en Inde du Centre-Est Résumé : Dans « What kinship is », Marshall Sahlins (2011) offre une nouvelle et stimulante réponse à une question ancienne et très débattue, en définissant la parenté comme une « mutualité d’existence ». Pour les locuteurs Halbi du Plateau du Balstar de l’Inde du Centre-Est, la parenté est définie par le toucher: les jeunes honorent les anciens par des gestes tactiles de respect familial auxquels il est répondu par des gestes tactiles d’amour familial. Dans certaines occasions rituelles, ces salutations se parent de fleurs multicolores, de plats goÛteux, d’eau purifiée, d’encens aux parfums doux, de paroles agréables, et de sentiments sincères. Ceux qui ne sont pas parents, par contraste, sont définis par des gestes de respect mutuel qui ne sont pas tactiles. L’implication générale apportée par cet exemple pour l’étude de la parenté en tant que « mutualité d’existence sensible », pour détourner légèrement l’expression de Sahlins, implique un déplacement des études sur la parenté en tant qu’abstractions sémantiques de terminologies de référence vers une réflexion sur les pragmatiques en jeu dans les interactions sensibles de face à face entre individus. Peu de recherches ethnographiques ont été conduites sur celles-ci. Le mot « parenté de peau » (skinship), dérivé du japonais, par son évocation du toucher et de la parenté comme venant ensemble, engagerait une approche originale de l’analyse de la parenté. Chris Gregory has appointments in anthropology at the Australian National University and the University of Manchester. He has conducted fieldwork in india, Papua New Guinea and Fiji. His research has focused on the relationship between kinship, economy and religion. He is the author of Gifts and Commodities (1982), Observing the Economy (with J Altman 1989) and Savage Money (1997) and Lachmi Jagar: Gurumai Sukdai’s Story of the Bastar Rice Goddess (with H Vaishnav, 2003). ___________________ 1. This article contains a mixture of Halbi, Hindi, and Sanskrit words. I have decided not to use diacritics to avoid confusion for the general reader. For example, some Halbi vowels are long while Hindi vowels are short for the same word.

Sahlins: Twin-born with greatness This work is licensed under the Creative Commons | © Marshall Sahlins. Attribution-NonCommercial-NoDerivs 3.0 Unported. ISSN 2049-1115 (Online) Twin-born with greatness The dual kingship of Sparta Marshall Sahlins, University of Chicagowith the assistance of Philip Swift This article examines the comparative configurations of diarchy by means of an extended analysis of the Spartan dual kingship in ancient Greece. Twinned and inseparable, both human and divine, the Spartan kings were themselves descended from celestial twins, hence it is argued that the Spartan diarchy is an empirical instantiation of the king’s two bodies—the dual kingship as an expression of sovereign twinship. The essay goes on to consider other royal twins of Greek mythology, one of whom was usually descended from a god, and argues that such myths of dynastic origin constitute a cosmology of sovereign right in which the Spartan myth of stranger-kings of divine descent was opposed to the Athenian ideology of autochthony. Keywords: kingship, ancient Greece, Dumézil, dynastic succession, mytho-praxis The one phenomenon which remains a complete puzzle [in Sparta] is the survival of kingship, worse still, of a dual kingship. I have no explanation to put forward, but I will suggest that “survival” may not be the precisely correct word. Moses Finley (1981: 39) Is it conceivable that before the eunomia was established, or, as Herodotus would say, before Lycurgus instituted the gerousia and ephorate, Sparta was ruled by two kings, each having sovereign power over the state? We have all heard about kingdoms divided between two princes, or about co-regencies of father and son, but one can hardly visualize a single state ruled by two sovereigns. Robert Drews (1983: 81) L’historien ne peut pas donner la raison de ce partage de la royauté. L’attribuer à un calcul de politique est une pure hypothèse. Charles Daremberg (1904: 892) It may be useful to begin by putting the dual kingship of classical Sparta into a comparative frame. If it has analogies elsewhere, or better if it appears as a variant of more familiar systems of sovereignty, then it begins to shed its enigmas. Dual kingships are found in many different civilizations and in a variety of forms. In virtually all diarchies, one or the other king is superior by virtue of a closer relation to divinity; but otherwise the kingship varies according to two distinct principles of sovereign dualism. The two kings are sometimes different and complementary in function; or else they are in-alike, sharing the same powers. Let us call the first a complementary or asymmetrical diarchy, referring in this way to the organic division of the sovereign powers, as between a war-king and a sacerdotal or peace-king. As they differ qualitatively, each is supreme in his own function. But in diarchies of symmetrical form, the one king is the functional image or twin of the other. If differing in rank, they are in all other respects the same in privilege, as they are in sovereign function. This is the Spartan kingship, admittedly more rare than complementary diarchies. Yet both types can be found, in diverse institutional expressions, among the ancient Indo-European peoples. The Romans knew both, though they put only one into practice. In Roman kingship traditions everything happens as if the Spartan concept of twinned rulers were consciously rejected in favor of a complementary dualism. According to the well-know legend, a functional division of sovereignty was introduced in Rome precisely through the failure of the joint rule of twins, Romulus and Remus -failure of the kind of diarchy that in Sparta was the beginning of dynastic wisdom. Sired by Mars, Romulus and Remus had “an unsociable love of rule” (Dion. Hal. Rom. Ant. 1. 85). When sent off from Alba by their royal grandfather, together with his own rebellious subjects, the twins divided their party in two with the intention of stimulating a useful rivalry. But the effects were ultimately fratricidal: fatal to Remus and to the project of a twinned kingship. Instead, Rome was founded by a combination of peoples of different qualities, the militant Latin invaders and the (re)productive Sabine aboriginals, whose respective kings Romulus and Tatius initially shared power as joint rulers. Thereafter, Rome would be alternately governed by kings of the violent Latin type and the more judicious Sabine type—in Dumézil’s terms, celeritas kings and gravitas kings—who thus incarnated the cultural dispositions of the two founding peoples. As the tradition goes, Romulus presumably killed his co-king Tatius (according to Livy 1.10), and then himself disappeared without issue. The kingship devolved upon the Sabine, Numa, whose reasoned and ritualized reign makes a strong contrast with the “sacred violence” of Romulus. Unlike Romulus, whose creative acts of sovereignty included rape, ritual sacrilege (falsification of the auguries), fratricide, regicide, and a thirst for conquests, Numa weaned Rome from war and instituted the cults of order and prosperity. Numa was succeeded by the warlike Tullus Hostilius, the latter by the peaceable Ancus, and so character of kingship alternated to the second Tarquin and the end of the monarchy. For Dumézil, the contrast between Romulus and Numa in particular was characteristic of the complementary dualism of Indo-European sovereignty (1949: 143-59). In various writings (1948; 1949, etc.), Dumézil develops this contrast between the magical war-king and the judicial peace-king, between celeritas and gravitas, by means of a series of correlated oppositions. The two types are contrasted as sacred force is opposed to reasoned order, youthful warrior to venerable legislator, will to intelligence, act to decision, and other-worldliness to this-worldliness (or divine to human). For Rome, Romulus and Numa are the prototypes; but Dumézil finds the most general expression of this complementary dualism in another realm, the famous couple of Indic sovereign gods Mitra and Varuna, even as the same distinctions may be discovered in other systems of complementary rule, as between the kshatriya and the brahmin. And there is still another kind of variation: the different ways such dualism is institutionalized as kingship. The complementary powers may or may not be realized in diarchy, a double kingship. In the instance of Romulus and Tatius presumably they are; and one may add less equivocal Indo-European examples such as the Celtic chieftain and tanist. (Of course, there are numerous non-Indo-European analogues, as the sacred and active ruling chiefs of Polynesia, or emperor and shogun in Japan.) But in Rome from the time of Numa, there was but one king, and the two modes of sovereignty appeared in alternation, over time. Moreover, it would be easy to round out the set of permutations by instances of a unique king who synthesizes the creative violence and constitutional order of sovereignty in his own person. Descended of the Mayors of the Palace, whom Pirenne once styled the shoguns of the Merovingian king, Charlemagne as rex francorum was particularly endowed to so embody the royal duality. The Roman and related traditions thus become instructive on several scores. First, even if it were only imagined in the native ideology, the change from a twinned kingship of the Spartan sort to a complementary dualism confirms that we are indeed dealing with a family of related structures. Asymmetrical and symmetrical diarchies belong to the same structural universe, as conceivable—and historically possible—transformations. Second, a distinction needs to be made between sovereign dualism as a structural principle and the manner in which it is institutionally expressed, the actual configuration of kingship. The same complementary opposition of royal powers may be variously manifest in a system of two kings who differ in function, in a dynastic succession of unique kings alternating in character, or at one and the same time in the duplex political being of an exclusive monarch. And finally, the Dumézilian parallels between ruling kings and sovereign gods—Numa : Romulus : Fides : Jupiter :: Mitra : Varuna, etc.—suggests a third conclusion, which will go a long way towards situating the symmetrical kingship of Sparta in the structural group at issue. There are two distinct principles of sovereign dualism: duality of the sovereign person, and duality of sovereign powers. Their intersection makes up the structural group. The complementary distribution of powers between two rulers is one thing; another is the doubling of man and god entailed in concepts of divine kingship. The first is a political division of labor, a functional dualism, but the second is an ontological principle. The divine king is in some sense the double—the living form, the earthly successor, or the incarnation—of the sovereign god. Metaphysics differ: if we can again bring in comparative examples, Maori say that the sacred ruler (ariki is the “resting place” of the god; Fijians, that the king is a “man god.” We can resume the variations by saying that the divine king is a twinned being; he is “twin-born with greatness.” This dédoublement may easily escape notice as a dual kingship proper, insofar as the king’s two natures are combined in a single royal person. But if we can overcome our own dualistic prejudices about mind and matter, spirit and body, and privilege rather the structural principle over the institutional form, then the famous medieval doctrine of the king’s two bodies finds its place in the structural set. And the Spartan dual kingship then appears as a humanized version of the twinned sovereignty, an empirical expression of the king’s divinity by a mortal doubling of the royal person. Here it is simply that the king’s two bodies are tangibly present to experience. In this perspective, the theological speculations of the Norman Anonymous (c. 1100 CE), cited by Kantorowicz, become paradigmatic: We thus have to recognize [in the king] a twin person, one descending from nature, the other from grace....One through which, by the condition of nature, he conformed with other men; another through which, by the eminence of [his] deification and by the power of the sacrament [of consecration], he excelled all others. Concerning one personality, he was, by nature, an individual man; concerning his other personality, he was, by grace, a Christus, that is, a God-man (Kantorowicz 1957: 46; original emphasis). The Norman Anonymous, moreover, goes on to explain that this christomimetes entails a dual ontology of both man and god, a metaphysical point also pertinent to the twinned kinship of Sparta (as we shall see). The god is (was) also a man, as the man is also a god—thus proving the interchangeability of the mortal and immortal sovereignties. Kantorowicz comments: He [the king], the anointed by grace, parallels as a gemina persona the two-natured Christ. It is the medieval idea of Christ-centered kingship carried to an extreme rarely encountered in the West. The king is a twinned being, human and divine, just like the God-man. . . . (1957: 49). In the same way that the correlated dualism of celeritas and gravitas on the axis of complementary powers is realized in several configurations of monarchy, and not always in diarchy, so this principle of the king’s two bodies varies in social expression. Nor are the representations of divine mimesis so far noticed—the two-in-one (medieval Europe) and the one-in-two (classical Sparta)—the only possible permutations of a twinned sovereignty. Where the king is the living image of the god, the whole field of iconography is potentially in play: all the parallel figurations of the god as an empirical image (e.g., as in stone or wood) associated with king and kingship. The subset is complex, ranging from the doubling of the king in an icon of the sovereign god—the former metonymically as well as metaphorically identified with the latter through the mediations of sacrifice—to the exclusive rule of a sovereign image standing in for the king. Such image-inations were unsuppressible even in a Christianized Europe: not only in the commonplace figurations of Christ as the celestial monarch, but in extremis the replacement of the defunct king by a wooden image to whom all royal honors were accorded. The deceased Francois I was so iconically “impersonated” for over ten days (Kantorowicz 1957: 425-26)—as it were, a perfectly logical inversion of the king’s two bodies, the natural form of the divinity endowed by human grace with the spiritual guise of royalty. In Sparta likewise, if a king died in battle, the funerary honors were paid to a statue of him (Herod. 6. 58). When situated thus in a comparative field, the dual kingship of Sparta begins to lose its strangeness. Indeed, if we are prepared to so enlarge the perspective, it becomes possible to claim that not even was the mortal doubling of the king unique to Sparta. The substitution of a human alter-image of the ruler is a common feature of world (and kingship) renewal rituals. In these rites of cosmic rebirth, including the Saturnalia and its carnivalesque reflexes, the king’s double acts often as the sacrifier and characteristically as the divine victim. But we need not rehearse here the whole The golden bough. The point is that Spartan diarchy is intelligible as a permanent human instantiation of the king’s two bodies, with the same sense of a divine legitimation. Since Aristotle’s reference to the Spartan belief that two kings made for the stability of the state (Pol 1271a26), scholars have sought the raison d être of the diarchy in its supposed functional or real-political values. (The common invocation of “survival” can be included here, as this is merely the limiting case of the same paradigm, i.e., non-functionality.) It is speculated, for example, that inasmuch as the two kings were a single legal person Sparta originally had one king, and then invented the second as a curb to individual ambition (Hooker 1980: 121). But how did Sparta alone come to this inventive solution? The Spartan kingship seems indeed more a problem of intelligibility—“the one phenomenon which remains a complete puzzle” (Finley 1981: 39)—than of functionality, except insofar as the latter also entails legitimacy. Even then, we shall see, the meaning of the dual kingship has more to do with Sparta’s historical pretensions to supremacy in Hellas than with her internal problems of statecraft. On the other hand, neither can the Spartan diarchy be resolved into a normal Indo-European form of complementary sovereignty. Not Mitra and Varuna, but Castor and Polydeuces, who were alike and inseparable. Equal in privilege and identical in function, themselves descended from an original pair of royal twins—which in Greek mythology generally is the sign of a double fatherhood, divine and human—the two Spartan kings by their very resemblance proved they were indeed born of Zeus, hence uniquely entitled to hold the sceptre in the Peloponnese. Nature of the Spartan dual kingship The Spartan king, says Thucydides (in the singular, but it was true of both), is “the seed of the demigod son of Zeus” (5.16.2). The allusion is to Heracles. With the “Dorian invasion” and the elimination of the Atreids—the house of Agamemnon and Menelaus, which had held the sceptre of Zeus—the Heraclids were the last Greek royalty that could claim power by devolution from the Olympian sovereign. Indeed they alone could have been the last to do so by direct patrilineal descent, since Heracles’ mother Alcmene was the last mortal woman with whom Zeus lay. With the destruction or demise of the collateral branches of Heraclids in Messenia and Argos, the Spartan rulers by c. 600 BCE were the only surviving Zeus-born lineage in the Peloponnese. And Macedonian claims to one side, by the beginning of the classical period the Spartan kings were the sole blood heirs in all of Greece to the Zeusian sovereignty. Hence they were, “of all men the most blue-blooded” (Ste. Croix 1972: 139; cf., Isocr. Ep. 9.3). We are not (of course) taking the mythological and genealogical ground of the foregoing statements as “true history.” But there can be no doubt that this mythopoetic consciousness of Spartan kingship was alive and well in the political life of the Hellenes through the Peloponnesian war and beyond. This suggests that the theory and practice of kingship in Sparta have to be understood at least as much from her external relations to other states—most notably her long-standing projects of domination—as from the internal relations of the Laconian polity. By the same token, the myths and rituals of Spartan kingship become historical “truths,” at least as relevant as the kings’ “factual” powers. Three general characteristics of the Spartan diarchy stand out in the descriptions left by Herodotus, Xenophon, Plutarch, and the ancillary standard sources: (1) The divinity of the two kings: their exclusive association—and in certain respects, identification—with the sovereign god Zeus, and with the Dioscuri, Castor and Polydeuces; (2) The universal scope of sovereignty: sacrificial mediators between culture and cosmos, the kings’ own lives were ritually and politically identified with the life of society; marked by wealth, their powers also included judicial and sacerdotal aspects, with a special emphasis on the external-military or protective functions of Zeus and the Dioscuri; (3) The symmetrical or twinned nature of sovereignty: the two kings were only minimally differentiated by descent while otherwise the same in privilege, and acted officially in concert. These dimensions of Spartan dual kingship are interrelated; each is testimony to the others. In the discussion to follow, they are only nominally taken up in order, it being impossible to separate them absolutely. Each of the Spartan kings is a double being. At all public feasts they received a double share of everything. Herodotus distinguished the religious occasions involving consumption of the sacrifice, where the kings are served first “and twice as much of every dish as everyone else,” from ordinary state dinners, where they are likewise “served with double quantities”—adding that they again enjoy the same privilege at private dinners (6.57). (Some of this may already be a reference to the Dioscuri, as: “The real guests at the entertaining of gods, theoxenia, are the Dioskouroi. They are celebrated above all in the Dorian area, in Sparta” [Burkert 1985: 107]). Xenophon commented that the lawgiver (Lycurgus) accorded the kings such privilege not that they might eat twice as much as others, but so that they could honor whomever they pleased (Rep. Lac. 15.4). While taking note of the important point that the Spartans in this way and others gave their kings the means to appear as sources of largesse, such practical values do not explain the matter, since the kings also acted as two-fold persons in their legislative capacities. If absent from the gerousia a king is represented by his nearest kinsman among the elders, who is thus entitled to cast two votes in addition to his own (Herod. 6. 57; our emphasis). Nor would the intention to honor the king account specifically for the two-fold nature of the respects, as opposed to some other (and larger) numerical sign of the people’s esteem, unless what was thus being recognized was the double nature of the royal person, part mortal and in part divine. Such issues are ignored by the familiar sociological observation—of which Xenophon’s is an early example—that the double honors accorded the kings have the value simply of a status distinction, marking thereby the exalted position of royalty. We have to examine the precise content of such honors, to see just what is being signified. Benveniste offers a relevant analysis of the pertinent term geras: the “honor” or the “honorific part” (as of booty) allotted to kings (basileis) in Homeric literature (Benveniste 1969: 43-50). He cites in this connection Herodotus’ description of the Spartan kings’ double portions, as well as their places of honor at public games, care of the oracles, rights to the victims of animal sacrifice, and the like. “Each term [of Herodotus’ description] seems to be made to illustrate a Homeric text” (1969: 47). Yet in certain of the Homeric texts adduced by Benveniste to support the point that geras refers distinctively to royal dues, the privileges in fact are being offered to the gods, in sacrifice. Benveniste fails to comment on this equivalence between royalty and divinity as destinaire of the geras. Yet the equivalence concerns even the parts of the sacrificial animal. In the Homeric hymn to Hermes (122, 128-129), the divine messenger divides the slaughtered cattle into twelve portions as offerings to the gods, adding to each the honorable part or geras. The geras in question is the chine or back-portion, which is precisely the portion of the animal reserved: (1) to the Homeric king (Od. 4.65-66); and (2) to the two kings of classical Sparta (Herod. 6.56). “Why do they honor us like gods?” asks the Lycian king Sarpedon, referring to the prestations (gera) in cuts of meat, seats of honor, lands, and the like given to himself and his co-ruler Glaucus (II. 12.310f). Once more the honors in question fit the Spartan kings—as do Sarpedon and Glaucus, on which more anon. There is something more to the sacrifice. Xenophon tells that in the compact made by Lycurgus with the king(s), the lawgiver “ordained that the King shall offer all the public sacrifices on behalf of the state, in virtue of his divine descent” (Rep. Lac. 15.2; our emphasis). Apart from the usual (Maussian) association of sacrifier (and sacrificer) with the god through the mediation of a victim consecrated to the latter and identified with the former, the Spartan kings had an a priori suitability for the priestly office by virtue of their consubstantiality with Zeus: i.e., by descent. Thus they detained in particular the priesthoods of the celestial-encompassing and terrestrial-localized aspects of the god, Zeus Uranius and Zeus Lacedaemonius (Herod. 6.56; were these two priesthoods respectively held by the senior and junior branches of the Heraclid lineage?). Moreover, according to Herodotus, the kings were allotted the skins and chines of all animals offered in sacrifice (6.56-57; cf. Xen. Rep. Lac. 15.3). One cannot be certain but the statement appears to cover the offerings made by others, on occasions at which the kings were not present or officiating. At any event, the same significance attaches to the double portions due the kings at sacrificial feasts: the kings were not sacrifiers merely, but in such rites played the role of the god, the one who consumes the offerings. There is nothing unusual or distinctively Spartan in the fact that men ritually consume the sacrifice, thus partaking of the divine benefits of the consecrated victim in a commensality with the god. But if the gera of the Spartan kings is testimony to their double being—i.e., being that includes a divine nature—then their double portion of the sacrificial offering makes the feast which follows more than a figurative communion with an unseen god. The sacrificial feast is an empirical communion partaken with existentially present gods, the kings. Hence the fact, already noticed, that the Dioscuri were the honored guests at the Dorian theoxenia, the “ubiquitous feasting of the gods.” For the Dioscuri, we shall see, were the alter-images of the Spartan kings, even as the term means literally “sons of Zeus.” The texts on royal privileges in Sparta allude to several corollary intimations of Zeus. For example, in battle the kings were accompanied by a bodyguard of one hundred men, chosen from Sparta’s finest—thus the “hundred-handed ones,” offspring of Uranus, whom the latter had cast into the nether regions of Tartarus, whence they were rescued by Zeus and became his fighting allies in the war with the Titans (Apollod. Bibl. 1.1–2). And terrible strength was in their mighty forms. (Hes. Theog. 153) The Spartan bodyguard would sooner suffer their own disgrace than allow the king to meet death (Isocr. Ep. 2.6). Likewise for any soldier it was a greater dishonor to fail to sacrifice oneself for the king than to throw away one’s arms (Isocr. de Pace 143). Even the enemy, “fearful of spilling the blood of a Spartan king,” would generally tend to avoid direct combat with him (Plut. Agis 21). So within Sparta the king’s person was inviolable. Until the murder of Agis IV (241 BCE), and as documented even in Plutarch’s description of the deed, it was hardly conceivable for anyone to lay violent hands on the king (Plut. Agis 19.9). A human modality, all this, of divine immortality—indicative also of the cosmic significance of the king’s life for the existence of society. Also corollary: the king’s body was (in principle) without blemish, and his conduct without blame. One judges from the arguments over the accession of the lame Agesilaus (early fourth century BCE) that a physical defect in a potential heir was a disqualification, albeit in this instance not definitive. Especially inauspicious would be the defect of lameness, as it is the mythical sign of the earth-bound and earth born (cf. Lévi-Strauss 1977: 215-16)—hardly the pride of the immigrant cum celestial dynasty of Dorian Sparta, however much it might be honored in “autochthonous Athens.” The faultlessness of the king was bound up with his role as sacrifier for the kingdom, thus with his metonymic connection to the god established through a victim that was likewise without blemish. But then, the blamelessness of a king would extend to his political conduct, on pain of incurring divine wrath. The effect is a Frazerian relation between the king’s goodness and the welfare of the state, to the extent that the king becomes politically accountable and subject to removal by other powers-that-be. Every ninth year, reports Plutarch, the Spartan ephors consulted the celestial signs regarding the king’s conduct, and “if they chance to see a shooting star, they presently pronounce their king guilty of some offence against the gods, and thereupon he is immediately suspended from all exercise of regal power, till he is relieved by an oracle from Delphi or Olympia” (Agis 11.3). In a celebrated incident, Lysander as ephor claimed to have seen the inauspicious sign, temporarily bringing about the deposition of the King Leonidas, whose faults were that he had violated the ancient laws forbidding any of the royal blood of Heracles to settle in a foreign country or to sire children by a foreign woman (Agis 11.4). Speaking of Frazer, the practice can be considered a Greek version of famous ideologies of regicide, the doing-in of the failing ruler, based on the same correspondence between the king’s perfection and the society’s eunomia, but here modified by the inviolability of the royal person. All this has bearing on current scholarly debates about the political powers of historical kings in relation to the other organs of rule in the Spartan polity: the ephors, the gerousia and the assembly (e.g., Drews 1983: 78-85; Ste. Croix 1972: 138-48). The issue is whether kingship was mainly (or merely) “symbolic,” or else a real force in the decisions and affairs of state. Dispute is joined over the known or presumed influence of kings such as Agesilaus, and what this might imply about more obscure rulers and the kingship in general. We would not here enter the lists, except to say that, clearly, some did and some didn’t. Some (as Cleomenes I), did exercise considerable authority and had significant historical effect; others (as Leónidas II) could not even prevent their own deposition, banishment, recall from military command or liability to legal censure and penalty. Yet this variability, as it cannot be accounted for from constitutional principles, itself requires explanation. And what has been said here about the king’s sacred status may offer some contribution. For in the correspondence between the existence of the king and the well-being of society, there is room both for the ruler’s exercise of charismatic leadership and his political neutralization. Both would come from this very correspondence, with its implications concerning divine favor and divine wrath, which a gifted king might mobilize to take command but rival forces could invoke to hold the king responsible. Embodying the polity, the king could take it in charge, but he was also then accountable. Otherwise neither the constitutional division of powers nor the presence (or absence) of outstanding personalities will be sufficient to explain the known variability of royal authority. Only within the encompassing doctrine of the king’s divine powers are such factors given license to determine the de facto distribution of political powers. It follows that this distribution, as between the kings and others, will always be contingent and contextual, i.e. historical. Operative thus in history and event, the structures of sovereignty are periodically subject to contingencies that, by the correspondence doctrine, become crises of major proportions: the death of the kings. In principle, society dies with the king, to be reborn with the installation of a successor, which is what Herodotus describes in the royal mortuary rituals of Sparta (6.58). Herodotus likened the unusual rites to barbaric practices, specifically to Asian (Persian) models (6.59). Perhaps significant historically, the observation is surely acceptable anthropologically, since the ceremonies are of the kind known to divine kingships worldwide. The particulars of the funerary rites given by Herodotus are few, even peculiar; but in this wider context they become intelligible. At any rate, our interpretations can be confirmed by the parallels between the signs of royal power displayed at death and the documented functions of Spartan kings in life. The main interpretations of Herodotus’ description of the funerary ceremonies are as follows: (1) The catastrophe of the king’s death is cosmic or universal, which is to say in political terms, imperial. Herodotus is careful to note that it not only concerns Spartans, but helots and country people (perioikoi, “dwellers around”) from all over Laconia. A number of these conquered and outlander peoples are obligatorily required to attend the ceremonies, mourning intensively alongside Spartans proper in a “huge crowd” of many thousands. This is one of the similarities to the death ceremonies of “Asian” kings. It speaks to a comparable character of Spartan sovereignty: its universalist pretensions, like the Persian “king of kings.” Remarkable from an Hellenic point of view, this quality of Spartan kingship is not unrelated to its other “anomalies,” such as the attempt to make good by myth and war its unique claim to the succession of Zeus. (2) The catastrophe involves the fertility or reproduction of the social order. Hence the curious stipulation that two people from each household, a man and a woman—i.e., a reproductive couple—put on mourning, on pain of heavy fine. Herodotus also observes that “men and women together” strike their foreheads and otherwise display their grief at the ceremonies proper. The implication is that the king is the condition of the people’s fecundity and prosperity. (3) Another implication is that society dies with the demise of the king. Smiting their foreheads, while wailing and singing the praises of the dead king, the people inflict punishment upon themselves, in this way symbolically sharing the fate of the sovereign. But the suspension of the society has also a direct representation: for ten days after the funeral, only mourning prevails; all public meetings and elections—the normal affairs of society—are prohibited. Without the king, no social order, only the disordered excesses of grief. (4) The new king, upon installation, re-establishes social life as a new beginning. “When a new king comes to the throne on the death of his predecessor, he follows a custom which obtains in Persia on similar occasions: he remits, that is, all debts owed by Spartan citizens either to the king or to the treasury. This corresponds with the Persian custom whereby a king, on his accession, remits arrears of tribute from all his subject states” (Herod. 6.59). The installation of the new king is in this way the logical complement of his predecessor’s death rites: if society dies in the latter, it is reborn in the former. What we say of Herodotus’ text is generally supported by Xenophon. He remarked that the funerals of Spartan rulers are more like those of heroes (demigods) than of men (Rep. Lac. 15.9). And of the death ceremonies of Agis II (c. 401 BCE), they were “more solemn than belongs to a man” (Hell. 3.3.1). Our interpretations might even so seem far-fetched, were they not also supported by the privileges and powers attested for living kings. The mortuary ceremonies are the negative imprint only of the sovereignty; on the other side of the coin are the positive attributes of Spartan kings. Just as the funerals depict the kings as the condition of fertility and social reproduction, so the ruler in his style of life as well as judicial offices is specifically associated with fecundity, both sexual and proprietary. Plato says that the kings were the richest men in Sparta (Alc. 122c-3a). True or not, this is the impression the Spartans would manage by the economic prerogatives accorded their rulers: as many cattle as they wish, choice lands in perioikoi districts, their own lake near their house, not to neglect the double portions. Again the commentators (Herodotus, Xenophon, etc.) understand these arrangements as practical means of ensuring that the kings will be able to discharge their duties as sacrifiers, as well as do honors to others. But they also mean that the kings will appear the embodiment of wealth by their property and the fount of largesse by their generosity. These material expressions amount to the same thing as the kings’ sacrificial function, the finality of which is likewise the general prosperity. The sovereigns’ peacetime judicial rights are again similar. Surely they are restricted by comparison to the legal prerogatives of other organs of state. But if reduced to a minimum the kings’ own magisterial roles are a significant—i.e., a signifying—minimum. The two kings were juridically concerned with familial reproduction and the transmission of estates, that is, in default of the normal domestic mechanisms of continuity. They selected spouses for unbetrothed women who had inherited estates, and attested to adoptions. Besides, the kings ruled legally on all matters concerning public roads, another reference of their relation to the social totality. We have mentioned their charge of the public cult and their command in war, again real-political expressions of the cosmic sovereignty represented in the royal rituals of death. Without denying that all this may leave the kings with only limited constitutional powers, warfare perhaps excepted, it should be noted that their powers are in kind universal. In Dumezilian terms the kings are both celeritas and gravitas: they exercise both functions of Indo-European sovereignty, judicial-sacerdotal and external-military. Or else, their dominion spans all three functions: priest, warrior and producer (source of wealth). And as both kings detain the same functions in all domains, the Spartan diarchy is not complementary. It is a universal and symmetrical sovereignty, encompassing all aspects of social life, with a special emphasis in the protective-military dimension—hence specially universal or encompassing in relation to outer nature, other people, and rival kings. On military campaigns, the king(s) take(s) command. The sovereign leads on the march, issues the orders of the day, decides the places of encampment and (in principle) battle. The strategist in war, the king’s judicial powers are also total in this context: he holds court on all matters of dispute, booty, etc. arising on campaigns. Likewise for his sacrificial offices. Before the army departs, the king makes offerings for success to Zeus “and the gods associated with him” (identified by Marchant [1925: 178-79, note 1] as Castor and Polydeuces). At the borders of Spartan territory he sacrifices again in order that the army may cross over into foreign land—if the sacrifice is unsuccessful, they all go home—and he makes the offerings too before battle. Regarding these sacerdotal activities, we underscore several features. First, the king rules transgressions of the border: he is the condition of the possibility of a Spartan imperial presence in Greece and beyond. This is once more akin to the funerary rites: the kingship, and only the kingship, and only the Spartan kingship has transcendent powers. Like Zeus. But also like the celestial twins, the Dioscuri: in battle, the kings particularly take part of Castor and Polydeuces, and vice versa. This relationship merits some comment. The divine twins Castor and Polydeuces were kinsmen of the Spartan kings, at once through Zeus and in the human line of Heraclidae. The guiding stars of Spartan arms, who disappeared before the defeat of Leuctra (Paus. 4.26-27.3) and appeared upon Lysander’s victory at Aegospotami (Plut. Lys. 12.1), the Dioscuri were also physically present in the Spartan host, in direct association with the Spartan kings. Their image accompanied the kings in battle: the dokana, a double cross-piece of two vertical and two horizontal staves (cf. Waites 1919; also Plut. De fat. amore 478a). Apparently a double-barred figure-H (#), the dokana, by its celestial and terrestrial orientations, seems a perfect icon of the dualities involved. After the falling out of Cleomenes I and Demaratus in campaign against the Athenians (c. 506 BCE), it was made illegal for the kings to take the field together. The image of the Dioscuri was then divided between the one who stayed in Sparta and the other who led the army. The Dioscuri, however, were not merely protectors or tutelaries of the kings; in war the kings played the role of their divine counterparts, most particularly of Castor, who was the son of a human father and the strategist. When their army was drawn up in battle array, and the enemy near, the king sacrificed a goat, commanded the soldiers to set garlands upon their heads, and the piper to play the tune of the hymn to Castor, and himself began the paean of advance (Plut. Lyc .22.2–3; n.b., the composition of the war paean was attributed to Castor). Protected by twins, identified metaphorically with twins, the Heraclid dynasty of Sparta was born of twins (fig. 1). These were Procles and Eurysthenes, the sons of Aristodemus, a direct descendant of Heracles (in the fifth generation). Aristodemus had been heir by lot to Sparta in the so-called Heraclid return. According to the usual account, however, he died before the conquest; the twins, ruling jointly, were the beginning of the dynasty. Herodotus retails another version which, he says, the Spartans tell themselves, differing by the assertion that Aristodemus did reign over Laconia, only to die shortly after the birth of his sons and successors. The apparently minor difference in fact encodes another claim to legitimacy, apart from conquest, and we shall return to it presently. Here we would complete the general discussion of Spartan kingship by remarks on its character as a symmetrical diarchy. Figure 1. Genealogy of the royal Spartan twins. Rather than a complementary dualism of sovereign functions, the Spartan is a mirror kingship. The kings are mirror images of each other in action as they were in origin. The original twins, Procles and Eurysthenes, Herodotus recounts (6.52), could not even be told apart: “they were both the same size and each exactly like the other.” Nor could (or rather, would) their mother tell the difference, thus frustrating the Spartans’ desire to make the elder one the king. The Delphic oracle advised that both be kings, allowing the elder only the greater “honour.” Still stymied by this, the Spartans on the suggestion of a Messenian man watched the mother—who in fact knew the difference well enough—to see which one she fed and washed first. So they were able to determine that Eurysthenes was the elder and had him brought up at the public expense. His descendants thenceforth ruled as the senior line, the Agiadae, but jointly with the cadet branch from Procles, the Eurypontidae. The kingship was minimally differentiated. It even began in twin mothers as well as twin fathers: Procles and Eurysthenes married twin sisters, Lathria and Anaxandra, who also descended (in the paternal line) from Heracles (Paus. 3.16.6). Thereafter, the single important difference between their respective descendants was seniority in the Heraclid line. However, this is one of those differences that make a difference, a distinctive feature (structurally speaking) signifying “greater honor,” which is to say in Greek closer to god. On the other hand, since the two kings are otherwise identical, the sequitur must be that the god = the man. In this mirror sovereignty, each king not only sees himself in the other, but a reflection of divinity. This helps account for the Spartans’ insistence, throughout the history of the dynasty, that there must be two kings, and that they must respectively derive (in the paternal line) from the two Heraclid branches, the Agiads and the Eurypontids. If a king died leaving a minor heir, a regent was appointed from among his close kin. If a king were dethroned, he was succeeded by another from the same house. The one exception is relatively late and of the kind that “proves the rule.” Cleomenes III (236-222 BCE) attempted to make his own brother Euclidas partner in the throne (Plut. Cleom. 11.3). But precisely this was part of Cleomenes’ vain project to revive the ancien régime: he would begin the kingship anew. Twinned by origin, the two kings were indistinguishable in authority and in action. Except that one was the priest of Zeus Uranius and the other of Zeus Lacedaemonius, no distinction appears in the texts in their sovereign powers, whether juridical, sacerdotal or military. Nor is there any indication that one was superior to the other in social privilege or political authority. Until the rule that only one could take the field at a time, they appear to have done everything official together; they even lived and messed together in the same tent (as suskenoi, Xen. Hell. 5.3.20; Rep. Lac. 15.5; Plut. Ages. 20.5). Herodotus surely exaggerated when he said that the descendants of Procles and Eurysthenes quarrelled ever after, as those two did all their lives (6.52; cf. Ste. Croix 1972: 140). Even when they were at odds politically, as were Agesipolis I (395-380 BCE) and Agesilaus II (400-360 BCE), they could remain intimate personally. One might have expected, says Xenophon, that Agesilaus would have been pleased when he heard of Agesipolis’ death, as one is at the death of a rival, but in fact he wept and mourned for the loss of a comrade; for, of course, the Spartan kings mess together when they are at Sparta. And in all their conversations about their young days, hunting, horsemanship or love affairs, Agesipolis was excellent company for Agesilaus. He also treated him, as the elder man, with becoming respect in all relations which arose out of their shared quarters (Hell. 5.3.20). The constitutional necessity that there be two kings, of two lines, sharing both public functions and private lives, implies that they were a single royal person. Herodotus (6.50) recounts certain incidents involving the disposition of hostages in which the other cities concerned, Aegina and Athens, refused to comply with Spartan demands on grounds that they had been made by one king only; whereas, admitted the Aeginetans and the Athenians, they would have been liable if the Spartan kings had acted jointly. Ste. Croix’s probable conclusion about these episodes seems of more general applicability: in principle, only on the condition that they acted in concert could the king represent the Spartan state (1972: 150-51). Just what did this principle mean? And why were the Spartans so tenacious of it? Certain answers, we shall argue, can be deciphered from the precedents and paradigms of Hellenic myth. Myth tells that the Spartan kings were not the first such twinned rulers in Hellas. At the same time, these ancient mythic prototypes give the sense of Spartan historical practices. For they organize the historical experience and practice of kingship. The mytho-praxis of dual kingship We take Spartan kinship as a good instance of what has been called “mytho-praxis”: the projection of the relationships of cosmological myth onward into historical action (cf. Sahlins 1981, 1985). This is not to prejudge whether any such relationship found in myth is “historically true,” wholly or partially. Nor do we allege that the Spartan kingship is the mere “survival” of Homeric or other venerable prototypes, without functional significance. Our understanding is that the mythical tradition of dual sovereignty was selectively and uniquely elaborated in Sparta by an unfolding relation to the historical conjuncture. It was a claim, based on ancient pan-Hellenic authority, to the legacy of “Dorian,” and beyond that “Mycenaean,” dominance in the Peloponnese. And as the Mycenaean (or Achaean) kingship had even greater pretensions, the same claims could be made, as the occasion arose, to leadership against the Persians and superiority over earth-born Athens. Dynastic succession in the Peloponnese Sparta’s pretensions as against Athens, the claims of the allochthonous people over the autochthonous, invoke a theory of sovereignty widespread in ancient Hellas, not to mention Indo-European peoples generally (cf. Sahlins 1985, chap. 3). The concept was especially marked in the Peloponnese. The so-called Dorian conquest was the last only in a series of dynastic successions of the same type. Nor does it exaggerate to say that in terms of the categories at issue the prototype was Cronus’ emasculation of the Sky (Uranus) and appropriation of the fruits (the daughter, Rhea) of the Earth (Gaia). Zeus immediately followed with a repetition of such exploits at the expense of his father Cronus, whence the universal domination of the Olympian sovereign god. What marks the Peloponnese in general and Sparta in particular is the claim of a sovereign devolution from Zeus, hence a parallel human hegemony over other kings and peoples, whether born of the earth or of other heroic ancestry. Cosmogony is translated into an epic tale of dynastic succession: the advent of a stranger king of violent dispositions and Zeusian antecedents, who typically marries the daughter of an earlier or indigenous ruler, assassinates the latter, and so gains the kingdom. Such usurpers are foreign, celestial or what is a transformation of the last one on a human plane, migratory. Their predecessors are aboriginal, or relatively so by contrast, and terrestrial. Twin kings are part of this theory of sovereignty, we shall see, as are certain forms of royal incest and royal endogamy (father’s brother’s daughter marriage). And all these structural features conspire to make usurpation itself the principle of the legitimacy. The seeming paradox of a “legitimate usurpation” expresses the double descent of the conquering dynasty. On the one side, through the appropriation of the indigenous princess, the stranger-king bestows the royalty of his predecessors on his descendants, i.e., through matrilateral affiliation. On the other side, the paternal, the dynasty is heir to the favor and charismatic powers of the sovereign Zeus. Such was the double legacy of the Heraclids, as of the Perseids and the Atreids before them. Ancestor of the Dorian conquerors, Heracles of valorous feats was the son of Zeus by the woman Alcmene (fig. 2). Alcmene was the daughter of a Mycenaean king, Electryon, and through Electryon the granddaughter of Perseus, founder of the city. As for Mycenae in pre-Dorian Argos, it was the seat of Agamemnon and the ancient Achaeans, thus the legendary (as well as historical) source of a kingship supreme among the peoples of Greece. The founder Perseus, whose own legendary feats Heracles would equal (e.g. slaying of the gorgon Medusa) and surpass, this Perseus, says Herodotus, had “no human father by whose name he could be called,” but only Zeus (6.53). And Perseus had come to power by a crime against kinship, purportedly accidental: he killed his mother’s father Acrisius, effectively putting an end to the earlier dynasty of the Danaides. On the one side, affinal succession. On the other, conquest, regicide, and more: the anti-structural exploits of a usurping king that are the proof of his own transcendent lineaments. The stranger-king is worthy of his descent from Zeus. Figure 2. Genealogy of Heracles. Between the house of Perseus and the return of the Heraclids (the Dorian conquest), the Atreids ruled in the Peloponnese (fig. 3). Agamemnon, son of Atreus, held the sceptre of Zeus, as Homer tells in a well-known passage (II. 2.100-108), and therefore the leadership of all the Greeks against Troy. The Atreids had achieved their distinction by the same combination of marital and martial exploits. Their migratory ancestor was Pelops, Atreus’ father, author of a famous ruse that allowed him to carry off the royal woman Hippodameia and cause the death of her father Oenomaus, king of Pisa and Elis. Thereafter Pelops’ sons spread over the Peloponnese, taking the kingships of many cities and elevating their father to the status of eponym of the whole region. Pelops’ own father Tantalus, a king in Asia Minor (Lydia), was said by some to be a son of Zeus—thus accounting for the passage of the sceptre to Agamemnon via Pelops, as Homer tells. The sinister character of the Atreids was passed down in the same line. Their history was an unending tale of incest, fratricide and parricide (cum regicide); and of the acquisition of kingship by marriage—Menelaus and Orestes in Sparta, Aegisthus in Mycenae. The historic Spartans were eager to identify with these Achaean kings, and with their legendary sway. They would even appropriate and take to Sparta the bones of Orestes and again of his son Tisamenus, so metaphorically capturing an Achaean ancestry (see below). Figure 3. The line of Tantalus. The pre-Perseid dynasty of the “Egyptian Danaides”—that of Perseus’ grandfather Acrisius—had had similar stories to tell of their own success. Acrisius indeed had married Eurydice, daughter of Lacedaemon, eponym of the Spartan kingdom (Lacedaemonia). And with Lacedaemon we reach the original and minimal form of Peloponnesian usurpation: the replacement, through the mediation of marriage, of the indigenous son of Earth by the immigrant son of Zeus (fig. 4). So run local tales of the first times in Argos, Arcadia, Laconia and Messenia. Later royal houses, such as the Heraclids, formulated an ideology of dominance that could apply to the Peloponnese or Hellas in general, especially as the successors of Mycenaean rulers. But in the myths of early kingship, as collected by the renowned tourist Pausanias, these encompassing claims diverge into the local pedigrees and founding traditions of independent peoples or cities. Still, as closer to the gods and to the Hesiodic Golden Age, the protagonists of the initial dynastic dramas represent abstract concepts only slightly more delimited than those figured in the gods at the beginning of world. In the core regions of Arcadia and the southern Peloponnese, Pausanias gathered several such traditions, recounting the coming of heroes from elsewhere who replaced the aboriginal kings of Pelasgian or Lelegean stock. We would privilege these tales over versions sometimes found elsewhere (e.g., in Apollodorus or Hyginus): not only because of their local provenience, but for their logically motivated relations at once to ancient cosmogony (the Theogony) and to the traditions of later royal lineages (the Atreids, Heraclids, etc.). The whole set of myths then forms a series of transformations, built up recursively on the same basic theory of sovereignty. The passage from cosmogony to “history” sees a progressive expansion of the scale of sovereignty, but also its humanization, which is a reduction of its conceptual or categorical scope. Figure 4. Peloponnesian usurpation The first human times are epitomized in the antecedents of Sparta particularly (fig. 4). Lacedaemon, the aforementioned, gained the kingdom that would bear his name by marrying the royal woman Sparta, giving her name to the city. Sparta was the daughter of [the] Eurotas [river]—“it was Eurotas who channelled away the marsh-water from the plains by cutting through to the sea” (Paus. 3.1.1). For his part, Eurotas was the son and heir of Myles, and Myles of the original king Lelex. Lelex was autochthonous, a son of Earth. Just as the universe effectively begins by the union of male Heaven (Uranus) and female Earth (Gaia), so in Lacedaemonia the landscape and polity are constituted by the conjunction (through marriage) of celestial and terrestrial lineages. At the same time, the kingdom myth evokes usurpations of cosmogonic memory: by Cronus and Zeus successively, who through the mediation of women (their mothers), dethroned and killed their respective fathers and took the latters’ daughters (their own sisters). Pausanias had similar tales to tell of the other major Peloponnesian countries. Arcadia became known as such from the ancestor Arcas, a son of Zeus by the woman Callisto. She was a descendant of Pelasgus, who was the first king of the region and another son of Earth (Paus. 8.1.4-8.4.1). Similarly, the eponymous king Argus was a son of Zeus by the woman Niobe (2.22.5). She was the daughter of Phoroneus, the first man and ruler of the country (2.15.5). In another context, Pausanias brings the aboriginal Pelasgians into the Argive account, as he notes that the acropolis at Argos was named from Larisa, a daughter of Pelasgus (2.23.9). The Messenian story is incomplete and obviously reflects the historic subjugation by Sparta. The aboriginal Leleges who founded Messene were a branch of the Spartan Leleges. But the fate of the earliest kings is not recorded. In Eleia we find a transformation characteristic of the Peloponnesian peripharae. Here the first king (Aethlius) was said to be a son of Zeus by the daughter of Deucalion, but his line was superseded through affinal succession by the eponymous Eleius, who was a son of Poseidon (Paus. 5.1.8). This opposition of Poseidon and Zeus is worth a brief digression. Poseidon rather than Zeus figures as the divine ancestor of kings in several places around the central and southern Peloponnese (e.g. Pylos or Troezen). Poseidon was likewise the ancestor, in Egypt, of the pre-Mycenaean Danaides, who migrated thence to the Peloponnese (to be superseded by the Zeus-born Perseus). Again in later periods, Poseidon was adopted as the answer to Zeus by historic enemies of Sparta: Idas of Messenia, rival to the Dioscuri, is said (by some) to be a son of the ocean god, as also Theseus of Athens. But then, the rivalry of the divine brothers goes back to the beginning, when Poseidon, dissatisfied by the sovereignty accorded to Zeus, joined in an unsuccessful revolt against him. (The ambitious Poseidon went on to contest with various others for the patron status in Corinth [against Helius], Argos [against Hera] and Athens [against Athene].) We say “ocean god,” as Poseidon was of course, but it is notable that his force takes the form of earthquakes, if that of Zeus the celestial form of thunderbolts. One senses a translation, in terms of the Olympian gods, of the ancient opposition between Zeus-born heroes and autochthonous kings. Such again would be the main ideological issue in Sparta’s conflict with Athens. Whatever the final judgment on Poseidon, it is clear that the dominant royalty of the Peloponnese preferred to calculate their sovereignty from Zeus. And it is from Zeus that they derive the structural characteristics by which they were mythologically and historically known. This includes the twin kingship. Heracles himself was a twin. But he was conceived on Alcmene by Zeus; whereas his brother Iphicles was sired by the human husband, the Perseid king Amphitryon. Such was the prototype of Dorian twinship; however, the paradigm is most developed for the Dioscuri, the famous doubles of the Spartan kings. We turn first to their story. The Dioscuri and the war of twins The divine twins Castor and Polydeuces (Pollux) have mythical cognates across the old Indo-European world, from Vedic tradition to the Scandinavian—including the Theban twins Zethus and Amphion. The Dioscuri’s own cult was widespread in classical Greece (as well as Rome). But this cult was centered in the Dorians of the Peloponnese, specifically in Sparta: “Il faut encore remarquer le caractère essentiellement dorien du couple des Dioscures” (Daremberg 1892: 253). The Dioscuri were native to Sparta, indeed of ancient Lacedaemonian lineage. They were born to Leda the wife of the Spartan king Tyndareus; hence, according to the pedigree collected by Pausanias (3.1.1-5), they were of the dynasty founded by Lacedaemon himself. An allusion of Pausanias’ in this context indicates they succeeded Tyndareus as kings of Sparta. (This would have to be a Spartan variant, making them rather than their father the last rulers of the Lacedaemonian line.) Except for their exploits on the voyage of the Argonauts, the twins’ famous battles were fought on behalf of the Spartan kingship. They recaptured their sister Helen, with whom they had been raised in the house of Tyndareus, from the Athenian hero Theseus; they defeated the rival royal twins, Idas and Lynceus, kings of Messenia. The Dioscuri presided over Spartan games as well as Spartan battles. And as we already know, as tutelaries of the historic kings of Sparta, their image accompanied the latter to war. The name Dioscuri means “sons/youths of Zeus” (Burkert 1985: 212). In Laconian inscriptions they appear as the Tyndaridae, “sons of Tyndareus.” The apparent ambiguity is already meaningful: the sign of a double nature, human and divine. The same ambiguity attends their parentage in classical mythology generally. Alternately (according to the version), they were both sons of Zeus or both sons of Tyndareus, although in either case Polydeuces was the elder. But the version that accords with their ritual presence and mythical action in Sparta, as well as with other royal twinships of Peloponnesian legend, is that Polydeuces was fathered by Zeus, who first lay with Leda, Castor by the human sovereign Tyndareus (Apollod. Bibl. 3.10.7). Polydeuces, then, is full brother to the divine Helen, Castor to the more infamous Clytemnestra (fig. 5). This motivated tradition is also logically the most general, as it includes the transformations represented by a uniquely Zeusian or uniquely Tyndarean parentage. The twins were the social sons of Tyndareus, if Polydeuces was the natural son of Zeus, and both were eventually translated by Zeus to immortality—thus “youths of Zeus.” Like the two Spartan kings, the Dioscuri were both divine and human; or alternatively, one was divine but the human one was his very image. So like the Spartan kings, they were minimally differentiated by some mark or attribute, in other respects they were the same. In the representations of the Dioscuri on ancient coins, reliefs and the like, there is often some detail that distinguishes them, but evidently there was no fixed tradition in this regard (Daremberg 1892: 253). Sometimes Polydeuces is shown as a boxer, consistent with the Homeric description: “Kastor, tamer of horses, and the strong boxer, Polydeukes” (Il. 3.237). But in these terms of martial capacities, perhaps the more pertinent contrast is that Castor was the general (he taught the arts of strategy to Heracles), and Polydeuces the fighter (he had scars of battle on his face). Polydeuces appears to play the protective part of Zeus, Castor the human cum social part. Thus Castor proved to be mortal in the fight with the rival royal twins of Messenia (see below), but Polydeuces avenged him and survived. Yet in all their adventures the Dioscuri acted as one, and in the end not even Castor’s death would separate their fates. Figure 5. Parentage of the Dioscuri These adventures included the rescue of Helen, who had been abducted by Theseus (Apollod. Bibl. 3.10.7-8). Foreshadowing the Haa—Helen’s abduction by Paris and rescue by the brother-kings Agamemnon and Menelaus—the incident establishes an analogy between the Dioscuri and the Achaean heroes. Transitively, then, this makes another connection between the two kings of Sparta and the legendary hegemony of the pre-Dorian Achaean rulers. We reserve this point for the moment, to attend to the other defense of Spartan honor by the Dioscuri: their battle with the Messenian twins Idas and Lynceus. This war of twins, as we shall call it, was a revelatory affair. The myth puts royal twinship in the context of a number of structural elements that are all (we claim) related parts of the same general system of sovereignty: capture of the ancestry of the established dynasty through appropriation of royal women, incest, and the marriage of parallel cousins (FBD marriage). Besides, paralleling the historic conquest of Messenia by Sparta, the narrative demonstrates the supremacy over other kings that could be claimed by royal twins of Zeus. The genealogy of the war of twins presents the political issues en jeu. There are two principal variants: one consistently developed by Pausanias (3.1.4-7; 4.3.1), and an alternate interwoven with the first, sometimes confusedly, by Apollodorus. Essentially, they come down to local and regional versions of the same conflict. By inserting Perieres into the Lacedaemonian line, as father to the Spartan king Oebalus (Apollod. Bibl. 3.10.4), the Apollodorian variant confines the struggle for supremacy within the Spartan dynasty. The Pausanian account, however, by making Perieres king of Messenia and the successor of its own Lacedaemonian royalty, opposes the hegemonic claims of the Messenians and the Spartans (Paus. 4.2.2). But in either case, at issue is superiority in the Peloponnese, that is, as affinal successor through Perseus’ daughter Gorgophone of the founder of Mycenae in Argos. Set before the advent of the Heraclids, the war of twins becomes the mythic charter of the later struggles between the Dorian kings of Argos, Messenia and sparta. As the outcome is the triumph of the spartan Dioscuri, who nonetheless disappear from the mortal stage, the myth sets up the universal pretensions of their earthly counterparts, the twinned kings of sparta. Figure 6. Genealogy of the war of the twins I We will follow the Pausanian version, being the more general and of definite local origin, although the variant Apollodorian genealogy leads to the same conclusions (on a reduced scale). The key figure is the royal woman Gorgophone, she of excessive marital relations (fig. 6). Again, Gorgophone was the daughter of Perseus, the son of Zeus and founder of Mycenae. By Pausanias’ telling (2.21.8), Gorgophone was the first widow to remarry. Wedded initially to Perieres, king of Messenia, by whom she had twin sons, Aphareus and Leucippus, she upon her husband’s death married the Spartan king Oebalus, bearing him Tyndareus, Hippocoon, and Icarius. Gorgophone thereby effects the triple conjunction of the Argive, the Messenian and the Laconian kingships, setting the stage for the last two to fight it out over the ancestral legacy of the first. The struggle unfolds by repetition of the same issue, the capture of ancestry through women. Gorgophone’s Messenian and Spartan descendants, the rival sets of twins, fight it out over women who are their own sisters and cousins. Whoever takes these women will become son-in-law to the other lineage: classic relationship of the change of dynasties. And there is also more than the suggestion in the texts that the joint kingship of brothers signifies such encompassment. Aphareus and his younger brother Leucippus appear to rule Messenia together, as (less certainly) do Tyndareus and Icarius in Laconia. In the succeeding generation, where the cokings of Messenia, Idas and Lynceus, do battle with Castor and Polydeuces, this suggestion becomes compelling. Moreover, it matches the oral-historical record. When the famed wars between Dorian Messenia and Sparta began (in the eighth century BCE), both kingdoms were ruled by co-kings, according to Pausanius: “when Teleklos’s son Alkamenes was king of Lakonia, and the king of the other family was Theopompos, the sixth in line from Eurypon, when Messenia was under Antiochos and Androkles the sons of Phintas, the mutual hatred of the Lakonians and Messenians came to a head” (Paus. 4.4.4). But in the course of their losing struggles, the Messenians saw their own dual kingship desolve into the unique rule of one king, and eventually into the dictatorship of the valiant Aristomenes. With the decline of kingship, the Messenians also abandoned all claims to sovereign devolution from Zeus: “the Messenians,” Pausanius found, “do not foist Aristomenes on Herakles or on Zeus, as the Macedonians do Alexander on Ammon... I know myself that when they pour the ritual wine the Messenians call Aristomenes the son of Nikomedes” (4.14.8). Parenthetically, one has to wonder how much the glories of the Spartan kingship contributed to the decline of Hellenic monarchies generally, among states rival to Spartan power. For if it were true that “Zeus is king in heaven by the universal reckoning of mankind” (Paus. 2.24.5), as all Hellas acknowledged, still no other people could hope to match Sparta’s rights to the kingship of Zeus on earth. Except, perhaps, by the denial of the sceptre-doctrine altogether, in favor of some other. If any other people had a king, he had a king in Sparta. No doubt there were internal reasons that monarchy outside Sparta (and, e.g., Macedonia) had become obsolete by the classical period. But beset by turbulent Sparta—as well as Persia with its own archetypal king of kings—other Greeks must have discovered that the royalty they had been dealt was not the winning hand. Conversely, as the Spartans realized the historic strength of their Zeus-born kingship, they were satisfied to maintain it, until it was virtually alone. Could such a process confirm Finley’s suspicion that “survival” is not the right term? Figure 7. Genealogy of the war of the twins II We close the parenthesis and return to the mythical war of twins, taking note that it was also marked by a certain incestuous relation and a specific form of royal marriage, between the children of brothers (FBD marriage). Excessively close from the point of view of kingship, the incest was also a political excess on Messenia’s part. Aphareus the Messenian king married his half-sister Arene (fig. 7). She was, however, the daughter of the Spartan oebalus—and we know what kind of challenge this means to an established lineage. (Arene appears to be the mother of the Messenian twins, Idas and Lynceus, according to Apollod. Bibl. 3.10.3) In the next generation, the marital/martial issue turns on father’s brother’s daughter marriage. Idas and Lynceus are betrothed to their parallel cousins (FBD) Phoebe and Hilaira, priestesses respectively of Athene and Artemis. The two women, however, are successfully abducted and married by Castor and Polydeuces. Sometimes alleged to be the causus belli of the war of twins, this abduction again entails father’s (half-)brother’s daughter marriage: the women are also parallel cousins of the Dioscuri. We shall examine the mythical analogies and meanings of such marriage practices in the next section. Suffice it to note here that they are imitations of the incestuous excesses of Zeus. But then, the Messenians who lost the battle, are also said to have given up the pretext of kingship from Zeus: Idas, the elder of the Messenian twins, was a son of Poseidon, according to some (cf. Apollod. Bibl. 3.10.3). The proximate cause of war was either the Spartan twins’ abduction of the Messenian princesses, or their quarrel with Idas and Aphareus over some cattle that had been taken in a joint raid on Arcadia. In the latter tale, the Dioscuri had been tricked by their Messenian counterparts, who made off with the booty, upon which the Spartans marched on Messenia and seized the cattle. The symbolic issue in the two versions, however, is not that different, since either may signify a contest of kingship. The right to distribute the booty—which in this case the Messenian Idas had taken on himself, only to cheat the Dioscuri—is a royal prerogative. Besides, cattle are the most noble sacrifice (recall the historic privileges of Spartan kings in this regard); whereas, women are the means of acquiring the descent of royal predecessors. To complete the symbolic triad: “Dans une civilisation masculine comme celle de Grèce,” cattle and women are ritual equivalents, and equally the object of heroic acts of seizure. So Vernant tells us: Dans sa forme la plus ancienne (et dans un milieu de noblesse que la poésie épique nous fait atteindre), le mariage est un fait de commerce contractuel entre groupes familiaux... Parmi les présents... il y a une prestation qui a valeur spéciale parce qu’elle a lieu, de façon expresse, en contrepartie de la femme dont elle constitue le prix : ce sont les ἕδνα [hedna]. Il s’agit de bien précieux meubles, d’un type très défini: bêtes de troupeaux, spécialement des bovins... Par la pratique du mariage par achat la femme apparaît équivalente à des valeurs de circulation. Mobile comme eux, elle fait comme eux l’objet de cadeaux, d’échanges et de rapts (Vernant 1985: 170-171).In its oldest form (and among the nobility to whose circles epic poetry introduces us), marriage is a formalized transaction between family groups. . . . Among the gifts exchanged. . . there is one of particular value because it is explicitly given in exchange for the woman, and is in fact the price paid for her. This is the ἕδνα [hedna], a very valuable commodity of a very definite type: prize animals from the flocks and herds, especially male cattle. . . . By this practice of marriage by purchase, the woman appears equivalent to the values in circulation. Being mobile in the same way, she is similarly the object of gifts, exchanges, and abduction (Vernant 1983: 139; translation modified). The events of the final battle are also variously told, but the dénouement is generally agreed upon. Idas killed Castor; thereupon Polydeuces, with the help of Zeus (n.b.), destroyed the Messenian brothers. Polydeuces then pleaded with Zeus to share his brother’s fate. Acceding to the request, Zeus ruled that the two should pass their days alternately living as the gods on Olympus and buried as men under the earth: Turn and turn about they passOne day with their loving father Zeus,The other hidden by earth in Therapne’s cavernsAnd fulfill a like fate.This life, and not to be fully a god and live in the sky,Polydeuces chose, when Castor was killed in war.(Pind. Nem. 10.55-59). In the end, both Polydeuces and Castor acquire a double nature, mortal and immortal. The complex onto-logic is the same, we say, as attends their historic alter-egos, the two kings of Sparta. One, the senior is more godly; yet as the two are to all social appearances identical, the godly one is also human, and the human one also godly. Like the metaphysics of the king’s two bodies as perceived by the Norman Anonymous, both king and god are twinned persons. The twinship can then function to proclaim the legitimacy of a given ruling line. These twinned kings seem to intervene at critical points of royal genealogies: to transform the original usurpation of a divine heir into a principle of dynastic continuity; or else to transfer (via the mother) the sceptre from one human line to another. In either case the political virtue is a reign at once of established human pedigree and of Olympian descent. Imitations of Zeus The several aspects of the Peloponnesian theory of sovereignty—including twrnship, usurpation and the exploits of conquest and assassination—are interrelated by the common signification of a human succession of Zeus. Hence the iconic resemblances between cosmogonic myth and kingship legend. The argument can be extended to the mythic testimony of royal incest and parallel cousin marriage. In fact father’s brother’s daughter marriage, we shall show, is a humanized mode of sororal incest. This kind of transformation progressively appears over time in the mythic corpus, a humanization of divine practice, affecting also the character of royal twinships. The anti-structural outrages of the first gods are gradually sublimated into human-social customs, until among the protohistoric and historic kings they present symbolic reflexes of their original forms. Let us go back to the beginning. We analyze the kinship relations of the genealogical line that leads down from the gods to Hellen, ancestor of the Greeks, as recorded in the Theogony and by Apollodorus (1.7.2-4). Here are found the same kinship practices as informed the war of twins, but expressed in cosmological dimensions that lend them a greater significance and a readier intelligibility. In the broadest terms, the permutations of marriage practice in these first generations of the human career represent the working-out of a Hesiodic degeneration from the Golden Age. Progressively removed from the immortals, the practices of human kings become ever more distant imitations of Zeus. Consider for example the progression from divine incest—more precisely, its extreme form, parthogenesis—to the father’s brother’s daughter union that gave rise to Hellen. As a social reproductive capacity, incest is a modality of the famous autonomy of the olympian gods, their self-sufficiency, of which their freedom from labor is another. In social terms, incest is a denial of dependence on others for reproduction, hence another aspect of immortality. And as it is among humans a crime against kinship, a transcendent anti-structural act, it becomes for the gods the proof that they are stronger than society—and thus able to constitute it. Conversely then, the replacement of divine incest by human parallel cousin marriage proves that mankind is condemned to a state of dependence. Just as men, unlike gods, are dependent on nature and cannot live without travail, so they are dependent on others for the (natural) means of their reproduction, with all the problems such alliances breed. Exogamy is the analogue of human mortality, or even in myth—e.g. the war of twins—the cause. On the other hand, the solution of parallel cousin (FBD) marriage is a close human approximation to the sexual excesses of the olympians. Hence its incidence among the gods’ royal successors. In structural terms, this passage from cosmogonic incest to human marriage is a transformation constructed on a critical invariant. Taking place between the children of twin brothers, who are thus one-in-two, parallel cousin marriage amounts symbolically to the incestuous union of siblings. The genealogy of Hellen is careful to make the point, as it takes the development of human marriage step-by-step through a finely graded series of decreasing incest. First, parthogenesis: the creation of male Heaven, Uranus, from female Earth, Gaia, which is immediately followed by the sexual union of the two. The mother-son relationship of Gaia and Uranus is followed by the brother-sister mating of their offspring, oceanus and Tethys (Hes. Theog. 337). Brother-sister marriage gives way to union of brother’s daughter and father’s brother, Asia and iapetus. And two generations later, the series culminates in the marriage of Deucalion to his father’s (younger) brother’s daughter, Pyrrha—the “primal couple”, as West (1985: 139) describes them. The marriage is indeed prototypical. In later kingship myths, as we have seen, it recurs between the children of twin rulers. Something of the same sort is being said cosmogonically, inasmuch as Deucalion and Pyrrha (his FBD) are the children of the brothers Prometheus and Epimetheus. While it is not said that the latter were twins, their fraternal solidarity and resemblance is well-enough remarked (Forethought and Afterthought). Also like the twin kings of later fame, Prometheus and Epimetheus were minimally distinguished on the axis of divinity and humanity, even as in tandem they were intermediate between the two, and effected the transition from one to the other. Prometheus, the elder, is more like the gods; while Epimetheus, with his well-known failings, proved all too human. Parallel again to the senior of twinned kings, Prometheus especially acts as a being of double nature. A Titan who challenges Zeus, he does so on behalf of humanity. According to some accounts, he fashioned the first human beings out of clay. But Epimetheus, by virtue of his imperfections, could not resist the “beautiful evil” Pandora, the first woman made of such mortal stuff (Hes. Theog. 585; cf. Vernant 1979: 98-101). The consequences were tragic for all of us. The Deucalian flood is not numbered among them by our authors, but it wipes out Prometheus’ creative work, leaving only Deucalion and Pyrrha to recreate men and women from stones—with the help of Zeus. But then, does not the marriage of Deucalion and Pyrrha, children of the fraternal pair of Titans, appear as a human translation of the sororal incest for which Zeus was well enough known? This must be the reason that of Hellen, Deucalion’s supposed son, “some say Zeus was his father” (Apollod. Bibl. 1.7.2). The subsequent success of Hellen and his sons was another reproduction of Zeus, of a kind with the achievements of Lacedaemon and kindred heroes who replaced the indigenous kings of the Peloponnese. Except that the Hellenic conquest, eventually covering all of mainland Greece and the islands, is set on a wider scale. Giving rise to the several Greek nations, the descendants of Hellen everywhere superseded the aboriginal inhabitants: Pelasgians, Carians, Lelegeans, and other sons of the soil. But as generations passed and the distance of mortals from the gods increased, such counterparts of Olympian exploits in human practices became ever more reduced versions of the divine ideal. The differentiation of a specifically human nature continued, until even kings had to reconcile themselves to exogamy. The rule of twin kings and parallel cousin marriages are specifically associated in legends of the pre-Dorians. The combination of twinship followed by father’s younger brother’s daughter marriage makes a nice ideal type. It not only represents brother-sister marriage in a humanized form, but by virtue of the hypogamy retains the sovereignty in the more godly line, while it recapitulates within the dynasty the appropriation of the earthly/feminine by the celestial/masculine side. In at least one such affair, concerning the children of Aegyptus and Danaus (see below), the demand of cousin marriage (FyBD), is taken as the sign of the senior brother’s attempt to secure the whole power. In the traditions of Spartan kingship, there is a remarkable testimony to the continuity of parallel cousin marriage between the children of twin rulers, but in a still more humanized—i.e. exogamous—form. Like the Dioscuri, the original royal twins of Sparta, Procles and Eurysthenes, married twin sisters, Lathria and Anaxandra (fig. 1; Paus. 3.16.6). The sisters were relatives of their royal husbands in a collateral branch of the Heraclids: they descend in the paternal line, from Heracles’ second son (Ctesippus); while the Spartan kings trace to the eldest son (Hyllus). Thus the initial royal marriage of Sparta is a “classificatory” father’s brother’s daughter union, a marriage of women of the cadet branch to men of the senior. But if this is an exogamous version of parallel cousin marriage, it should be remembered that true cousin marriages among royal twins had not always worked out well. Both the Aphareids and the Dioscuri had tried it, yet the descendants of neither would inherit the throne over which the rival twins had given battle. There is a double bind in Greek mythology, which is the mechanism of the Hesiodic distantiation from the Golden Age. Descended from the sovereign gods, the human kings presumed to duplicate the exploits of the Olympians. But such is the essence of hubris, and they suffer for it. In the event, such feats as incest or practices as cousin marriage became a mythical memory: doings of the ancestral predecessors that could validate the divinity of the royal lineage, but in later times were not repeated, and in the historical accounts not remarked. The twinned kingship itself suffered a similar symbolic reduction. In Sparta it continued to serve as a legendary precedent, but of course the historic kings were twinned rather in function than by birth. Heracles himself could have been a lesson to Spartans in the hubris of real twinship and actual incest. His mother Alcmene of the Persidae had married her mother’s brother, who was also her father’s brother’s son (fig. 2). Or taking it from the point of view of her royal husband Amphitryon, an eldest son and descendant in the senior line from Perseus, he married his sister’s daughter, who was also his parallel cousin (FyBD). However, Zeus was the father of Alcmene’s oldest, Heracles, as he disguised himself as Amphitryon and lay first with her; whereas, Heracles’ twin Iphicles was sired by the human husband. Neither of the twins succeeded to Perseus’ legacy. Jealous with Zeus for his affair with Alcmene—his last, recall, with a mortal woman—Hera delayed the birth of Heracles in a celebrated ruse, depriving the hero of his Myceneaen birthright, which passed instead to Eurysthenes. Iphicles never amounted to much. Besides, Heracles killed Iphicles’ sons as well as some of his own (by the oldest daughter of the Theban king) when driven mad by Hera. Before Perseus, the Danaids had ruled Argos, and practiced endogamous marriages with similar tragic consequences. Even in the Egyptian prehistory of the dynasty, the twin sons of Poseidon disagreed and split the rule, Agenor going off to Phoenicia, leaving Belus to reign alone over the Nile kingdom. As for the twin sons of Belus, Aegyptus and Danaus, their quarrels were fierce, and turned specifically on the father’s (younger) brother’s daughter marriage (fig. 8). Figure 8. Origins of the Danaids Aegyptus had fifty sons by many wives, as likewise Danaus had fifty daughters. But when Aegyptus asked for his brother’s children as wives for his own sons, Danaus fled to Argos with his daughters, fearing (n.b.) that Aegyptus was thus plotting to take the whole kingdom. By Pausanias’ account, Danaus was accorded the throne (of Gelanor) by virtue of his descent from an Argive royal woman (Io). But Aegyptus’ sons did somehow manage to marry Danaus’ daughters, who thereupon slew their husbands on the wedding night—all but the oldest daughter who spared her cousin-husband. The famous descendant of the surviving couple was the ill-fated Acrisius: from the womb he quarrelled with his twin brother Proetus (Apollod. Bibl. 2.2.1), with whom he was finally forced to divide the kingdom; nor, for all the cruelties he practiced on his daughter Danae, could he avoid being slain by the latter’s son Perseus, as the oracle had foretold (fig. 9). One is naturally reminded of the Theban Oedipus, whose incestuous hubris (mother-son marriage) could rival Heaven (Uranus) himself. His own twin sons likewise failed to alternately share the kingdom, and ended by killing each other. There were other twins who did not get along too well, e.g. Pelias and Neleus of Pylos. The mutual love of the Dioscuri seems exceptional and, in light of their special relation to the Spartan kings, significant. But by historic times, so far was humankind distanced from the marvellous exploits of Zeus on the wives of kings, true royal twinship was rather in disrepute, even in Sparta. Hence Herodotus’ repetition, with respect to the original Spartan twins, of what was by now a mythical saw: that Procles and Eurysthenes had quarrelled with one another all their lives, as, he added, have their respective descendants ever since. The last half of the observation was patently false; while the first, invoking the disadvantages of twinship, neglects the singular structural parallel preserved in myth between the dominance of the Spartan twins among the Dorians, and the earlier dominance of the Achaeans among all the Hellenes. Figure 9. Antecedents of Perseus. We refer to the recurrent set of relationships among the sons of Hellen and the sons of Heraclid Aristomachus (fig. 1). Hellen is the proximate ancestor of the main Greek peoples; Aristomachus, of the Dorian conquerors of the Peloponnese. The story of the first refers to the original advent of the Greeks, of the second to the so-called Heraclid return, several generations after the initial conquests of Heracles’ son Hyllus. The two migrations have very similar genealogical codes. They are also versions, set on different scales and in different eras, of the same concepts of political hegemony. Both indeed speak to a two-fold supremacy: of the foreign invaders over the settled inhabitants of the country, and of one group among the invaders over the other, collateral groups. The several nations of Greeks sprung from these descendants of Hellen and were named after them: the Dorians from Dorus, Aeolians from Aeolus, Achaeans from Achaeus, Ionians from Ion. (Hellen was king of Thessaly; of Xuthus we shall speak presently). The sons of Aristomachus the Heraclid likewise divided the Dorian conquest—this by lot—each taking one of the three main countries of the southern Peloponnese (Apollod. Bibl. 2.8.4; Paus. 4.3.3-5). Temenus, the oldest, received Argos, traditionally the leading kingdom; Cresphontes, by trickery in the draw, got fertile Messenia; and the twin sons of Aristodemus, Procles and Eurysthenes, took sparta, perhaps the least desirable. In the same way Hellen’s patrimony had been divided among three sons, one of whom was a father of two. So far: so alike—but there is more than formal parallel. Following the Dorian conquest, Sparta embarked on a career of sustained aggression against the brother kingdoms of Messenia and Argos. (we are passing now from myth into history.) The Messenians were subjugated completely and their Heraclid kings eliminated. The Dorian Argives drove out the descendants of Temenus, and by the time of the Persian invasion repeated wars with the Spartans left the Argolid prostrate. The dual kings of Sparta were then the sole survivors of the royal lineage of Heracles, thus the exclusive heirs to the Mycenaean legacy of which Heracles had been deprived. Here, then, is a triple correlation between: (1) the distinctive kingship of Sparta, alone among the Dorians to begin in twins; (2) the unique derivation of sovereignty from Zeus, which such twinship is known to signify (cf. the Dioscuri); and (3) the pretensions of dominance over other kings and kingdoms, a human parallel to Zeus’ divine sway. Politically as well as genealogically, the system of relationships among the descendants of Hellen is virtually the duplicate. Of Hellen’s three sons, only Xuthus is noted (Apollod. Bibl. 1.7.3) to be the father of two (i.e., in this context of constituting myth and eponymous classification). Achaeus and ion are not reckoned as twins, so far as we know, but the pair achieve a political position in the Peloponnese—and later in all Hellas—analogous to Aristodemus’ twin sons, or again to the Argive brothers Agamemnon and Menelaus, who effectively connect the two genealogical charters. Like Aristodemus, Xuthus will soon disappear from the scene. Xuthus steals his brothers’ inheritance, is exiled, and goes to Athens. Notice that Aristodemus was struck down by a thunderbolt, according to the prevalent account (Apollod. Bibl. 2.8.2), a sure sign that he had invoked Zeus’ displeasure. The key term of the sibling triad, the father of the two sons, has a sinister character. But then, Xuthus and Aristodemus are the fathers of conquests, and of aggressions against their fraternal kinsmen. The disposition of the ancestor is realized in the political expansion of the line. Xuthus had received the Peloponnese in Hellen’s division of the country. Before the Heraclid return, the Achaeans and Ionians, who sprung from Xuthus’ sons, respectively dominated the southern and northern Peloponnese. And by a set of equations well known in the classical period, the myth could retain a certain currency, as it had by then a much wider political import. By classical times, the myth of Hellen’s sons could oppose the Spartans to the Athenians, as Achaeans to Ionians. Descended from Dorus, the Spartans however had absorbed the identity of the Achaeans. From Achaean predecessors they took the sceptre of Zeus—and a certain identity. They reburied the bones of Orestes and Tisamenus in Sparta, acts that gained them certain victories and the imagery of paternal descent from the house of Atreus. Pausanius’ comparison of the Spartan king Agesilaus (early fourth century BCE) with Agamemnon seems more than a simile of his own invention: “Agesilaos thought himself king of a richer city than Agamemnon and like him the lord of all Greece, and he believed that to overpower Artaxerxes and possess the riches of Persia would be a more glorious achievement than the destruction of the kingdom of Priam” (3.9.4). So would the Spartan Cleomenes I, of equally megalomaniac reputation, announce to the Athenians (late sixth century BCE) “I am no Dorian, but an Achaean” (Herod. 5.72). For their part the Athenians, by a mytho-praxis at least as arbitrary, took themselves to be Ionians, thus lords of nearly all as far as Asia Minor. This identification too was a kind of mythical passe-passe—and contradictory to their self-proclaimed autochthony. Neither Ion nor Xuthus ever ruled Athens. Ion, however, once served as Athens’ military lord (polemarch), an episode deemed sufficient to give the Athenian phylae their Ionian names and Athenian colonization its “Ionian” character. The arbitrariness of all this is, from another perspective, its logical value. The two sons of Xuthus could thus continue to represent the dominant stocks of all the Hellenes. And the Hellenic charter structure turns out to be very like the Dorian, not in form only, but in linking a political hegemony with a sovereign duality. The traditions of Hellen and Xuthus open into another set of mythic permutations, again involving royal endogamy, but only to radically distinguish Athens from the whole Dorian concept of a Zeus-born usurpation. To spell out all the transformations would need another essay. Here we concentrate merely on a few incidents of early kingship legend in Athens. Hellen and Xuthus have their parts in these events, but the parts are negative. The episodes in question constitute an explicit disavowal of such stranger-kings in favor of the aboriginal sons of Earth. Figure 10. Athenian permutations I Athenian permutations Not that the early Athenian kingship was innocent of dynastic succession through the woman. Cecrops (I) replaced the original ruler Actaeus through marriage to the latter’s daughter. And Cranaus was elected to succeed Cecrops, though not apparently related to him. But if Actaeus was a “son of the soil,” so were Cecrops and Cranaus; there is no question here of the accession of an heroic outsider. The question does arise, however, with the coming of Hellen’s brother Amphictyon (fig. 10). Amphictyon migrates to Athens from the Peloponnese, marries Cranaus’ daughter (Atthis), expels his father-in-law, and becomes king. But only for a brief interval. Amphictyon is driven out by Erichthonius, the great chthonic hero and king of Athens—he was half serpent. Erichthonius was born of Earth, in consequence of the hilarious and futile attempt of the limping god (another chthonic sign) Hephaestus on the chaste Athena. In disgust, Athena brushed Hephaestus’ seed from her leg, whence it fell to the ground, inseminated Earth and issued in Erichthonius (Apollod. Bibl. 3.14.6). So by driving out Amphictyon, Erichthonius preserved Athens’ own purity. The Athenians succeeded in doing what the aboriginals of the Peloponnese (Pelasgus et. al.) could not: rid themselves of would-be heroic usurpers (Lacedaemon et. al.) The Athenian legends go on to mark the contrast by inventive permutations of the common Peloponnesian versions of dynastic origins. Erichthonius’ son, Pandion (1) marries his mother’s sister, a matrilateral incest as opposed to the typical patrilateral forms of the olympians and their royal avatars (fig. 11). As we might expect, Pandion’s wife Zeuxippe (a suggestive name) then gives birth to twin sons, Erechtheus and Butes. Indeed “Butes, it is said, was a son of Poseidon: so Hesiod in the Catalogue” (Cat. Wom., Frag. Hes. frg. 223 ). Still in the same line of mythical thought: Erechtheus and his twin Butes share Pandion’s kingdom—but in a way quite different from the joint rule of their Dorian and Achaean geminal counterparts. Figure 11. Athenian permutations II The Athenian tradition now produces the great transformation of dual sovereignty: the complementary rule of king and priest rather than the symmetrical, twinned kingship. Erechtheus’ share is the kingship, and Butes’ is the priesthood of Athena and Poseidon-Erechtheus. The Eteobutadae continued to administer these central cults of the Acropolis into the historic period, by which time the kingship itself was resolved and differentiated into the complementary offices of the archon, basileus, and polemarch. Finally, let us go back to the wandering Xuthus, son of Hellen. He married Erechtheus’ daughter Creusa; it was by her he had Achaeus and Ion. It seems a perfect set-up for usurpation in the classic Peloponnesian style. Except that Xuthus, when called upon to adjudicate the succession of Erechtheus, chose the latter’s eldest son Cecrops (II), thus depriving his own sons and Cecrops’ younger brothers. For these good offices, he and his sons were thrown out of the city. And Athens could sustain its claims to autochthony, an ideology of no little service in the historic opposition to Sparta. The ancestors of the Athenian dead, said Socrates, were not strangers, nor are these their descendants sojourners only, whose fathers have come from another country, but they are children of the soil, dwelling and living in their own land. And the country which brought them up is not like other countries, a stepmother to her children, but their own true mother; she bore and nourished them, and in her bosom they now repose. . . . And a great proof that she [Earth] brought forth the common ancestors of us and the departed is that she provided the means of support for her offspring. For as a woman proves her motherhood by giving milk to her young ones. . . . so did this our land prove that she was the mother of man. . . . And these are truer proofs of motherhood in a country than in a woman, for the woman in her conception and generation is but an imitation of the earth, and not the earth the woman. (Plato, Menex. 237b-238a). Dual kingship in the Iliad The “Achaean” claims of the later Spartans pose the vexed question of dual kingship in the Iliad. This will be our final consideration—and our own ultimate hubris. It does not seem possible that anything we can say about Homeric kingship has not already been said by our scholarly betters. Still, I risk. . . . The problem of diarchy in the Iliad breaks down naturally into two parts. First and most obvious, the relationship of Agamemnon and Menelaus. Second, and perhaps less remarked, the several instances of dual leadership among other forces than the Achaeans proper, as noted in the Catalogue of Ships and elsewhere. This dualism occurs alike on the Argive and Trojan sides, particularly among the Lycian allies of the Trojans. It is customary to refer to such as dual “captaincy,” as if it concerned leadership in war only, but this is to ignore the explicit designation of the personages concerned as heroes and royals (basileis. Nor should Odysseus’ celebrated remark that “many lords are not good,” that there should be but one king “to whom crafty Cronos’ son gave the sceptre” (Il. 2.204-6) be taken as definitive, since it was addressed to the unruly mob. It appears rather that the Spartan dual kinship had many more—and more widespread—ancient precedents than are generally acknowledged. Agamemnon, ruler of Mycenae, was of course primus inter pares with respect to Argve kings, and his brother Menelaus was ruler of Sparta. The “secundogeniture” (as Nilsson called it; 1972: 69) need not detain us, since it is well known that Agamemnon was at home in Sparta, where his son Orestes succeeded (through a father’s younger brother marriage), and it is well attested as a division of rule among twin kings of greater mythical antiquity (e.g. Belus and Agenor, Acrisius and Proetus, etc.). Interesting in this connection is Eric Hamp’s argument that the two names “Agamemnon” and “Menelaus” are based on the same root-term for “ruler”: “It seems that in Indo-European society the term for supreme ruler or chief among petty rulers was of the form *GREAT + the term for ruler; this seems to have formed a close compound” (personal communication). Elsewhere Hamp writes: “as Heubeck. . . has pointed out, *Mένμων is nothing but an apocopation of the name-set epitomized by *Μενέλαος; it is therefore the formulaic equivalent so to speak. *Ἀγα-μένμων is then ‘great, Ober-*Mένμων.’ He is the principle wanakts of the coalition, whatever such a contemporary office exactly was” (Hamp 1971: 24). Such etymology is functionally expressed in the iliad by a number of references to the joint responsibility of Agamemnon and Menelaus for the expedition, their shared status as its leaders, and their coupling in these respects by opposition to other kings of the Argive host. Among the relevant passages are: (1) The invocation of Atreus’ two sons by Chryses in Book 1 when he offers to ransom his daughter, allotted to Achilles in the division of booty. Chryses supplicated the Achaeans,but above all Atreus’ two sons, the marshals of the people [kosmetore laon]:“Sons of Atreus and you other strong-greaved Achaians” (Il. 1.15-17). (2) It is specifically Menelaus’ and Agamemnon’s honor that is being served by this war, Achilles says in his protest over the loss of the woman (Il. 1.159-160); and then in explaining his sorrow to his goddess-mother, he confirms that Chryses has “supplicated all the Achaians, but above all Atreus’ two sons” (Il 1.374-75). (3) This sense that the expedition or the rescue of Helen belongs jointly to Agamemnon and Menelaus and redounds specifically to their honor is repeated a number of times: at the end of the Catalogue: “Tell me then, Muse, who of them all was the best and bravest . . . who went with the sons of Atreus” (il. 2.760–762). at the death in the first great battle of orsilochus and Crethon, twin sons of Diocles of Phere (n.b.), who had followed along to Ilion (Troy), “winning honor for the sons of Atreus, Agamemnon and Menelaus” (il. 5.552-53). after the battle, Antenor counsels the Trojans to give back Helen “to the sons of Atreus” (Il 7.351). with Agamemnon out of action, Menelaus, “shepherd of the people” assumes command, and calls upon the Achaeans to help prevent Hector’s taking Patroclus’ body: “Friends, o leaders and men of counsel among the Argives, you that beside Agamemnon and Menelaus, the two sons of Atreus, drink the community’s wine and give, each man, his orders to the people” (Il. 17.248f). (4) But in fact, Agamemnon and Atreus do not drink wine like the other kings. When the ships came with wine from Lemnos, sent by Jason’s son Euneus, the sons of Atreus were given theirs apart and as a gift, unlike the other Archives: “Apart to the sons of Atreus, Agamemnon and Menelaus, Jason’s son had given wine as a gift, a thousand measures; and thence the rest of the flowing-haired Achaians bought wine” (Il. 7.470-473). (5) Agamemnon and Menelaus are of twin-minds and shared yet unspoken disposition, that is, by nature. When Agamemnon calls a counsel of battle, Menelaus appears without a summons: “of his own accord came Menelaus of the great war cry, who knew well in his own mind the cares of his brother” (Il. 2.408-409). After Achilles had refused to rejoin the fight, Agamemnon could not sleep (Il. 10.3); neither could Menelaus (Il. 10.5). The Iliad also suggests a more direct connection between the sons of Atreus and Sparta’s twinned kings: through mutual parallels to the Dioscuri. The whole story recapitulates the earlier rescue of Helen by her brothers Castor and Polydeuces, she having been abducted while still a maid by Theseus. And then, the text itself uniquely identifies the Atreid brothers with the Dioscuri as “marshals of the people” (kosmetore laon). Both Chryses and Achilles so refer to Agamemnon and Menelaus in Book 1, a designation not otherwise found except when Helen speaks of Castor and Polydeuces. This is in Book 3, as Helen attempts to descry the Lacedaemonian twins from the Trojan ramparts, not knowing they are already dead: Yet nowhere can I see those two, the marshals of the people, Kastor, breaker of horses, and the strong boxer, Polydeukes, my own brothers, born with me of a single mother (Il .3.236-238). Agamemnon together with Menelaus and Castor with Polydeuces may be the only “marshals of the people,” but they are hardly the only pairs of royal brothers in the Iliad. There were a number of such on both sides, both from Greece (including the islands) and from Asia Minor. Among the Argives, They who lived in Aspledon and Orchomenos of the Minyai [adjoining Boeotia], Askalaphos led these, and Ialmenos, children of Ares (Il 2.511-12).Schedios and Epistrophos led the men of Phokis (Il. 2.517).Idomeneus the spear-famed was leader of the Kretans, those who led Knossos and Gortyna [etc.]. . . .Of all these Idomeneus the spear-famed was leader, with Meriones, a match for the murderous Lord of Battles (Il. 2.645-51).They who held Nisyros and Krapathos and Kasos, and Kos, Eurypylos’ city, and the islands called Kalydnai, of these again Pheidippos and Antiphos were the leaders, sons both of Thessalos who was born to the lord Herakles (Il .2.676-79).They who held Argissa and dwelt about Gyrtone [in Thessaly]. . . . of these the leader was Polypoites, stubborn in battle, son of Peirithoos whose father was Zeus immortal . . . not by himself for Leonteus was with him, scion of Ares (Il 2.738-45). There may be a few other such pairs—the descriptions are sometimes unclear—and a few Argive groups were lead by three or more. While it is certainly true that those brought to Troy by a single king were more numerous, the characterizations of dual leadership are reminiscent of (other) mythic paradigms of twinned sovereignty, including the differentiations of senior and junior and references to descent from gods. Perhaps not too much should be made of the last, since in principle this was a war of heroes. Yet in at least one instance on the Trojan side, the description of dual sovereignty becomes classic: it could have been lifted from Herodotus on spartan kings. Among the Trojan allies there were the co-leaders: Adrestus and Amphius of Adrestia (near Troy); Hippothous and Pylaius (“scion of Ares”) of Pelasgian Larissa; Odius and Epistrophus from the shores of the Black Sea; Nastes and Amphimachus who were Carians of Miletus in Crete; and last-mentioned by Homer but not least for our purposes, sarpedon and Glaucus of Lycia in Asia Minor. If anything, dual leadership is more common on the Trojan side than among the Argive host. On the other hand, sarpedon and Glaucus, the most Spartan-like kings here, descend from Aeolus, son of Hellen, and more proximately from forebears who ruled in Argos. This if one takes the Homeric genealogy; by other variants, they are linked to the ancestors of the Peloponnesian Danaids, and to Minos and Crete. Crete historically had a strong Dorian component, and a constitution famously like Sparta’s (cf. Aristotle, Pol. 1271b20). In the long history of Classical scholarship has not someone remarked that the “godlike” Sarpedon’s speech, rousing his co-king Glaucus and the Lycians to the attack, describes a dual sovereignty that point-for-point matches Herodotus (6.5657) on the prerogatives of the Spartan kings? The scene is in Book 2 of the Iliad, before the defensive ramparts of the Achaeans, who had rallied there to protect the ships and temporarily halted the Trojan onslaught. Sarpedon now urges the Lycians to make a rush, addressing Glaucus: Glaukos, why is it you and I are honored before the others with pride of place, the choice meats and filled wine cups in Lykia, and all men look upon as if we were immortals, and we are appointed a great piece of land by the banks of Xanthos, good land, orchard and vineyard, and ploughed for the planting of wheat? Therefore it is our duty in the forefront of the Lykians to take our stand, and bear our part of the blazing of battle, so that a man of the close-armoured Lykians may say of us: “Indeed these are no ignoble men who are the lords of Lykia, these kings of ours, who feed upon the fat sheep appointed and drink exquisite sweet wine, since indeed there is strength of valor in them, since they fight in the forefront of the Lykians” (Il 12.310-321). The pride of place—presumably the seats of honor as well as the priorities in the feast—the choice meats and filled wine cups, the rich lands set aside on a body of water (recall the Spartan kings’ lake), the duty to lead in war, and all these the joint privileges of two kings whom men look upon as if they were immortals: the resemblances to Spartan kings are surely striking. When it is added from the genealogy recited by Glaucus, son of Hippolochus, that his senior co-king Sarpedon is the son of Zeus, the paradigm is virtually complete. Here is Homer’s genealogy, given by Glaucus before he is about to do battle with Diomedes (fig. 12): Figure 12. Genealogy of Glaucus (according to Homer) Bellerophon, the ancestor of the Lycian kings, was a Heraclean-like figure. Driven from Argos by the Danaid ruler Proetus over a marital contretemps, he was sent by the latter to his father-in-law, the king of Lycia. Bellerophon carried a secret message from Proetus to the Lycian king with instructions to do Bellerophon in. The Lycian king set him a series of dangerous tasks, the slaying of monsters, as Eurystheus did to Heracles. But when Bellerophon accomplishes these impossibles, the tasks turn out to have been marriage ordeals: the king accords him his daughter and the succession to the throne of Lycia. Bellerophon has three children by the Lycian princess. The first Isander was killed by Ares, the second Hippolochus sired Glaucus, the co-king of Sarpedon, who was the son of Bellerophon’s third child, a daughter Laodamia. The latter lay with Zeus and bore his son (Sarpedon). Between this and the ideal Peloponnesian genealogy of dual sovereignty, there is one significant difference only. Glaucus and Sarpedon are children of a human brother and sister, the senior Sarpedon being father’s sister’s son to Glaucus. If this were a Dorian succession, one would expect the two kings to be sons of the brothers Isander and Hippolochus. Rather, everything happens as if it were consistent with another social transformation for which the Lycians were uniquely known: they were, as Herodotus told, matrilineal “during the rule of Sarpedon.” They came from Crete, according to Herodotus, and in their customs they resemble the Cretans in some ways, the Carians in others but in one of their customs, that of taking the mother’s name instead of the father’s, they are unique. Ask a Lycian who he is, and he will tell you his own name and his mother’s, then his grandmother’s and great-grandmother’s and so on. And if a free woman has a child by a slave, the child is considered legitimate, whereas the children of a free man, however distinguished he may be, and a foreign wife or mistress have no citizen rights at all (Herod. 1.173). Hence, the permutation of the paradigm: isander is eliminated without issue, killed by Ares; and instead of his son, the child of the sister (and of Zeus) is superior to his human co-king, the child of the brother. Conclusion The mention of Crete, the distribution of Homeric dual kings in Asia Minor as well as mainland and island Greece, all this could lead to speculations about the historic sources of the spartan kingship. Such speculations seem to be a favored game of classical scholarship, as if the location of the origins, preferably somewhere outside Greece, would somehow be a satisfactory explanation of the spartan case. Some such external source, Phoenicia for example (Drews 1983: 81), would have the special advantage of accounting for the uniqueness of spartan kingship among the Hellenes, while at the same time it invokes the apparent self-evidence of explanations by cultural diffusion. Diffusion may be marginally better than “survival,” but for several reasons we would not follow these common historicist modes of “explanation.” For one reason, like “survival,” diffusion merely postpones the problem rather than resolves it. Just as survival tells us nothing unless we know the contemporary values of the institution, the meanings and functions that give it continuity, so diffusion must contend with the fact that borrowing is always selective, hence likewise depends on the cultural system of the borrowers. Even if it were true that sparta got dual kingship from somewhere else, why just sparta, and of sparta, just why? Besides, everything in myth looks like dual kingship of the Spartan kind was ancient and widespread in Greece, much more so than is commonly supposed. And in any event, the problem of dual kingship can never be solved by its historical traces alone. We have to understand its structural values: in the broadest sense its meaning in the given social context. From the meaning we can make more sense of the history than vice versa. We have tried to show that Spartan dual kingship is a mytho-praxis, endowing the Laconian sovereignty and its existential situation with a treasure-house of mythic values. It evokes famous exploits of conquest and hegemony, the usurpations of indigenous kings, and reminiscences of universal domination. Its structural features are imitations of the sovereignty of Zeus and implications of the dominance of Mycenae. When Athens, in opposing such pretensions, put forth the opposed ideology of autochthony, the two ideals of sovereign right could only reinforce each other. The contemporary political contest evoked a dialectic present from the beginning of human history. And the more the Athenians insisted on their own superiority as native sons of the soil, the more virtues the Spartans could find in an ancien régime of (immigrant) heroic kings. “Survival” can be applied to the doctrines of both sides, but the doctrines survived because they were functionally predicated on each other. The Athenian apologist Isocrates once remonstrated with the Spartans, that they ought to stop splitting hairs about leadership in Hellas. The Lacedaemonians, he wrote, have inherited the false doctrine that leadership is theirs by ancestral right. If, however, one should prove to them that this honor belongs to us rather than to them, perhaps they might give up splitting hairs about this question and pursue their true interests (Isocr. Paneg. 18). Yet for Spartans, splitting heirs was hardly splitting hairs. Their dual kingship was proof that leadership in Hellas belonged to them: by ancestral rights that went back on the human side to hegemonic kings of yore, and on the divine side to the universal sovereignty of Zeus. Bibliography Benveniste, Émile. 1969. Le vocabulaire des institutions indo-européennes. Vol. 2: Pouvoir, droit, religion. Paris: Les Éditions de Minuit. Burkert, Walter. 1985. Greek religion: archaic and classical. Translated by John Raffan. Oxford: Basil Blackwell. Daremberg, Charles , ed. 1892. Dictionnaire des antiquités grecques et romaines. Vol. 2, Part 1. Paris: Librairie Hachette. ———. 1904. Dictionnaire des antiquités grecques et romaines. Vol. 3, Part 2. Paris: Librairie Hachette. Drews, Robert. 1983. Basileus: the evidence for kingship in Geometric Greece. Yale: Yale University Press. Dumézil, Georges. 1948. Mitra-Varuna: essai sur deux représentations indoeuropéennes de la souveraineté. Second Edition. Paris: Gallimard. ———. 1949. L’heritage indo-européen à rome. Paris: Gallimard. Finley, M. I 1981. Economy and society in ancient Greece, edited by Brent D. Shaw and Richard P. Saller. London: Chatto & Windus. Hamp, Eric P. 1971. “The variants of Ἀγαμέμνων.” Glotta 49 (1-2): 21–24. Hooker, J. T. 1980. The ancient Spartans. London: J. M. Dent & Sons. Kantorowicz, Ernst H. 1957. The king’s two bodies: A study in Medieval political theology. Princeton, New Jersey: Princeton University Press. Lévi-Strauss, Claude. 1977. Structural anthropology. Translated by Claire Jacobson and Brooke Grundfest Schoepf. Harmondsworth, U.K.: Penguin Books. Marchant, E. C., trans. 1925. Xenophon: scripta minora. Loeb Classical Library. London: William Heinemann. Nilsson, Martin P. 1972. The Mycenaean origins of Greek mythology. Berkeley: University of California Press. Sahlins, Marshall. 1981. Historical metaphors and mythical realities: Structure in the early history of the Sandwich Islands kingdom. Ann Arbor: University of Michigan Press. ———. 1985. Islands of history. Chicago: University of Chicago Press. Ste. Croix, G. E. M. de. 1972. The origins of the Peloponnesian war. London: Duckworth. Vernant, Jean-Pierre. 1979. “A la table des hommes: mythe de fondation du sacrifice chez Hésiode.” In La cuisine du sacrifice en pays grec, edited by M. Detienne and J.-P. Vernant, 37–132. Paris: Gallimard. ———. 1983. Myth and thought among the Greeks. London: Routledge & Kegan Paul. ———. 1985. Mythe et pensée chez les Grecs: études de psychologie historique. Revised edition. Paris: Éditions La Découverte. Waites, Margaret C. 1919. “The meaning of the ‘Dokana.’” American Journal of Archaeology 23 (1): 1–18. West, Martin L. 1985. The Hesiodic catalogue of women: Its nature, structure, and origins. Oxford: Clarendon Press. Naitre jumelé avec grandeur : la double royauté à Sparte Résumé : Cet article examine les configurations dyarchiques par le biais d’une analyse détaillée de la double royauté en Grèce ancienne. Jumelés et inséparables, à la fois humains et divins, les rois spartiates étaient eux-mêmes vus comme descendants de jumeaux célestes. Dès lors, on peut penser la diarchie spartiate comme une réalisation empirique des deux corps du roi — la double royauté comme gémellité souveraine. Cet essai se poursuit par un examen des autres jumeaux royaux de la mythologie grecque, dont l’un était habituellement vu comme descendant directe d’un dieu, et avance l’idée que de tels mythes sur les origines dynastiques constituent une cosmologie du droit souverain par laquelle le mythe spartiate des rois-étrangers d’origine divine était opposé à l’idéologie athénienne de l’autochtonie. Marshall Sahlins is the Charles F. Grey Distinguished Service Professor Emeritus of Anthropology at the University of Chicago. He is the author of numerous books, including Islands of history, How “natives “ think: about Captain Cook, for example, and Apologies to Thucydides.

de Vienne: “Make yourself uncomfortable” This work is licensed under the Creative Commons | © Emmanuel de Vienne. Attribution-NonCommercial-NoDerivs 3.0 Unported. ISSN 2049-1115 (Online) “Make yourself uncomfortable”: Joking relationships as predictable uncertainty among the Trumai of Central Brazil Emmanuel de Vienne, Université Paris-Ouest Nanterre La Défense This article explores the shapes taken by joking among the Trumai Indians and other groups of the Xinguan Indigenous Park (Mato Grosso, Brazil). This social practice often opposes persons who are in open-ended relationships, and can thus be defined as the default mode of relation: while it occurs prototypically between male cross-cousins, it is also common with Indian or non-Indian outsiders. Contrarily, one demonstrates both shame and respect toward real affines. Identifying this system of attitudes gives no account, however, of either the pragmatic properties of joking, nor its specific social efficacy. Joking, in particular, is remarkable by its inescapable ambivalence, both moral and functional. This characteristic is closely linked to the frame of interaction that joking is built upon, which manages, in the same time, to both follow highly conventional patterns and produce deep destabilization. This paper thus tries to explain the paradox of what could be called a predictable uncertainty and convey, partly from my own experience, what it is like to be part of such a play where the opposition between the failure and success of an interaction becomes blurred. Keywords: joking, laughter, affinity, uncertainty, Upper Xingu, Amazonia “No, I really think I should stop. This is the tenth joke I’ve told from my country and no one is laughing at all. I’m starting to feel ashamed.” This was the very first thing I said that provoked laughter from the assembled participants in this peculiar joking session during which I tried to introduce my Trumai hosts to selected examples of French humor. Up until then, each joke had been met by a severe silence, followed by my painful attempts to make the joke more explicit, and then by a laconic response from the village chief: “Yes, it was indeed funny. Tell us another one.” Apart from showing—fairly predictably—that blonde jokes don’t fit into Amazonian cosmology, this situation (both the insistence that I continue and finally the laughter when I confessed my shame) allows us to address interesting questions regarding joking in the Upper Xingu, the multilinguistic society to which the Trumai belong. This anecdote is, of course, symptomatic of the endless gringo-bashing that characterizes the socialization of Euro-Americans in the region—a process that I will have the dubious privilege of exploring further over the course of the article. But why should laughter, rather than other forms of communication, be the principal tool of such socialization, and why is it so systematic and intensive in the Xingu? In other words, the overrepresentation of the ethnographer in the data about joking is precisely the proof that he occupies a perfect position to document this social practice and inquire into the wider scope of Xinguanian sociality. Joking activity in the Upper Xingu is closely linked to the cross-cousin relationship (amipine in Trumai), which perfectly fits Radcliffe-Brown’s famous definition: “a relation between two persons in which one is by custom permitted, and in some instances required, to tease or make fun of the other, who in turn is required to take no offence” (Radcliffe-Brown 1940: 195). According to his theory, institutionalized joking occurs as a way to stabilize relationships that combine elements of disjunction and conjunction. Some authors have tried, by dint of wide-ranging and systematic comparative work, to test his hypotheses in a Murdockian frame of analysis (Brant 1948, Alford and Alford 1981), but such functionalist approaches are widely criticized today. As Dan Rosengren has noted (2010: 104), other classical works on joking or laughter also tend to reduce it to a single or, at most, a few functions, be they social (Radcliffe-Brown) or psychological (Mary Douglas’ Freud-inspired anti-rite, Clastres’ anti-fear defense mechanism). Recent works have shown how joking might allow one to negotiate or comment on political relations (Canut and Smith 2006), inequalities (Morton 2008), and how laughter can be seen as linked to power (Lagrou 2006). Its ambivalence is also nowadays commonly accepted. In the introduction of a special issue on joking in the Anthropologgical forum, John Carty and Yasmine Musharbash insist, for instance, on the close link between laughter and social rupture: “Laughter is dangerous. Laughter is a boundary thrown up around those laughing, those sharing the joke. Its role in demarcating difference, of collectively identifying against an Other, is as bound to processes of social exclusion as to inclusion. Laughing ‘with’ some people usually entails laughing ‘at’ others…” (Carty and Musharbash 2008: 214). In Amazonia, the opposition between the positive and the disjunctive side of laughter also appears when comparing authors. Claude Lévi-Strauss’ analysis of myths that address the theme of laughter takes the side of those being laughed at. Its link to revenge makes it a potentially dangerous form of expression and therefore something that has to be controlled (1964). In contrast, more recent work on laughter in Amazonia has underlined the central role of shared laughter in conviviality, and therefore in processes of kin making (Overing 2000). Even Rosengren’s paper (2010) on absurdity among the Matsigenka—though it begins by underlining the polyvalence of laughter—ultimately reduces his analysis to a single episode that expresses conviviality. The aim of this paper is to give a wider ethnographic account of the social lines that both circumscribe and are drawn by joking, to look into “the intimate and ingenious play of individuals with and within the structures of their social world” (Carty and Musharbash 2008: 215). This implies drawing connections between social relations organized into a structured system of shared terms, proscribed forms of behavior, and actual interactions in all their complexity. The complexity of joking interactions stems first and foremost from the specific relationship they establish between communication and metacommunication, which is common to what Gregory Bateson defines as the “play” frame. Watching two monkeys playing, he observed that for him, as for them, what they do looks like combat but is definitely perceived as “not combat.” Therefore this phenomenon “could only occur if the participant organisms were capable of some degree of metacommunication, i.e., of exchanging signals which would carry the message ‘this is play’” (1972: 139). He further noted that this message creates a paradoxical frame of the Russelian or Epimenides type, which he gnomically glosses in the following terms: “these actions in which we now engage do not denote what those actions for which they stand would denote” (ibid.). Thus the playful act both refers to and differs from its literal version. Erving Goffman, through the notion of keying, significantly improved our understanding of how participants in an interaction define their attitude toward such a “primary frame” (1974). With specific regard to joking, Don Handelman and Bruce Kapferer (1972) have shown that it corresponds to two different frames: setting-specific and category-routinized. While the former relies on local resources found in the setting itself and is highly fragile, the latter is firmly rooted in social conventions and is therefore more resistant. In both case, their aim is to understand how participants establish, maintain, subvert, or destroy the joking frame. However, their analysis takes for granted that one always identifies the situation as a “joking” one, which is not always the case in the Upper Xingu. On the one hand, it is true that the notion of stable behavior, once stripped of its functionalist or teleological connotations, is a good description of frequently predictable and routine-like joking interactions between cross-cousins. But, on the other hand, it does not do justice to the tension that appears no less frequently between certainty and uncertainty, prediction and surprise. As we will see, the explicit intention of these interactions is precisely to embarrass and destabilize the other. For this purpose, they mobilize a more complex frame, also mentioned (but only in passing) by Bateson, which “is constructed not upon the premise ‘This is play’ but rather around the question ‘Is this play?’” (1972: 141). This significant change makes room for variety, strategies, and pragmatic modulations within an explicit and rigid system of attitudes. To say it “leaves room,” however, is not really to do justice to the shift involved, which takes as its frame the very lack of an obvious frame. In so doing, as I shall show, it places uncertainty at the core of the social productivity of joking. Shame and respect The Trumai live along the tributaries of the Xingu river (Mato Grosso, Brazil). The area was first described by Karl von den Steinen toward the end of the 1880s (von den Steinen 1886, 1894). After a period of relative isolation, it began to be visited with increasing frequency, notably by the Roncador-Xingu expedition guided by the Villas-Boas brothers. They created the Xinguanian reservation in 1961, access to which was restricted to anthropologists and medical officers. The “Xingu Indigenous Park” is divided into a northern and a southern area. While the former offers no cultural unity, the latter, commonly called “Upper Xingu society,” consists of ten groups representing four linguistic families: Kamayura and Aweti (Tupi); Kalapalao, Kuikuro, Matipu, and Nahukuá (Carib); Yawalapiti, Wauja, and Mehinaku (Arawak); and Trumai (a linguistic isolate). The Upper Xingu is characterized by a certain degree of cultural similarity, and is doubtless the result of a long and ongoing process of integration, which both links the units (through marriages, economic ties, ritual, and observance of a shared ethos) and preserves them as they are, at least linguistically speaking (Franchetto 2001). Archaeological data shows that this grouping has an Arawakan substrate that succeeded in integrating and transforming (as well as being transformed by) successive arrivals of Caribe, Tupian, and Trumai groups (Heckenberger 2001). The Trumai were the last to arrive, in the middle of the nineteenth century, and, as a unit, they still occupy a marginal position in local society. Culturally fragile since they ceased ritual activities more than twenty years ago, and losing their language, today they perceive themselves as being in a critical situation (de Vienne 2011). Though excluded from the ritual cycle, the Trumai still participate in everyday Upper Xinguanian sociality. Radcliffe-Brown’s insight that one has to consider the joking relationship as part of a wider system of behavior remains valid. In particular, as it is a relationship of “permitted disrespect,” our theorizing concerning it “must be part of, or consistent with, a theory of the place of respect in social relations and in social life generally” (1940: 196). As mentioned in the introduction, in the Upper Xingu, joking and teasing occur primarily between cross-cousins (amipine in Trumai), that is to say virtual affines (see figure 1).1 This contrasts with behavior toward actual affines, which is characterized by marks of respect and shame. In Trumai these actual affines are even called ha falτïts chïk, “those toward whom I feel shame.” This shameful relationship is always mutual and, when it exists between people of different generations, also asymmetrical, as the younger partner shows more shame and therefore more respect than his elder. While in many Xinguanian languages, shame and respect are both referred to by the same term, among the Trumai falτï (shame) is different from oxe (respect). If the former involves the latter (shame is a display of respect), the reverse is not true. Some kin are said to be respectable without being treated with shame. Figure 1: Shame relationships. Shame relationships invariably involve avoiding the other’s name, restraining from laughter and joking and, to varying degrees, physical avoidance. When permitted (between same-sex in-laws) or absolutely necessary (between generations), forms of address must be rendered oblique. This can be done by teknonymy and also—between same-sex in-laws—by the terms of address inetl ka’am (brother-in-law) or inatletl ka’am (sister-in-law), contractions of “it is to him/her that I am talking.”2 It is between Ego and in-laws of the opposite sex at G+1 that the shame is strongest. In the past, Trumai men even had to head into the forest when they saw their mother-in-law on the path to the river in order to avoid meeting her. Today they avoid close copresence and eye contact. The Trumai’s comments on these rules are strictly faithful to Radcliffe-Brown’s (1940) theory, according to which institutionalized shame appears as a way to stabilize a relationship that combines both conjunction and disjunction. In their own words: “You can’t argue with someone you don’t talk to.” Shame relationships thus constitute an interaction frame of a special kind: embodied as an emotion, they prevent interaction explicitly to preserve the relationship. Apart from defining a kinship relation, shame is a key moral value for the definition of Xinguanian identity. For example, Ellen Basso defines the ifutisu feeling among the Kalapalo as “respect for one another” (1973: 7) between kin, and more generally as a moral stance that refuses aggressiveness and favors continuous displays of generosity. To publicly accuse someone of stealing, even when true, shows a lack of ifutisu, while accepting a bad bargain shows that one possesses this quality to a large extent. Therefore the ideal man is ready to sacrifice his own good for the sake of peace, proof that he has “shame.” Puberty seclusion aims not only to produce beautiful bodies (desirable, which for women are fit for procreation and for men are fit for traditional wrestling), but also to generate this feeling of shame that children lack. Young persons are isolated behind a wall in their house for several months (up to two years for future chiefs who are supposed to possess these Xinganian qualities to the highest degree). During this period, they avoid sunlight, cannot see anyone that is not a close consanguine, and should only talk when absolutely necessary and in hushed tones. These behavioral rules are converted into an actual feeling of shame when they emerge from seclusion. During the emergence ceremony, the secluded adolescent is exposed on the central plaza, in broad daylight, usually during an important intertribal ritual. Women, who have not cut their hair since their first day of seclusion, exhibit a long fringe that hides their eyes. This veil is then publicly cut, an act that gives its name to the end of seclusion ceremony: “the cutting of the fringe” (puk naha). Briefly put, shame and respect in the upper Xingu extend far beyond the sphere of kinship. They are the moral marks of humanity, which are used to define the Xinguanians as opposed to savage (bravas) tribes and white people. Laughing with one’s cross-cousins, therefore, cannot but make constant and implicit reference to these respect and shame relationships. Jokes are funny if only because they are acts of subversion, pragmatic comments upon, or ironic décalage vis-à-vis shared knowledge of what it is to behave properly. Indeed, this is how Murray Garde interprets joking relationships (associated with virtual affinity) among the Bininj Kunwok in Australia; that is, as “the inversion of constrained behaviours to index pragmatically the absence of actual affinity” (Garde 2008: 237). The driving force of humor is thus defined negatively, as the relief felt at “the breaking of conventions” (ibid.: 248). These considerations, however, imply that joking, though creative and inventive, consists in letting off steam, in following one’s “natural” urges. As illustrated by the systematic references in the literature about joking to invisible lines and boundaries to be crossed (funny) or not to be crossed (improper), joking better appears as a positive performance, somehow perilous, which requires the mastery of certain positively valued skills. The cross-cousin relationship: How far can you go too far? The amipine (bilateral cross-cousin) category includes a large number of individuals, since in the Upper Xingu everyone in Ego’s generation is, for the purposes of addressing them, either a classificatory parallel cousin considered a brother or sister (ha pisi in Trumai, “brother/sister,” or “cousin brother / cousin sister” in Portuguese), or a classificatory cross-cousin (simply called cousin in Portuguese). The types of behavior associated with this cross-cousin relationship vary significantly depending on gender. It is principally male cross-cousins who engage in joking behavior. Female cross-cousins avoid making jokes at one another’s expense. Cross-sex cross-cousins have, qua potential spouses, the option of flirting with one another, which may take the form of gentle teasing but this largely takes place in the absence of the anthropologist. The ability to joke is also partly linked to age. Adolescents of both sexes who have just emerged from seclusion are too stricken by shame to be able to joke publicly. This ability is only acquired little by little and normally much later for women. Joking, however, remains mainly a masculine activity since it is paradigmatically associated with relationships involving alterity, a domain controlled by men. For this reason, and also because I had less access to the feminine world, this paper focuses primarily on interactions that involve men. Between male cross-cousins, exchanges involve systematic ribald comments, feigned aggression, horseplay, teasing, and in the past even a form of “goosing” or playful penis-grabbing. The cross-cousin is thus the closest other, and this relationship cannot be reduced to virtual affinity as its productivity goes far beyond matrimonial exchange. Javari festivities, for instance, involve an important intertribal ritual partly structured around the amipine relationship. It involves dueling with spears and insult contests opposing amipine from different villages. This relationship is also close to that of friendship, which is evident in reflexive comments about it as well as in the etymology of the term: -pine means “friend, partner.” This behavior is reflexively theorized. In Trumai, it is called either taraiw or hopeta. When joking involves speech, it is called ami taraiw (talk joking), though one cannot say ami hopeta. This detail aside, the meaning of both verbs is the same: joking or mocking. They can both be used in intransitive as well as transitive constructions, each construction involving a semantic nuance. Ha a hopeta or ha a taraiw (1Ps dual joke) means, “we both joke,” in order to define the nature of the relationship between the speaker and his cousin.3 Here joking is reciprocal. Ha hopeta / taraiw ine-tl (1Ps mock 3ps-acc.), on the other hand, means, “I make fun of him” or “I mock him” and therefore involves an asymmetry. This ambiguity is at the heart of the relationship since its mutual and symmetrical dimension is not given, but results from constant dueling. Officially, one has no choice but to accept cheerfully that one is the target of such mockery, and I frequently heard sentences like “you have to know how to [take a] joke,” “you mustn’t take offence,” or “this is what cross-cousins do,” and so on. Perhaps the paradoxical nature of the joking relationship frame should be clearly underlined. On the one hand it is shared by everyone and completely explicit, and involves to a large extent conventional patterns, themes, and lines. As noted by William McGregor, many writers have insisted on the fact that jokes tend to become conventional, and ultimately involve “extremely formulaic exchanges” (McGregor 1989: 86). On the other hand, the content of the explicit frame contradicts the very notion of frame, saying more or less, “Do what you want and be ready for anything, but don’t get mad.” In what follows, I explore the nature of the uncertainty produced by this duality by showing that it rests less on invention or creativity than on daring and the dissimulation of intention. Ribald wordplay is extremely common, often relying on the combination of more or less standard tropes. On one notable occasion, an elder described his mistress as “my sperm’s clothing.” Nowadays, however, humor tends to be derived from exposing the other to shame by setting him up for and then highlighting a (homo)sexual double entendre. This is mainly done in Portuguese—the dominant language in intertribal interactions—and between men. Traps are frequent—for example, “you like to suck?” (speaking of fruits), to which one carefully avoids giving a positive answer. This form of joking is so automatic among young people that they carefully avoid the verb “to give” (dar) because it has the figurative meaning of accepting sexual intercourse with a man. These examples can be examined according to classic theories of humor, as summed up by William Beeman (2000). Humor always implies the setting up of a surprise or a series of surprises, among which the most common is “incongruity,” defined as follows: A communicative actor presents a message or other content material and contextualizes it within a cognitive “frame.” The actor constructs the frame through narration, visual representation, or enactment. He or she then suddenly pulls this frame aside, revealing one or more additional cognitive frames which audience members are shown as possible contextualizations or reframings of the original content material. The tension between the original framing and the sudden reframing results in an emotional release recognizable as the enjoyment response we see as smiles, amusement, and laughter. This tension is the driving force that underlies humor, and the release of that tension—as Freud pointed out—is a fundamental human behavioral reflex. (Beeman 2000) Humor then consists in a double framing, laughter arising when the audience grasps the new contextualization of a previous utterance. If someone says, “I gave it to him,” somebody else might respond, “then you gave [put out], huh?” The pragmatic indicators (pause, smiles, blinking, and glances to others) allow to reframe the verb “to give” in its sexual meaning. Wordplay, however, is not the primary resource for joking. Since the main goal is to embarrass the opponent, one can use other ways to reach it. There will still be a process of double framing that generates tension and release, but it will concern the interaction itself, not a particular linguistic content. Cousins frequently play with kinship terms. The idea is to imply by the use of such and such false kinship term either that the addressee is older than he is (grandfather, uncle) or that one of his female consanguines is attractive (brother-in-law when it is his sister, father-in-law when his daughter, etc.). This is a pragmatic reference to other possible relationships and it underlines the openness of joking relationships, although such mockery is not really unsettling. Infidelity, on the other hand, provides an endlessly rich and destabilizing vein of humor, which is the driving force behind one of the most conventional and efficient embarrassment techniques. It occurs when two cousins meet in the presence of one of their spouses. The unaccompanied cousin will almost inevitably begin to relate his cousin’s infidelities to the latter’s spouse. All the actors involved recognize this as hopeta, and indeed most of it is pure invention, and is seen as such. Most of it, because the trick is to slip in actual examples of infidelity. Cousins frequently travel together in the park or to the city for political meetings, healthcare or school seminars, and so forth. During these stays away from home, they may mutually confess their good fortune with this or that Indian “cousin” or white nurse. They may also go to brothels together. Cousins, then, are frequently yutu, a term used to describe people (male or female) who have or had a common lover. The wife’s attitude when faced with these revelations must conform to the local ideal of emotional control. Jealousy, even with good grounds, is seen as improper behavior. This is why, most of the time, she says nothing and shows no reaction beyond raising her eyebrows, which might equally mean, “I am no fool, I know perfectly well,” “I don’t care,” or “I don’t believe a word of what you are saying.” The consequences, however, can be serious. The calmest wife, once she finds herself alone with her husband, may unleash her rage. A young Trumai told me for instance: “I used to joke with X and his wife, since he is my cousin. But I completely stopped. Once, I did that brincadeira (joke), you know, I said that he went with me to visit Y in Morena, I revealed everything, but not seriously, you know. I heard that soon after I left, she took his 22 (rifle) and pointed it at him, loaded, saying ‘you cheating bastard, I’ll kill you.’ I was shocked when I heard that [laughing]. He almost died because of me. Since then I’ve keep quiet.” The fun of the situation, even if conventionalized, is that uncertainty surrounds how far the teaser is prepared to go, how the victim will control his reactions, and whether the wife will say anything at all. As for the victim, one can well understand that the interlarding of true stories and false ones makes it all the harder for him. So long as they are false, he feels relatively safe, but when the speaker inserts a true story into the mix, he must make sure not to react differently than before, despite fear of his wife’s reaction and anger at his friend’s betrayal. Frequently there are more than just these core participants present. Joking is better with an audience, the idea being to expose the victim as much as possible. Observers are not passive, at least those who are not in a shame relationship to the core participants. They join the teaser in frantically blinking, with both eyes, as a sign of connivance, encouraging him to continue. They also refrain from laughing but in a conspicuous way, and eventually they enter the game. The goal is collective laughter, which breaks out when the husband, or his wife, finally reacts. The public act of denunciation is nothing more than the patient construction of an emotional tension, during which expectation grows and collective attention gets focused with more and more intensity on the target couple. The man may ultimately lose his temper in an exaggerated manner (though while still smiling or laughing) or he may vigorously deny the accusations. The wife may calmly state that she knows how worthless her man is or explain in a sad tone that he indeed wants to abandon her. In any case the reaction is the signal of a collective release. The session can proceed further following the same pattern, or shift toward more serious subjects. The role of the audience in this precise case is emblematic of joking or teasing events in general. The public obeys what we might call a “nonassistance clause” toward the victim. For the sake of the joke, even close consanguine refrain from helping and will eventually become accomplices. The motivation for this relentless mockery is rarely cruelty. These interactions have nothing to do with bullying either, since they do not intend to create recurrent scapegoats. Alliances are highly fragile and temporary. They can change in the blink of an eye, and the audience will apply the same rule of mute or active participation against the new target. It is true, however, that repetitive successful joking on the same theme and with the same people can have a strong impact on the victim’s reputation. I witnessed many cases where a joke ultimately morphed into gossip. Uncertainty about the joker’s intentions is also at the core of practical joking. During my most recent fieldwork in July 2011, an old Trumai told me about some of the good times he had had with an amipine who had died a couple of months before. Amati: The guy who died was a great amipine. Kamayura. One day he invited me to fish: “Hey makra [cross cousin in kamayura], let’s go fishing, you never ate something I fished or hunted.” I said, “Yes, true, nor did you eat something I killed. Let’s go.” “I think you never kill anything, that’s why.” “You never saw me, and I never saw you, that’s all. We are even then. Let’s go tomorrow, we’ll see.” So we went the day after. He asked: “who is going to fish, to kill fish with arrows?” I said, “You will, I will steer the canoe.” “Just because you suck when it comes to killing fish!” he answered. “No, if you appear to suck, we’ll switch places.” “Ok then.” There was a lot of “turtle’s food,” you know [name of a plant that grows in the water]. When the first [turtle] appeared, he missed it. Later, downstream, he missed another one. Pa! He missed three. They were really big! I was just watching in silence. He missed one more, then I said, “You are so bad at it, get ready to fall in the water!” “Don’t do that!” he answered. He missed and missed again. Then I saw the water was shallow. When he missed it, [he enacts the brutal paddle move he did] Tsum! He fell in the water…. “You are right, I keep missing, I deserved to fall,” he said. “Step in,” I said. He killed four or five then…. Then we made fire and ate turtle. When he finished he told me: “Amati take my arrows and my gun to the canoe.” “I am not your servant,” I answered. “No!” he said. “It is just to help me!” “Just this time then,” I said. I took them. He stayed behind to defecate. “You won’t be long?” “No I won’t.” Waiting for him I decided I wanted to test him, see how he would react. I said: “Man!4 “Yes?” “Bye bye! You can go back by foot! I am leaving.” “Don’t do that! I am just finishing!” “It is too late, I’m already on my way.” Then he shouted: “Amati! Amati!” “Yes?” I answered from downstream. “Don’t do that! come back!” “No, you can walk or swim back,” I said. “oh, I don’t know why I didn’t keep my rifle. I would have shot you in the ass!” Then I remained silent. Then I said just “ciao!” “Don’t do that or you’ll see, as soon as I get there I will beat you up!” “We will be even then, since if you beat me I’ll beat you as well, ciao!” Then he cried: “Amati! Amati!” [He imitates a tearful panicking voice and obtains laughter from his audience, two of his grandsons and me] Then I felt pity for him. “You shit in your pants out of fear, so I won’t do that to you.” “I really thought you were going to leave me. Just because you are amipine I won’t fight with you now. We came to fish, not to joke.” “You’re right.” It is like that. Poor him, he died. This narrative offers a good illustration of the mix of rivalry, mockery, and friendship that characterizes the amipine relationship. It consists precisely in demonstrating the opposite of kind and proper behavior: Amati acts offended when the cousin asks him to help him with his equipment, impatient when he is supposed to wait for him, angry when his cousin misses fish or turtles. The point is clearly to bring things just to the point where the other begins to take the other one seriously. This is especially the case when Amati pretends to leave his cousin on the bank. Only when the victim cries and shouts does he decide to stop the game and relieve the tesnion. The narration of his cousin’s panic is the high point of the tale, showing that the joke was a success, at least from Amati’s point of view. Joking interactions with cousins, in sum, could be described as composed of three steps. The first is a given, it is shared knowledge about the nature of the relationship: a joking one. The second is the utterance or gesture that manifests hostile intent (betrayal, offense, aggression, etc.). Teasing somebody about his infidelity must necessarily call into question the idea that the joker is “only” joking. He might actually want to betray you and tell your wife the truth. Mock aggression, feigned offense, and anger tend, in the same way, to break the joking frame itself. The third moment is metapragmatic. It consists in reframing the intention as fake, through laughter or explicit reference to the initial frame. This can be done by the offended, in a laughing tone (“You call that being a friend, eh?”) or more frequently by the offender (“I was just kidding”). In short, joking amounts to trying to convince the other you are not joking before saying that you were. The social performativity of joking The idea of a stable or conventional set of jokes linking a precisely defined set of persons according to kinship is, then, quite far from doing justice to the flexibility of the system. As for the content of the jokes, we have seen that even the most robust patterns can generate both good, clean laughter and deep embarrassment. On the participants’ side, one must acknowledge that joking is socially performative, and not merely a peripheral attribute of prior relationships. Pushing the inversion to its logical conclusion—and at the risk of exaggerating—one might say that cousins don’t joke because they are cousins, they are cousins because they joke. Of course, actual affinity and to a lesser extent consanguinity constitute objective limitations, but there is plenty of room for pragmatic manipulation. This manipulation is also present in relations between cross-cousins of opposite sexes. In such cases, the concerned parties can choose to think of themselves as sharing a sibling relationship or, alternatively, as remaining cross-cousins. This choice often boils down to a question of sex and refusals often take the form of kinship terms (“no, brother”). It is even possible to temporarily transform a consanguine into an affine, as one man intimated when he explained to me that one of his regular mistresses was “sister by day, cousin by night.” But to return to relationships between men, it is worth noting, first of all, that not every person considered amipine is equally fit to be a joking partner. As we saw in Amati’s narrative, some are “great cousins,” whom one especially enjoys laughing at or with. Some others are treated like siblings. Therefore the ideal amipine relationship can be produced to varying degrees. Second, joking is not limited to cousins. Within one’s kin circle, consanguines from Ego’s generation, and even from a different generation (to whom one ideally shows respect [oxe]) can also be teased or mocked but in a milder manner. This, however, is not officially recognized, as can be seen in this brief exchange with Amati: Me. One only jokes with amipine? Amati. Exclusively. Me. With one’s father, too? Amati. No. Me. Come on, as if you’ve never joked with your son… Amati. Which one? Me. Pedro. Amati. Ah, yes, we do joke. Some joke. Some joke; others don’t. As a matter of fact, the triangle formed by Amati, his elder son Pedro, and myself very much respected the pattern of the cross-cousin relationship. When one of us confessed some embarrassing anecdote to another, it was fairly certain that it would be revealed to the third person when we were all present. Even joking between brothers-in-law is possible, under specific conditions. A Trumai woman married a Kayabi man—Kayabi is a northern tribe that is not part of the Upper Xingu society, and has a different system of attitudes. Kayabi, unforgivably, joke with their in-laws and say their names out loud. This particular man, however, decided to show respect to his wife’s family and applied the Xinguanian rule. One of his brothers-in-law, however, was failing to reciprocate, and so one night, after a couple of drinks, he decided to address the problem. The brother-in-law in question reported the conversation: I really like him. He is really funny, I knew him before, we joke a lot…. One day he came to me, he was a little drunk. He said to me: “Hey, why do you always use my name? You don’t respect me! I don’t know why. None of your brothers use my name! Leandro doesn’t use my name, Edilson doesn’t use my name, Renato doesn’t use my name…” And at the same time, he was using all their names [he smiles]. “But you keep using my name! Why so? You don’t like me?” Then I kept serious. I looked at him and I said, “No I don’t like you.” [he makes a serious face]. He was stunned. He asked: “But why? Did I do something?” Then I said: “I am no boiola [gay, depreciative] who likes men. I like women.” Then he laughed, “Ka ka ka ka! You got me!” We joke a lot, I like him. This anecdote confirms the reflexive nature of joking: what is funny is to play around with whether or not the participants are joking. It also shows the openness of the system. This in-law was at once too distant to be respected (he couldn’t, in spite of his efforts, avoid his in-law’s names) and too good a joking partner to be dismissed as such. His wife’s brother thus decided to confirm him in this status of cousin-like brother-in-law. At the other end of the social spectrum, joking appears as a systematic interaction frame with native or nonnative outsiders. It is used in particular as a means of integrating the other. As we saw in the introduction, the anthropologist, in particular, may be submitted to severe treatment. Since Buell Quain, the American anthropologist who did fieldwork in 1938 among the Trumai, five researchers have worked or are working with them, a number that is sufficient to identify some regular patterns and to account for idiosyncratic behavior. The anthropologist, in addition, is only one of the faces taken as otherness for the Trumai. Joking with him or her is similar to joking with a Brazilian neighbor in the city, a doctor or nurse in the Park, an NGO employee, etcetera. In these cases, embarrassment and teasing rely primarily on the absence of shared knowledge, in particular regarding myths and language. The first type of joking consists of making the other repeat in Trumai sentences he cannot understand. As a consequence, the first lexical domain a foreigner will learn is the anatomy of male and female genitalia, the adjectives to describe their textures and smells according to age, and the metaphoric forms they can take in insults. What is seen as comic in these interactions is that the outsider says something embarrassing for himself like “my penis stinks.” After a few variations, when the laughter loses its intensity, someone else may intervene to help him. They communicate defensive lines, in a low voice, though clearly audible to everyone, that involve turning the line back on the offender with necessary alterations (e.g., “your vagina stinks”). Embarrassment ceases to be the only trigger of laughter, and is replaced by insult dueling that is very likely to escalate. Giving the stranger a Trumai name is also a systematic joking resource. The nicknames (which may last only as long as the utterance or which may become definitive) usually make reference to a mythical character via behavioral or morphological similarities. Here, the comic effect relies on the reframing of the situation through mythical knowledge. My Trumai name, Walatu (“frog”) was given after a couple of nights during which I stayed up late reading with my head torch. Walatu is a mythical character who similarly put glow worms on his eyelids in order to hide the fact that he was sleeping. He thus avoided being killed by Jaguar, who was waiting for him to fall asleep. The fact that the addressee does not know the cultural reference allows for efficient and effective humiliation. When the name is particularly inglorious, they try to make the fool think it is a proper Indian name. It usually works the first time, as the anthropologist can hardly resist this symbol of acceptance into the native social world. Only later is he or she informed of the meaning. This was the case for me with Ororok, a mythical character famous for his physical weakness and therefore systematic defeats in wrestling. Sometimes nothing motivates the choice of the name but its infamous or obscene connotations. Thus a nonnative female visitor was once given the name Werap, protagonist of a myth about female sexual frustration and frantic masturbation. The episode is still recalled with nostalgia. Mockery of strangers, then, relies on the same process of transformation of embarrassment into collective laughter that we saw with same-sex cross-cousins. Indeed, in the multilingual crucible of the Upper Xingu, the possibilities of linguistic incomprehension are also exploited in intertribal exchanges, and not only in those involving whites. That said, there are still a number of significant differences. The feigned hostility proper to cross-cousin exchanges cannot, for instance, be replicated in interactions with whites, at least not early on. Instead, one exploits the humoristic potential of integration and rapprochement: giving the outsider a local name, as we saw above, or touting the possibility of unlikely sexual liaisons. For example, a young female anthropologist spent last February in Canarana working with Amati, now aged over 60. She used to appear around noon in his house, a time when he had usually just got up and bathed. But one day, he was still in the bathroom and she had to wait. Amati’s son Pedro and sister Kumaru started to tease both her and Amati. They urged her, when Amati finally appeared, to say certain sentences to him in Trumai. She accepted the game. Pedro first taught her: “Are you up, my tender one?” (Aya hi lakida, ha pudits chïk?). Then Kumaru added a new version: “Are you up, Amati my lover?” (Aya hi lakida, ha el Amati?), which was finally transformed into: “Are you up, Amati’s pubic hair?” (Aya hi lakida, Amati honkos?). This last ribald comment illustrates the logic of gradation common to a teasing session. The plan was successfully executed. It was both a comment on this researcher’s constant desire to spend time with Amati, and on Amati’s tendency to sleep late in the morning. Even though I was in France at that time, I was roped into this session. Knowing that she was planning to contact me on Skype, they never told her the meaning of the lines, so that she would have to ask me. This would recreate, even in their absence, the discrepancy of knowledge and thus the slight embarrassment that characterizes Trumai humor. Humor involving non-Indian visitors, then, revolves around the performance of a sort of integration or rapprochement, but in such a fashion as to suggest that such a rapprochement is in fact impossible. Indeed, it is not possible to joke about a sexual liaison between an outsider and a Trumai of the same age, precisely because such a thing is possible. Similarly, the Trumai names given to anthropologists are never authentic anthroponyms, which are cognatically transmitted via alternating generations. Here, the humor integrates and draws the anthropologist in, but also maintains a minimal distance that can never be crossed. This is not to suggest that these two types of joking relationships (between male amipine and between Trumai and outsiders) are in fact radically different in nature. Rather, they constitute two poles of a continuum. Indeed, it is only a matter of time before the joking frame gets established with the stranger. Afterward, if he appears to “know how to joke,” they may push the cursor toward the “great amipine” level and pretend to take his belongings and leave him behind when traveling, or mention greeting him with a club for a cannibal ritual. Joking is thus a powerful tool of social integration. One observes whether the visitor is ready to accept cheerfully ironic comments and teasing, ready to learn how to reciprocate in order to create “sweet” laughter (tsi pom). The opposition, frequently noted, between laughing with and laughing at—the former being convivial and the second dangerous—is therefore unable to seize the dynamics of Xinguanian joking. The aim is precisely to accept both laughing with and being laughed at, and at the same time. The second reason the line between amipine and outsiders should be blurred is that the Upper Xingu is an open system, which incorporated former enemies and “made them human.” This transformation took place through ritual, trade, marital alliance, but probably also through humor. The Xinguanian cross-cousin is then the proximate and partially pacified incarnation of a form of generic alterity that can take many forms and that is thought of in terms of affinity. This is a common characteristic of Amazonian sociality, as Eduardo Viveiros de Castro has shown. 5 In many societies, affinal terms are used to describe relationships that go far beyond actual affinity, such that affinity can be seen as the “given dimension of the cosmic relational matrix” (2001: 21), which Viveiros de Castro proposes to call “potential affinity,” “in order to distinguish affinity as a generic value from affinity as a particular type of kinship tie” (ibid.: 22). Conversely, consanguinity is not seen as given: it must be actively produced by stripping away the initial part of affinity that was embedded in the relationship. Affinity, even when it stops being virtual, thus always remains ambivalent. For example, in Tupinamba tovajar means both “brother-in-law” and “enemy,” and it “expresses friendly alliance as well as mortal enmity outside, and certainly the other way around. It approaches and opposes in the same movement” (2002: 408–9). In Trumai, wayar, which is of Tupian origin, designates both enemies and the parents of one’s son-in-law or daughter-in-law. A recent paper by Carlos Fausto (2012) explores the friendship relationship (pajé) among the Parakana, and shows how it is “defined by the very fact of containing contradictory connotations,” such as hostility and intimacy, self and otherness. The friend, as he sums it up, is “the nearest enemy, the prey closest to hand,” and it is indeed possible to actually kill one’s best friend. Even if the Xinguanians rarely reach such extremes, one finds just the same duality in the cross-cousin relationship. What I aim to show is how it is established in the here and now of interaction. Even the closest cousin is reestablished as a stranger during the joke. What is important, apparently, is not the absence of enmity; it is the moment enmity is transformed into friendship. The dominant joking pattern—that of feigned hostility—expresses the constant enactment of this process of familiarization. Friendship is defined negatively, as the discovery that the other, surprisingly, has no hostile intention. Laughter is more than release; it is relief. Amati and his Kamayura cousin offer one last striking example of this dialectical move: Another day, this guy who died, he said to me: “Let’s have a pee, don’t you feel like you need to pee?” “Let’s go.” “I have a story for you,” he said. “Ok.” So we went outside. He started telling: “I heard that the Trumai who live in Diawarum are intending to kill all the Kamayura. But the Trumai are nobody,” he said. “How could the Trumai finish off the Kamayura? They are not men enough to do that. They aren’t. It’s the Kamayura who are real men.” He was peeing all the while. When he was peeing I pushed him, “tuf!” “It is the Kamayura who aren’t men!” I pushed him like that. “Wait a second! I’ll beat you up!” He slipped. “Tchuk tchuk tchuk!” He almost fell, stood up again, saying, “Wait a second! Wait a second!” [He laughs]. This was the last joke I had with him. It was in Pavuru. If co-peeing between men can be given as proof of intimacy, what we witness here is two intimate enemies joking about their enmity. The Kamayura are, for the Trumai, their “true enemies” (wayar pi’tsi), which also means their “true affines.” These two sides of the relationship are equally historically important. Systematic marriages over the last 150 years (many of them through intimidation) and episodes of coresidence have even had significant effects on the Trumai language, which has incorporated many Tupian words, including, ironically enough, wayar. In the same period the Kamayura repeatedly tried to “exterminate” the Trumai, threatening to attack and sometimes actually attacking their village, a process the Kamayura would describe as retaliation for witchcraft attacks. The perfect setting for a good laugh… One can now clearly see that joking with amipine and joking with outsiders are logically linked. In the first case, closeness is a given—because of the existence of real genealogical ties, actual intimacy, and a relationship stretching back to childhood. Indeed, they could have elected to consider each other as siblings, a safer but far duller course of action. The point then is to destabilize the relationship and reaffirm its inherent alterity by feigning enmity. In contrast, in the second case the starting point is radical difference and the humor, as we saw in the case of Amati and the young anthropologist, relies on suggesting an impossible alliance. In both cases joking preserves distance at the same time as it negates it. It is simply that depending on one’s point of departure, one either feigns to sunder so as to unite or feigns to unite so as to sunder. Joking, precisely because of the uncertainty it generates, is the means by which one establishes the “proper distance”—it is a pragmatic tool for bringing closer or distancing as the case may be.6 With the Kayamara, we have a situation of real affinity that is constantly transformed into ambiguous enmity, and with the anthropologist, we have a relationship of mistrust that can be overcome by aping an improbable affinal tie. Moral ambiguity It is not surprising if this ambivalence is reflected in reflexive moral comments on joking. On the positive side, joking skills are highly valued. The collective laughter they can produce is, just as anywhere else, a powerful tool for conviviality and social bonding. The inscription of successful joking events in the individual and collective memory, however, has not frequently been remarked upon in the anthropological literature. The first example given above, in which Amati recalled his best tricks with his late amipine, offers a good example. The beginning and the end of the narrative are references to his recent death (“poor him, he died”), and define it as a special type of eulogy. The fact that Amati (at least according to his memory) always gained the upper hand over him certainly contributes to his regret. But I noticed on numerous other occasions that a particularly successful joking session constituted a sort of hallmark of the relationship between its participants. It was endlessly reenacted and voluntarily recalled in later encounters, always accompanied by declarations of nostalgia. Old Kumaru was a dedicated joking teacher for me and this greatly contributed to my integration into her family during the very first weeks of my fieldwork in 2003. In 2011, during my most recent stay, our reunion took the particular form of retrospective teasing. It was around 11:00 PM, shortly after I arrived in the city of Canarana with three Trumai men from a distant village. We came to her house, where she lives with her daughter and two grandchildren. Her older sister Muchatari, the oldest person in the Trumai community, was also present. Kumaru, to the audience, in Trumai. He was scared of me. He was always scared. Muchatari, in Trumai. Scared that you would rape him, the little boy. Kumaru. Yes, scared I would rape him. Muchatari. Scared you would cut his growth. [During seclusion, young men are not supposed to have sexual intercourse, otherwise they will stop growing. End of seclusion is even called “cutting the growth.”] Me, in Portuguese. I did not understand that part. Would you translate for me Karu? Karuwaya, in Portuguese. They are saying she used to scare you, and she keeps laughing, “Walatu was frightened, thinking I was really going to grasp him.” Me, laughing, in Trumai. That is true. [This confession causes a collective outburst of laughter.] Muchatari, in Trumai. See!! He was scared. See, he’s telling you— Karuwaya. [at the same time] Waaa! He said it! Kumaru. That’s right, he was scared. Me, in Trumai. “Does she really want to have sex with me?” I was thinking. Pedro, in Trumai, in a serious tone. It’s true, she wanted to hunt him. [He uses the verb lax, which describes the act of lying wait, ready to pounce.] [Big success. Someone says “atsi!” or ‘yuck!’] Kumaru, in Trumai. Each time, he leapt to his feet. Muchatari, in Trumai. You should have grabbed him, you’re crazy. Kumaru. He would have had a fit… Muchatari. You could have kept on scaring him. Kumaru. I used to creep up on him and say: “Walatu!” [Impersonating my reaction:]“What? What?” He jumped up with a start. Muchatari. You should have grabbed him! Kumaru. He leapt to his feet saying, “You woke up?” He was lying [i.e., my polite question was a way to act normal], since he was scared. [Everyone laughs.] [After a pause:] Only when I die, I will forget. As this last comment testifies, these feigned acts of sexual aggression generate a feeling of both joy and nostalgia. This combination can be compared to the happiness provoked by collective rituals, which are all designated in Portuguese as festas, “parties.” There too, social bonds are embodied as intense feelings of joy and happiness, which are later carefully recalled. There is, however, more to it than that. The very family that so often praised my “knowing how to joke” also complained about it to other villages. I “joked too much” and therefore showed a lack of respect for the Trumai. This denigration, at the time I was staying in a village other than theirs, was part of the complex political game between the different Trumai factions, where the anthropologist sometimes appears as a stake. Nevertheless, the fact that joking could be mobilized as an argument in such a context pleads for its inescapable ambivalence. Another indication lies in the drastic differences between people toward joking. Some people are known as masters, while others claim not to joke at all. For instance Wayaku, a grandmother in her sixties, asserted: Wayaku. I don’t joke. Me. You don’t like to? Wayaku. Some Indians tease their cousins, and then they insult each other, [in Portuguese] they insult [each other]. Me. Really? Wayaku. That’s why I don’t like it…. I am no joker, Walatu, [with] my cousins, I don’t joke. Even if I have a lot of cousins, a lot of female Kamayura cousins, I don’t joke. I only discuss properly with them, only properly I discuss, only properly we joke about this exclusively. [hesitation] Me. Why? Wayaku. I really don’t like to. I only talk properly to people. I don’t joke. I don’t say foolish things. Me. Oh really? Wayaku. I only talk like this, I say proper things. When others joke, when they joke with other people, then I laugh. Then it is sweet for me. When they talk to someone else, to another person, when I hear it I laugh. Then it is sweet for me. But I don’t joke, I just listen. Wayaku. When they talk to other people then I laugh. But as for me I don’t joke. Interestingly, Wayaku ascribes her not liking to joke to moral concerns, rather than to a lack of humor, since she enjoys a good laugh. Her judgment opposes hopeta and taraiw to “proper” or “good” talking (det’a ami), thus implying that the former is ae tak, “bad.” Officially, the chief (called aek in Trumai, “he who is good”) is not supposed to joke in the Upper Xingu, as he represents local moral standards of conflict avoidance and generosity.7 In this short interview one easily sees the motives behind this condemnation of joking: its possible conversion into actual hostility, manifested by insults (wita wita), and thus its lack of fit with the Xinguanian ethos of control, peace, and shame. The word tsiro tsiro, here translated as “foolish things,” with its remarkable semantic reach, confirms the continuity between joking, ribaldry, and aggressiveness. This word commonly appears in judgments on coarse or gross conversations. For instance, a night when a Trumai man was teaching a Brazilian visitor, with vivid impersonations, the extraordinary variety of women’s reactions to sexual intercourse (according to tribe, race, and age), his wife marked her disapproval by saying: tsifan tsiro tsiro len kain hi ami katsi thing tsiro tsiro only emphatic 2Ps talk sitting You only say vulgarities Used with the verb tsin (to do, to cause), in the expression tsin tsiro, it can mean to “bother” someone on purpose. For example, the teasing intention is phrased this way, when two friends decide together to play a trick on a common cousin. In Portuguese, its conventional translation is “encher o saco,” literally to “fill someone’s balls,” (i.e., to piss him off). This is seen as permitted and friendly, but the exposure to public attention, the necessity to endure or retaliate in a witty manner is also stressful. Some people avoid particular villages because they “have too many cousins there,” or know that one particular family loves joking too much to guarantee a safe and relaxed visit. The same expression, tsin tsiro, also means to “persecute.” In historical narratives, it is employed to describe the Kamayura’s violent and systematic attempts to exterminate the Trumai. Thus the same term covers every situation from trivialities to genocide by way of mockery, revealing that they are indeed part of the same spectrum. This woman’s reluctance to joke also has another motive, linked to questions of reputation. She does not, indeed, joke easily. But this is not true of her daughters, who are masters of irony. But she underestimates this, or reluctantly admits that only one of them “jokes a little bit.” The reason is that joking is not only risky for its target; it can also disqualify its agent. Playing with boundaries is necessarily perilous. One adjective very commonly punctuates joking exchanges: ayoar. It literally designates the person in convulsions or epileptic fits, diseases that particularly threaten children. Most of the time, they are imputed to transgressions of couvade prohibitions identified by the shaman. Many animals are supposed to be ayoar and are therefore unfit for children and parents before the former reach the age of five or six. In each time, the Trumai can mimic the particular behavior that justifies the ayoar attribute: uncontrolled head movements of the capucin monkey, the toucan’s rolled-back eyes, the excessive drooling of the razor-billed Curassow (Mitu tuberosum), etcetera. One fish (unidentified) has no such strange or improper behavior but is equally ayoar, and therefore prohibited, because it is said to transform into a snake when it gets old. In its wider meaning, ayoar could be translated as “fool” (bobo in Portuguese) or “crazy,” with the same spectrum of uses the latter term has in English, (i.e., qualifying serious mental disorders as well as amusing tomfooleries). Thus joking is almost necessarily ayoar in a mild sense. But the boundary between weak and strong uses is fairly porous. For example, it is quite normal to joke about one’s mistresses in a ribald manner, but doing the same about one’s wife is ayoar in the strong sense, and will provoke disgust (yikpix) and embarrassment, as will violations of kinship terminology and proscribed behavior. In short, ayoar qualifies any behavior that shows a lack of control, either physical (spasms or convulsions) or social (lack of etiquette), any behavior that transgresses ontological or social boundaries. One then has to make sure not to go too far. But it is also important to not joke too often, on too many occasions, or with too many kinds of kin. Joking might then become a constant property of the individual and not a performance limited to particular contexts. This does not mean that everyone seen as a “joker” taraiwk8 among the Trumai is condemned or despised. But some of them are, at least according to some people. Someone too prone to transgression casts doubt on his ability to identify transgression as such. Others might suspect that he no longer has the necessary ironic distance to his joking: instead of acting like a fool he is seen as one. At this point we are able to identify the boundaries of joking: going too far can degenerate into insults and actual hostility. Joking too much can ruin a reputation. And acting too safely prevents one from being funny. Values of laughter However, this way of presenting things suggests that failure and success in joking is only a matter of controlling these dangers, as if everyone shared a common desire for social harmony, as if joking was only about laughing, was nothing but harmless fun that adds spice to an otherwise mundane existence. In this final section I would like to briefly introduce another level of uncertainty, which gives some idea of the contextual and teleological range of joking (i.e., when it can be used and what it can be used to do besides provoke laughter). In the first place, when laughing at someone, one can choose not to offer the victim the opportunity to show his capacity for self-mockery. Joking then indeed becomes exclusionary rather than conjunctive, as I witnessed on several occasions. One was particularly revealing. A woman from a different Xinguanian group invited herself to a Trumai village with her six-year-old son. She was there to ask one of the Trumai men to marry her and keep her in the village. She claimed that her son was his. For a couple of weeks, the whole family united in teasing him, urging him in a mocking way to accept her, and seriously evaluating whether the child looked like him and whether he should accept the offer. The situation was one of real uncertainty, at least for the woman and the man’s family. After this phase, however, and bearing in mind the man’s firm refusal and a few slightly inappropriate acts by the woman, they reached the collective conclusion that the boy definitely did not look like him. I then witnessed a mocking session in Trumai, in her presence, that made everyone laugh but her. She could not understand but clearly perceived that she was the subject of the conversation. She left a few days later. In a similar way, joking can be a way of accusing or deliberately shaming someone to make him actually lose face. One day, upon seeing an alleged sorcerer arriving in a FUNAI post, a Wauja man said to him in front of everyone: “You are here! So you flew? I’d love to do that. Let’s fly together, ok? Let’s fly!” This was a transparent allusion to “animal’s skin,” a sorcerers’ cloth that allows them to fly at night to faraway places. Witchcraft is normally a serious matter, since a reputation as a sorcerer can lead to death. Its target could hardly find it funny. “He just laughed” is the conventional description of someone who lost the interaction, someone who tried not to lose face by this defense mechanism but actually did. Between these two poles (joking as integrative on the one hand and joking as cruel mockery on the other) there are a few intermediate forms. For example, even in a “great amipine” relationship, there may be really hidden intentions, that will remain such, as shown in this final account from Amati: Amati. My cousin T. also said to me: “Let’s go bathe in the river.” “Ok.” It was early in the morning, around six o’clock. We lit a fire on the riverbank. Then came the women. He said to me, “Cousin you are going to help me.” “With what?” “Did you see that man over there, with his wife? Provoke him, and while you provoke him, I’ll have sex with his wife.” I said, “Ok.” “You’ll do that for me?” “I will.” And it worked perfectly, Walatu! So a few moments later the guy came along playing the flute. [T. said:] “Go! Go! Go! That looks like him, doesn’t it?” “I don’t know. I will imitate him [this is considered to be a lack of respect], and if it turns out not to be him, you’ll get me out of trouble by saying I thought it was our amipine.” “Yes, Ok,” he said. “Well then.” So I started imitating [the flute] “wa wa wa wa” [The amipine:] “Who is disrespecting me?” Then I started to shout and shout again, imitating him. Then I said: “Play more! Play more!” “Who isn’t paying respect? I am no child!” “Yes indeed, you are old, you are mira [grandfather in Kamayura]!” “If this is who I think it is, I will drown you in the river!” “Ok, I am waiting for you!” Then I jumped in the water and waited, among the other people who were taking their bath. He soon arrived. I stayed still. “Who was imitating me?” he asked. T. said: “It’s your cousin, our cousin over there.” He turned to me: “It was you?” “Oh! What a miracle, you go to take your bath early!” [This is a comment implying that he is both lazy and old as only elders are authorized to take their time in the morning.] “Jump in quick, you look cold.” “No, I am fine, but I’ll go right now indeed.” He jumped. Tsum! Every time he came close, I threw mud on him. “Wait a second! I’ll drown you.” He chased me all around, but he couldn’t grab me, until in the end I climbed out of the river. “You are no kind of a man!” “No, you are no kind of a man!” Then we ran, we ran. After a while I felt tired, I hid in the long-grass [capim]. [The chase goes on, in the banana plantation, and then Amati goes back to the village without his cousin noticing] I went straight to his house, on purpose. I stayed there, and his mother asked me: “Where is your amipine?” “I left him by the river, he was trying to beat me up.” “He doesn’t know that you came here?” “No, I don’t think he does.” Then it took a while before he arrived. His family said to him: “You only come now? Your amipine is already there, in his house.” [The family is taking the side of Amati for the sake of the joke.] Others said: “He wanted to beat him, to drown him, to beat him for real.” His mother said: “Joke properly, my son!” [The amipine:] “Ah, I’m tired, I kept chasing him like an idiot.” Then I answered: “Because you are one indeed!—Oh! He is here! Wait!” Then he came close, but when he was about to grab me I escaped and ran out. Everybody laughed. Me. And meanwhile T. was having sex with his wife? Amati. He had already finished! He had already eaten her. [laughs] A trick might hide another trick. The whole situation, understood by the victim as normal mocking and horseplay between two great cousins, is in fact a diversion. This difference in knowledge between the participants clearly means additional fun for the teasers and for those who hear the story. The intention is not actually hostile but will not be revealed to the victim, who will thus be excluded from the collective laughter. This does not imply a breach in the amipine relationship, at least from Amati’s perspective. Sometimes intentions are actually hostile. But once some time has passed, it is possible to pretend they were feigned. No one is taken in by it, but this allows everybody to keep up appearances. In 2006, I received an aggressive email. Its author accused me of stealing Trumai culture and, in the end, threatened to “mark [my] ugly ass with arrows.” I was not particularly at ease when I turned up at his village two years later. I was talking with his father when he entered suddenly, with a bow and tipless arrows, and said: “Didn’t I say I would mark your ass with arrows? Turn around.” The three of us forced a laugh and then he turned around and left. The structure is the same as a joking event: demonstration of hostility, a pause, and then requalification of it as a joke. The only difference is the remarkable delay (two years) between tension and release. Conclusion This description of one aspect of Xinguanian sociality allows us to contrast two very different interaction frames, both equally explicit, and both paradoxical in that they contradict the very idea of a frame of interaction. The first one, the affinity-shame frame, is extremely stable. It is, when all is said and done, nothing more than a set of rules for avoiding interaction and therefore for shielding its protagonists from shame. In this way, the affinal relationship can continue safely. The second one, on the contrary, sets up rules such that instability is inescapable, and allows for shameless attempts to shame the other. It consists in explicitly giving no limits, leaving to the actors the task of establishing them in the course of interaction, and of gradually building a relationship. The instability lies in the fact that the more stable the joking relationship between two people becomes, the further they must go to make each other unsure of the other’s intentions, which is the driving force of humor. This formulation is exaggerated for the sake of the argument. Not every amipine relationship ends in anger and rancor. But there is indeed a strong tendency to escalate, and as such sad endings exist as a possibility according to the rules of the game. In such a context of constitutive instability, joking can assume different functions and express a large spectrum of intentions. It is therefore difficult to decide what it is to fail or to succeed in such interactions, difficult to define dysphoria and euphoria when the very idea of a shared construction of common ground is explicitly discarded. It all depends on the teaser’s intentions, which are never, by definition, taken for granted. One might rightly argue that this aspect is not specific to the Trumai and their neighbors of the Xinguanian reservation, and that many examples of teasing given in the course of this paper could be found elsewhere under very similar forms. True indeed. And this is precisely why it is useful to model how uncertainty is built as constitutive of the joking frame, and to show how this can shed light on the culturally specific relationships joking is locally associated to (potential and virtual affinity in our case). It is now possible to formulate a few hypotheses about my introductory anecdote, and more precisely about why they insisted on me going on, and why everyone laughed when I said I was ashamed. I was assuming that they expected to finally find one joke funny. But they were more eager to maintain and feed my growing embarrassment, derived from this systematic and repetitive failure—which was funny per se. It was but one of many episodes in the joking relationship that was gradually constructed between the chief and me. Declaring my shame was, unbeknownst to me, the best signal for an end. Being metapragmatic, and proving my uncertainty about the situation, it showed that the teasing had already been a success. As a moral value, it was also the exact opposite of joking, and thus, perhaps, a reminder that I was not being treated with respect. But how could I possibly be sure? Acknowledgements This paper has had many readers to whom I am grateful for their insightful comments. I especially thank Matthew Carey, François Berthomé, Julien Bonhomme, Olivier Allard, Grégory Delaplace, and three anonymous reviewers. This paper was written during a one-year postdoctoral fellowship in the department of anthropology at the London School of Economics, and greatly benefited from the comments made by LSE staff and students after its presentation at the Research Seminar on Anthropological Theory. References Afford, Finnegan, and Richard Alford. 1981. “A holo-cultural study of humor.” Ethos 9 (2): 149–64. Basso, Ellen. 1973. The Kalapalo of Central Brazil. New York: Holt, Rinehart, and Winston. Bateson, Gregory. 1972. “A theory of play and fantasy.” In Steps to an ecology of mind, 177–93. San Francisco: Chandler. Beeman, William O. 2000. “Humor.” In Linguistic lexicon for the millennium, edited by Alessandro Duranti, special issue of Journal of Linguistic Anthropology 9 (1–2): 103–6. Brant, Charles S. 1948. “On joking relationships.” American Anthropologist, New Series 50 (1): 160–62. Canut, Cécile, and Etienne Smith. 2006. “Pactes, alliances et plaisanteries: Pratiques locales, discours global.” Cahiers d’études africaines 184: 687–754. Carty, John, and Yasmine Musharbash. 2008. “You’ve got to be joking: Asserting the analytical value of humour and laughter in contemporary anthropology.” Anthropological Forum 18 (3): 209–17. Clastres, Pierre. 1974. La société contre l’Etat. Paris: Editions de minuit. Douglas, Mary. 1975. “Jokes.” In Implicit meanings, 90–114. London: Routledge and Kegan Paul. Fausto, Carlos. 2012. “The friend, the enemy and the anthropologist: Hostility and hospitality among the Parakanä (Amazonia, Brazil).” Journal of the Royal Anthropological Institute 18 (S1): S196–S209. Franchetto, Bruna. 2001. “Linguas e história no Alto Xingu.” In Os povos do Alto Xingu, edited by Bruna Franchetto and Michael Heckenberger, 111–56. Rio de Janeiro: UFRJ. Garde, Murray. 2008. “The pragmatics of rude jokes with Grandad: Joking relationships in Aboriginal Australia.” Anthropological Forum 18 (3): 235–53. Goffman, Erving. 1974. Frame analysis. New York: Harper and Row. Handelman, Don, and Bruce Kapferer. 1972. “Forms of joking activity: A comparative approach.” American Anthropologist, New Series 74 (3): 484–517. Heckenberger, Michael. 2001. “Estrutura, história e transformaçao: A cultura xinguana na longue durée, 1000–2000 D.C.” In Os povos do Alto Xingu, edited by Bruna Franchetto and Michael Heckenberger, 21–62. Rio de Janeiro: UFRJ. Lagrou, Elsje. 2006. “Laughing at power and the power of laughing in Cashinahua narrative and performance.” Tipiti: Journal of the Society for the Anthropology of Lowland South America 4 (1): 33–56. Lévi-Strauss, Claude. 1964. Mythologiques, T. 1: Le cru et le cuit. Paris: Plon. Morton, John. 2008. “Poofters taking the piss out of Anzacs: The (un)Australian wit of Sydney’s gay and lesbian Mardi Gras.” Anthropological Forum 18 (3): 219–34. McGregor, William. 1989. “A structural analysis of a humorous story in Gooniyandi.” Journal of Literary Semantics 18 (2): 85–116. Overing, Joanna. 2000. “The efficacy of laughter: The ludic side of magic within Amazonian sociality.” In The anthropology of love and anger: The aesthetics of conviviality in Native Amazonia, edited by Joanna Overing and Alan Passes, 64–81. New York: Routledge. Radcliffe-Brown, Alfred Reginald. 1940. “On joking relationships.” Africa 13 (3): 195–210. Rosengren, Dan. 2010. “Seriously laughing: On paradoxes of absurdity among Matsigenka people.” Ethnos 75 (1): 102–21. Von den Steinen, Karl. 1886. Durch Central-Brasilien: Expedition Zur Erforschung Des Schingû Im Jahre 1884. Leipzig: F. A. Brockhaus. ———. 1894. Unter Den Naturvölkern Zentral-Brasiliens: Reiseschilderung Und Ergebnisse Der Zweiten Schingü-Expedition, 1887–1888. Berlin: D. Reimer (Hoefer and Vohsen). de Vienne, Emmanuel. 2011. Traditions en souffrance: Maladie, chamanisme et rituel chez les Trumai du Haut Xingu (Mato Grosso, Brésil). PhD dissertation. Paris: École des Hautes Études en Sciences Sociales. Viveiros de Castro, Eduardo, and Carlos Fausto. 1993. “La Puissance et l’acte: La parenté dans les basses terres d’Amérique du Sud.” L’Homme 33 (126–128): 141–70. Viveiros de Castro, Eduardo. 2001. “GUT feeling about Amazonia: Potential affinity and the construction of sociality.” In Beyond the visible and the material, edited by Laura Rival and Neil Whitehead, 19–43. Oxford: Oxford University Press. ———. 2002. “Capitulo 8. Atualizaçâo e contra-efetuaçâo do virtual: O processo do parentesco.” In A inconstância da alma selvagem, 401–57. Sâo Paulo: Cosac and Naify. « Mettez-vous mal à l’aise » : Relations à plaisanterie et incertitude prévisible chez les Trumai du Brésil Central Résumé : Cet article analyse les formes que prend la plaisanterie chez les Indiens Trumai et d’autres groupes du Parc Indigène du Xingu (Mato Grosso, Brésil). Cette pratique sociale semble accompagner les relations ouvertes, et peut être ainsi définie comme un mode relationnel par défaut : quoiqu’elle soit associée d’une manière prototypique à la relation entre cousins-croisés masculins, elle marque fréquemment aussi les relations avec les étrangers, Indiens ou Blancs. A l’extrême opposé, les relations entre affins réels sont définies par le respect et la honte. L’identification de ce système d’attitude ne saurait cependant rendre compte ni des propriétés pragmatiques de la plaisanterie, ni de son efficacité sociale spécifique. On insiste en particulier sur l’ambivalence—aussi bien morale que fonctionnelle—de la plaisanterie, laquelle dérive directement du cadre d’interaction qui lui est propre. Ce dernier est en effet capable d’instaurer des échanges et des formules très standardisés tout en générant un fort sentiment d’embarras et de déstabilisation. On s’efforce donc d’expliquer le paradoxe de ce que l’on pourrait appeler une incertitude prévisible et d’évaluer ses conséquences sur les notions de succès et d’échec dans l’interaction. Emmanuel de Vienne is Assistant Professor (maître de conférences) at the Université Paris-Ouest Nanterre La Défense. He conducted his fieldwork among the Trumai of the Upper Xingu (Mato Grosso, Brazil). His PhD dissertation focused on the experience of sickness and on the transmission of therapeutic shamanic traditions and ontologies. Continuing this interest in cultural transmission, he published on related issues such as the price of traditional knowledge (with O. Allard) and transmission and evolution of ritual chants. He currently works on the ongoing process of ritual revitalization among the Trumai. ___________________ 1. Following Eduardo Viveiros de Castro and Carlos Fausto (1993), I call virtual affines people that can actually become affines, reserving “potential affines” for more distant Others. 2. Inetl ka’am is a contraction of ine-tl kain ha ami (“to him indeed I speak”). Inatletl ka’am is a contraction of inatl-etlkain ha ami (“to her indeed I speak”). 3. Ha a (both of us) is exclusive. The inclusive form would be ka a. 4. He uses the Portuguese indeterminate term fulano, “Mr. So-and-So,” because it is prohibited to pronounce the name of a recently deceased person. 5. See, for example, Viveiros de Castro and Fausto (1993), as well as Viveiros de Castro (2001, 2002). 6. Here the uncertainty is precisely what makes the joking socially productive, as is the case in the love affair described by Matthew Carey (this volume). 7. Among the Trumai, however, contemporary chiefs do joke, just as the former great chief who died in 1995 used to do. I don’t know whether this is specific to the Trumai or is just another example of the discrepancy between norm and actual behavior. 8. Taraiw-k: joke-nominalizer, literally “the one who jokes.”

Corsín Jiménez: Daribi kinship at perpendicular angles This work is licensed under the Creative Commons | © Alberto Corsín Jiménez. Attribution-NonCommercial-NoDerivs 3.0 Unported. ISSN 2049-1115 (Online) Daribi kinship at perpendicular angles: A trompe l’oeil anthropology Alberto Corsín Jiménez, Spanish National Research Council (CSIC) What are the contours of such “knowledge” that does double duty both as a public good or commons and as a source of individual empowerment and liberty? This article offers an analysis of the epistemological organisation of the “knowledge economy” by shooting ethnographic work on Free Software through a playful trompe l’oeil of Roy Wagner’s classic piece on Daribi kinship. It offers a preliminary template for thinking of Euro-American knowledge as itself a trompe l’oeil device. Keywords: kinship, knowledge, commons, free software, perspectivism, perpendicularity, epistemology [T]he opposite of “kinship” is . . . total, unrestricted analogy. Wagner, Analogic kinship: a Daribi example (1977) Let us begin with the proposition that all forms of knowledge are analogous to one another. There are of course different kinds of knowledge. But we may think of these kinds as different expressions or manifestations of an underlying relational economy. In this economy, knowledge flows across and along different kinds of relationships—but it is the fact of its “analogic flow” that takes precedence and defines the underlying cultural logic. For in this system there seems to be a prerequisite and expectation that knowledge must be made to flow. Said differently, a “certain solicitude . . . is quintessential to all ideal [knowledge] relationships, regardless of how they may be defined or in what forms the solicitude is expressed. And this solicitude represents, as well as anything can represent, what I mean by the basic analogy of all [knowledge] relationships to one another.” In this analogical economy, then, there is little profit in thinking of knowledge as being divided up into discrete units or entities. However and whatever knowledge is differentiated into, it will always be so from the starting premise of its analogical flow. Thus, “the flow of relationship, and ultimately linearity—analogy across the [epistemic forms]—is integrally linked to differentiation. Lineality is not a separate, ‘political’ consideration . . . but is always an aspect of a totality that includes differentiation as well.” Such a totality is the underlying analogic flow of knowledge: the essential solicitude through which epistemic forms relate to one another. Readers may recognize in the above a paraphrasing of the opening sentences in Roy Wagner’s classic article, “Analogic kinship: a Daribi example” (1977: 623-24). Where Wagner spoke of “kinship,” however, I have substituted the term “knowledge”; where he spoke of “generations,” I have inserted “epistemic forms.”1 That the organisation of knowledge may be modelled on an analogical economy of relations is, however, hardly a novel claim. The digital commons of the Internet’s knowledge economy is the example par excellence. The commons is here actualized through a decentralized network of “knowledge-based communities” (David and Foray 2003: 21). The network enables “online exchanges” between diverse users, including designers, engineers or lay people on an equal footing, thus making it impossible to identify the exact location and identity of an innovation. The larger the number of people engaging in such peer-to-peer exchanges, the greater the chances of making the overall stock of knowledge grow. This is therefore an incremental model of knowledge-exchanges that relies on the very “wealth of networks” (Benkler 2006) as at once infrastructures for and analogues of knowledge. It is on this account that the networked economy of information may be seen as an analogical economy of knowledge, where activating a connection with a peer instantiates at once a relation and a flow (of bytes, files, information). The analogical qualities of such a networked economy of knowledge have been amply documented and commented upon. It is often noted, for example, that the structure of networked information—and in particular the negligible costs of reproducing digital information—instantiates a de facto regime of superabundant or “open knowledge” (David and Foray 2003; Foray 2006: 172-79), where subsequent attempts at appropriating or circumscribing the flow of information would in effect amount to a “tragedy of the public knowledge ‘commons’” (David 2000). The commons of the knowledge economy may also be thought of as a system of externalisations. The image is derived from economists’ own definition of public knowledge as, indeed, an externality: knowledge that, due to its intrinsic qualities, escapes proprietary circumscriptions, that “spills-over” market boundaries and enclosures (Cornes and Sandler 1996; Frischmann and Lemley 2006). Of course, what makes knowledge capable of “externalisation” is also what makes it “public.” In its travelling as an externality, knowledge effectively detaches itself from itself: its very flow defines (and contributes to enhancing) the largesse of knowledge from wherein it has departed. In their analogically detached form, then, the flow of all such externalities through exchange is constitutive of the whole relational economy. The opposite of property Although there is considerable agreement today over the epistemic richness of knowledge in the wealth of networks, it seems a little harder to agree on its allegedly expansive qualities, whence the proliferation of liberal designations: open, commons, public. “Why do [some people] say the public domain is a commons?” asks critical legal scholar James Boyle. “Why [do others] use the term commons to refer to any resource protected by a liability rule rather than a property rule? Why [do they] refer to open source software as a commons, though it is in fact protected by a property rule?” The answer, he adds, “lies in the particular functions and dysfunctions of property on which the theorist is implicitly concentrating while creating their image of property’s outside, property’s antonym” (Boyle 2003: 29). The liberal qualities of knowledge are therefore cajoled into existence, Boyle tells us, by “‘the opposite of property’—its outside, its limitations, negations, inversions and correctives” (31-32). The practice of opposition would therefore seem an expansive terrain in itself, of which perhaps the most famous of its currents today might be the culture of hacking. Accounting for such a diversity of hacking practices, Gabriella Coleman and Alex Golub (2008) have recently described three traditions of hacker ethics: three modalities of exercising “freedom” or “liberalism” in cyberspace, setting up in each case their own complex negotiations with the world of property. The tradition of crypto-freedom, for instance, promotes “an ethical sensibility that affirms the sacrosanct nature of individual privacy” (ibid.: 260). Thus, the concern of the volunteer association Cypherpunks is to protect individual autonomy from intrusive corporate behaviour, including governments, and in pursuance of this aim they develop technological solutions for securing personal privacy in the world of digital information. Whilst Cypherpunks are thus concerned with guarding freedom from institutional intrusion and coercion (known in political theory as “negative freedom”), Free Software advocates promote instead a “liberal version of freedom that invoke[s] the virtues of sharing and pedagogy”—thus a “more positive notion of liberty” (261). Moreover, the essential freedom that inheres in information imbues the practice of programming among Free Software developers with an experience of artistic and creative self-expression: a “romantic drive” for creativity whose goal is to enrich the awareness and bonds of communitas among fellow developers (263). Last, the ethics of the so-called “hacker underground” is characterized by a radical culture of transgression. Their moral practices engage in an ongoing struggle over the means and technologies that work to consolidate power. As Coleman and Golub put it, the “underground seeks to remind those in power that there are individuals in an unknown, cavernous ‘out-there’ who can and always will unsettle, even if only temporarily, the purported absolute power of ‘the establishment’.” The underground cherishes the thrill that derives from “the aesthetics of eating forbidden fruit into the human art of the short con” (264), an excitement that is expressed in idioms of pleasure, passion for exploration, even humour. The three moral genres of hacker practices therefore enact very different “opposites to property.” Coleman and Golub suggest that in the American context the hacker material may in fact be read as “a cultural case in which long-standing liberal ideals are reworked in the context of interaction with technical systems to create a diverse but related set of expressions concerning selfhood, property, privacy, labor, and creativity” (267). As an ethnographic insight, the opposite of property would in this context seem to emerge from/as an American tradition of romantic individualism, liberty and freedom. Although it is certainly not their interest to reify a notion of American Culture, overall Coleman and Golub’s study suggests that the techno-cultural form of hacking stands in a “dialectical relationship” to the general cultural template of liberalism (267). James Leach has also carried out work among Free and Open Source Software (F/LOSS) programmers, and his ethnography focuses likewise on the morality and aesthetics of freedom and liberty at play among developers (Leach et al. 2009). His interests, however, lie in the processes of emergence of “freedom” as a social form and action. Not, then, on the kinds of techno-cultural forms that stage a dialogue with, and contribute to, ongoing conversations about the qualities of liberalism, but on the internal enablement of hacking itself as an emancipatory form. A number of elements play a part in this entangled process of enablement.2 For one, hackers make serious investments in writing good code. Good code is code that works, which brings about the effects wished-for in the “external world.” Leach says that hackers have a “‘binary’ approach to judgement,” by which he means that in writing good code hackers are ultimately concerned with writing programming language that is machine-truthful. The goodness of code evinces in it being machine-readable. If the programming language is true, then the code will be good: “judgements of fitness or effectiveness merge with more subjective aesthetic judgements.” This binary form of judgment, Leach suggests, may be read as a “metonymical extension” of the binary code (zeros and ones; on and off) upon which all computing programmes run (58-59). If the natural morality of good code appears to uphold itself on the natural veracity of language statements, there are however certain forms of cultural exchange that help elicit such truthful code-facts. Leach points for instance to the practice of “flaming,” which involves the heated and rancorous exchange of messages between developers in public forums (60). Such tournaments of agonistic exchange are thought of by the F/LOSS community as occasions that contribute towards the making of more robust code. Code thus achieves the status of a self-evident “Good Thing” (61) when it can survive comparison and the agony of vociferous contrast. As a form of knowledge then, F/LOSS stands at the very frontier of our epistemic futures. The robustness, effectiveness and intrinsic goodness of F/LOSS translates into a sociologically self-evident potentiality: whose beauty, simplicity and effectiveness awaits all sorts of applications and eventful futures. As Leach puts it, “because code both enacts knowledge and makes knowledge evident, in an illocutionary way it reproduces the imagined space of a frontier by revealing the truth of what exists and thus new potentialities” (62). It is of course this sense of immanent futurity that informs the liberating and liberal qualities of F/LOSS as an epistemic object; that enables from within its projection as a political heir to and vehicular form of liberalism. Working upon code is analogous to working upon society. Every bit / byte that works fuels emancipatory social change. Analogy I opened the article with a paraphrase of Roy Wagner’s classic essay on analogic kinship. I would like to return to that piece again. At the end of my introductory section, I noted how in the imagination of the new digital commons, “the flow of all such externalities through exchange is constitutive of the whole relational economy.” This sentence, too, was a paraphrase from Wagner’s piece, who at one point observes that “the ‘flow’ of controlled analogy through exchange is thus constitutive of the whole relational matrix” (Wagner 1977: 631). What Wagner calls “controlled analogy” are “detached, partible things” (souls, women, pearlshells, meat) that circulate among Daribi kinsmen in their efforts at creating distinctions between kinds of persons and relationships. The flow of such analogical objects is controlled because what is at stake here is the lineality of kinship. Let me explain. Wagner makes a distinction between “vertical” or “lineal” and “horizontal” relationships. Vertical relationships, and vertical analogues, are those that contribute towards the life-sustenance of female and male lines. However, whilst femaleness is considered to be self-sufficient, the role of maternal blood (pagekamine) enough for conception, maleness is deemed on the contrary contingent. The seminal fluid (kawa) of a man is never sufficient for conception and must be augmented by the juices and fat of meat. As Wagner puts it, meat is “an adjunct to maleness and male reproductive potential: it is the partible and portable accessory to masculine continuity.” The conceptual world of Daribi kinship is therefore organized through the flow of substantial and partible analogues, although the moral force of the flow is in effect always represented and perceived as the “lineal flow of male substance” (628). Thus, for example, the parties to a betrothal are not simply engaged in a ceremonial exchange of prestations, but are striving to secure the continuity of their respective linealities. The objects exchanged are “adjuncts of female productivity (bark, cloth, string bags, and so forth)” in one case, and flows of maleness itself in the other, “giving meat and other adjuncts of maleness and male productivity (pearlshells, axes, bushknives)” (ibid.). The horizontal flow of such detached markers at once supplements and zooms in on the vertical flow of relational analogies (kawa and pagekamine). objects flow horizontally and in so doing promote or enhance the vertical flow of substances. They analogize each other at perpendicular angles. In Wagner’s words: “Exchange and descent, affinity and consanguinity, become metaphors for one another” (632). Analogies thus work within and across social forms. Vertical flows are analogous to horizontal exchanges; but verticality and horizontality are themselves internally analogized by mutual cross-references. Thus, pearlshells or meat flow horizontally to replenish and augment their verticality. The paradigmatic case of such an analogical duplex ordo is the birth of a child, who “becomes itself a point of analogic relation between the two linealities: relationship has ‘happened’ to the original demarcation between them. Because the same social persona stands in an analogic relation to both, the two are related analogically to one another” (634). A matter to be resolved then is the lineality (maternal or paternal) to which the child belongs. Although both sides see the child as a potential analogue to supplement male contingency, it “becomes a truly pressing issue for the father’s lineality, since for them male contingency is opposed to female sufficiency” (634). It is the paternal side, then, who claims the child through the presentation of a detached analogue for substantial flow (pagehabo). But the flow of analogies accomplishes a further analogy, for “inasmuch as the persona for whom the payments are given belongs simultaneously to both linealities, and relates them, the mediating payments that define this lineal affiliation are exchanges within a single lineality. They are ‘shared’ as well as exchanged” (635). Sharing and exchanging emerge in this context, therefore, as analogies of an analogy. From the point of view of the father’s line, the presentation of pagehabo expresses sharing (the amplification and distribution of kawa, now passed down the male line) through the idiom of exchange (mediating payments). From the point of view of the maternal line, it expresses exchange (mediating payments) through the idiom of sharing (the self-sufficiency of the child’s femaleness, which now supplements male contingency). Importantly, however, the circulation of pagehabo is seen simultaneously as a vertical and horizontal flow by both kinship lines. What the paternal line sees as a vertical flow (the flow of kawa), the maternal line will see as a horizontal flow, and vice versa. As noted above, we may therefore say that the flows are analogized at perpendicular angles with respect to each other. Alternatively, the flow of analogies effects an exchange of axes: as they move along the axes the analogues (meat, pearlshells, etc.) are at once “separating/holding parts of itself/things from one another/together (the sentence can be read in two ways, as indicated by the italics threading one of them)” (Strathern 2005: 163n3). At any moment during a transaction, any one analogue is keeping the axes differentiated (the vertical from the horizontal; the paternal from the maternal) and collapsing them onto a single lineality. “It is,” in Wagner’s words, “like virtually everything else in the relationship, ambiguous, and it acquires this character precisely because the relationship, qua relationship, is constituted by the summing together and mutual modelling of the two aspects” (Wagner 1977: 636). Trompe l’oeil Around 1670, Dutch artist Cornelius Gijsbrecht produced a number of paintings that played on the rhetorical power of perspectivalism to startling effect. Trompe l’oeil letter rack, for example, masterfully displays a three-dimensional letter rack that bodies forward in space. As in all trompe l’oeil paintings, we are tricked into seeing a number of objects that apparently hang out from the picture plane and invade our own physical space. The trick is here twice enfolded, because some of the letters and papers hanging from the rack are partially hidden or veiled by a curtain. The curtain supplements the perspectival trick: it generates a false sense of depth in a painting whose entire mission is to empty itself of depth. The illusion of the trompe l’oeil is sustained by making us believe that the objects are outside rather than inside the pictorial plane. Hanneke Grootenboer has commented on this painting by Gijsbrecht and other seventeenth century Dutch trompe l’oeils, a propos their capacity to turn perspective inside out (Grootenboer 2005: 54). In trompe l’oeil paintings, she argues, the vanishing point and the point of view are made to coincide. Thus, rather than being invited into the picture, our perspectival gaze is in both cases forced instead to bounce off the picture plane. The “vanishing point to which the viewer’s eye is directed,” writes Grootenboer, “can never be reached—or, for that matter, seen—and collapses with the point of view from which seeing is made possible” (54). We think we are looking at a perspectival painting and yet we are not allowed to rest our perspectival gaze at its natural resting place: the vanishing point. Our perspective is instead violently thrust back to us, or as Grootenboer puts it, is “turned inside out”. Trompe l’oeil paintings are a meeting place for what, following Lacan’s famous description in Seminar XI, Grootenboer calls the reversibility of the gaze. The surface of a trompe l’oeil painting captures the encounter of the vanishing point and the point of view: where the perspectival organization of painting displays simultaneously an inside/outside structure. She offers an eloquent description of how this structure operates with the example of a glove: If we turn the finger of a glove inside out, its structure will remain exactly the same but will now be articulated by its reverse side. previously enveloped by the leather, the lining of the glove will emerge. Exteriority and interiority are reversed without actually changing the structure of the glove. Elaborating on this metaphor, we may imagine the tip of the glove’s finger as a vanishing point that once pulled out reveals the other side of the point, which normally falls beyond the horizon. Moving the finger of the glove back and forth, we see how the vanishing point can merge with the point of view when the structure within which these points are normally separated is mapped out (55). The surface of a trompe l’oeil painting therefore becomes a meeting place for two perspectives: the perspective that is hollowed out in a forward direction from the painting’s interior and the perspective held by the viewer from outside the painting. We can no longer speak of a vanishing point and a point of view, for these are descriptors that properly belong to the tradition of linear perspectivalism: inside a painting and outside a painting are epistemological locations defined within the parameters of a classical Renaissance geometrical grid. One could venture instead the argument that what is in effect making an apparition in the trompe l’oeil painting are two distinct pictorial objects: one canvas, two pictorial forms. If we follow Grootenboer’s metaphor, the inside of the glove’s finger is a different object from its outside. They are objects that share a perspectival structure—they both have a point of view on what is inside / outside in each case—but they are ontologically distinct objects. These are not objects differently located within the same geometrical plane, but objects which summon different topological fields. Another painting by Cornelius Gijsbrecht can help illuminate this point. In The reverse side of a painting (c. 1670) Grijsbrecht presents a painting whose canvas has been turned around. Figure 1. Cornelius Gijsbrecht, The reverse side of a painting (c. 1670), Statens Museum for Kunst, Copenhagen. We see a wooden stretcher, six nails holding the frame in place, and the back of a canvas proper. But there is of course no image. Grootenboer has nicely described the viewer’s startled reaction when confronted with this picture: If we follow our inclination to turn this canvas around in order to see what is represented on its front side, its shock effect would reside less in the deception, and more in the discovery that there is nothing there to see. Nothing, except for the same image, back as front (59). The picture whose negation Grijsbrecht painted is of course a different object from the actual frame and stretcher we would encounter if we were to flip the painting around. They hold the same perspective (they index the same presentation of the world: the reverse side of a painting) but they do so from distinct epistemological and ontological positions: a human- and an object-centred point of view. Thus, whereas The reverse side of a paintingis the view we hold of the picture as viewers, that is, a view that obtains through an epistemological action (the act of eliciting the painting as object), the back side of the painting, on the other hand, elicits not an epistemological point of view, but an ontological position: an object (the wooden stretcher) that no longer requires the epistemological elicitation of a viewer to come into existence. The reverse side of a painting and the reverse side of a painting are therefore different epistemological and ontological points of view, respectively, on the same epistemic object. The reverse side of a painting's masterliness lies in Gijsbrecht’s sophisticated problematization of the representational strategies of the linear perspectivalist tradition. Like Velazquez’s Las meninas or Vermeer’s The art of painting, it is a cunning work of baroque intelligence that subsumes its representational motif in a dazzling display of double relations (inside : outside / object : human / epistemology : ontology), and in so doing endows the notion of the duplex itself with epistemic status. The reverse side of a painting calls out the reverse side of a painting: the painting and the wooden stretcher mutually require each other for completion. At the same time, completion as such never quite obtains, for it always depends on one additional reversion, another flip of the object. The object therefore inhabits and performs an aesthetic of permanent elicitation. Gijsbrecht’s boldness forces us to take residence in the epistemic space of the painting’s liminal screen as a guarantor and purveyor of reasonableness. Oscillating between an interior and an exterior perspective, an ontological and an epistemological point of view, the trompe l’oeil’s very reversible structure emerges therefore as the only possible comfort-zone for stabilizing, if in a transient and fragile manner, the turbulence and confusion of all such double movements. It is the painting’s reversibility that holds all such reversions meaningful, much like perpendicularity does for the relational matrix of Daribi kinship. If reversibility operationalizes the task of conceptualisation for Gijsbrecht’s baroque reason, perpendicularity does so in Wagner’s axial world, where the horizontal elicits the vertical, and vice versa. Reversibility / perpendicularity is a concept that allows us to flip between an epistemological and ontological point of view. At perpendicular angles Gijsbrecht’s trompe l’oeil baroque cunningness reminds us of the Daribi flow of analogies at perpendicular angles. We may remember how in Daribi betrothal prestations the vertical flow and the horizontal flow were each a “model ‘of and model ‘for’” one another (Wagner 1977: 631). In this mutual self-containment, the work of the analogical resembles therefore a sort of trompe l’oeil effect, whereby the horizontal (say, the perspectival vanishing point) collapses onto the vertical (the self-standing point of view). The vanishing point is hollowed out in a forward direction and made to coincide with the point of view. The surface of the trompe l’oeil canvas contains thus two perspectives (the vanishing point and the point of view) within a single lineality (the surface of the trompe l’oeil painting itself), not unlike how a Daribi “controlled analogue” allows for conceptual flips between the horizontal and the vertical, the paternal and the maternal, sharing and exchange, kawa and pagekamine, “within a single lineality” (ibid.: 635). Eduardo Viveiros de Castro has recently produced an original analysis of anthropological theory (an anthropological theory of anthropological theory) that echoes some of the baroque motifs and aesthetics of the trompe l’oeil technique. He has described the transformational conditions of anthropology as the “discursive anamorphosis of the ethnoanthropologies of the peoples thus studied” (Viveiros de Castro 2010: 17, emphasis added). A word on the relation between the anamorphic and trompe l’oeil in this context is therefore in order. An anamorphic projection is one in which the point of view and the vanishing point of / on an object are deliberately disjoined. We look at a painting but our gaze finds no rest in its vanishing lines and objects. It bounces off the canvas and isthrust back upon us. When looking at one such painting one therefore requires the use of a special device (say, an optical lens) or manoeuvre (for instance, approaching the painting at a skewed angle) to restore the alignment of the vanishing point and the point of view, and thus the proportional semblance of an otherwise distorted projection. Thus, in front of an anamorphic painting our subject position is determined by the point of view itself: We may opt to take a position perpendicular to the painting’s surface, in which case the anamorphic projection will appear smeared and distorted. We will be seeing nothing. In effect we would have therefore opted for assuming a subject-less position: the position from wherein nothing becomes visible. Alternatively we may engage with the painting and attempt to assume its own point of view. This might require our bending the lamina or coming up close to the canvas and skewing our vision diagonally across its surface. Slowly, gradually, the anamorphic distortion will recompose. The painting will disclose and elicit its own point of view. We will now be seeing the distorted image but it will have become visible “from the painting’s point of view” (Grootenboer 2005: 131). We may therefore say that anamorphosis and trompe l’oeil techniques both work as analytical devices that elicit “perspectivalism” itself as an analytic.3 A point of view on/of the point of view: the perspective’s perspective.4 Although neither anamorphic paintings nor trompe l’oeils necessarily require the splitting of the vanishing point and the point of view at perpendicular angles, 5 the image of the perpendicular split is useful nonetheless as an analytical heuristic. Not surprisingly, Viveiros de Castro has resorted to this image in his elucidation of how an anthropological anamorphosis functions in his own models of Amerindian perspectivism: “Amerindian perspectivism,” he notes, “is an intellectual structure that contains a theory of its own description by anthropology; because, that is, it is itself an anthropology situated at a perpendicular angle with respect to ours” (2010: 60, my translation). The “perpendicular” is of course also the neatest exposition of how the “relational matrix” of Daribi kinship obtains when the vertical and the horizontal cross axes. In his recent book, Cannibal metaphysics (2010), Viveiros de Castro offers an original reinterpretation of Wagner’s model of Daribi kinship in terms analogous to the perpendicular perspective offered here. He picks up on Wagner’s description of kawa and pagekamine as “simply two ways in which the vertical flow of analogy” may be represented. Mutually analogized at perpendicular axes, we may recall, these “amount to the same thing seen, as it were, from different angles” (Wagner 1977: 628). Viveiros de Castro follows up on this remark by observing that in the “interpretation of Melanesian gift-exchange as being defined through the tension of an exchange of perspectives... it is the notion of perspective that determines conceptually that of exchange, and not the other way around” (2010: 133, my translation). What looks like the flow of kawa from one angle is seen as a flow of pagekamine from a different axis. The exchange of prestations is therefore elicited by this mutual analogising of the same thing from different points of view. It is the angular aperture, so to speak, that makes the elicitation visible. My overall point, then, would be to draw one further analogy: between the flow of analogies, the exchange of perspectives, and perpendicularity itself (the angular aperture) as a trompe l’oeil effect. What Viveiros de Castro calls an “exchange of perspectives” performs the same function to what I referred to earlier as an “exchange of axes.”6 As my detour through Cornelius Gijsbrecht’s work brought into focus, however, the substitution or displacement of “axes” for “perspectives” brings to the fore the purchase of perpendicularity itself as a concept. In Gijsbrecht’s paintings, the canvas’ surface both gathers and keeps distinct, in properly baroque and ambiguous fashion, the horizontal (vanishing point) and the vertical (self-standing point of view) axes. When views meet, the axes collapse, and the perpendicular emerges. The surface of the trompe l’oeil painting becomes therefore a meeting place for two perspectives/axes. Analogously, we might say that in the encounter between an Amerindian vanishing point and a Daribi point of view, the perpendicular is what surfaces as the very concept of anthropological analogy. Enablements In the terms in which Wagner carries out his analysis, the flow of “controlled analogies” delineates kinds of relationships whose “essence” constitutes the kinship “system.” Kinship is the flow. As such, the analogues/analogies that flow are always partial versions of much larger analogical economies: an analogy is always an analogy of something else. That is why kinship surfaces as a controlled flow, for if the “essence of ‘kinship’ is restriction . . . the opposite of ‘kinship’ is therefore total, unrestricted analogy” (Wagner 1977: 639). In the knowledge economy, the analogue is the “externality”—the form of knowledge that self-detaches (spills over) itself and in so doing contributes to the largesse of public knowledge.7 Knowledge is the flow. Thus, if the opposite of kinship in the Daribi conceptual world is total unrestricted analogy, such totality, in the Euro-American knowledge economy, obtains at the expense of property. The opposite of property is total, unrestricted access to the knowledge commons. Working our way down from an idealized and unrealizable knowledge commons into the partial analogies that sustain it is, of course, what the task of an anthropological analysis ought to amount to.8 Unlike legal scholars, for instance, who tend to populate the detotalization of the commons with property rights and licenses, anthropology would turn instead to what Coleman and Golub called its techno-cultural forms, and what Leach dubbed its forms of enablement. Coleman and Golub’s ethnography elegantly reminds us that when information is at stake, our Euro-American liberal tradition finds it close to impossible not to theorize its practices of self-description as practices of democratic and ethical self-fashioning. Our theories of knowledge are theories of political ethics (Corsín Jiménez 2008): we own up to our information in the same terms through which we describe—i.e. analogously to—what we owe to each other.9 Sometimes we privilege privacy, sometimes we honour communication; sometimes we take pride in individual autonomy, sometimes we elaborate on the virtue of pedagogical transformation. Knowledge is both a means of transgression and an endpoint for romanticism. It is the Euro-American analogy par excellence of self and other. It does both vertical work (enhancing or supplementing our ethical pedigree) and horizontal work (a commodity that mediates further transactions). It fares as both theoretical commodity and methodological infrastructure. As Charlotte Hess and Elinor Ostrom have put it, it has “dual functionality as a human need and an economic good” (2003: 13). Knowledge is our (what we have in) common(s), a point to which I shall return in the conclusion. Thus, in the world of software, there are those, as Chris Kelty puts it, that see software as a product and those that see it as “part of the human itself, constitutive of our very freedom and, hence, inalienable. Extending software, through collective mutual aid, is thus tantamount to vitality, progress, and self-actualization” (Kelty 2008: 309). It is in this context that I find particularly useful James Leach’s study of F/LOSS in terms of enablement. Hackers, Leach observes, work with a binary form of judgment. This simple binarism compounds an ethics and a pragmatics—it offers justification both for moral and rational action, and in so doing generates the conditions of possibility for future actions. Such a binary model locates the actions of hackers in an infrastructural media—the media of the Internet. It is the binarism of the Net that mediates the social hopefulness of collaborative work. It is the “naturalisation” of machine language that gets “metonymically” extended by programmers into a module of “external reality” (Leach et al. 2009: 58-59). The machine capillarizes the world. Leach sees some cause for concern in F/LOSS’s binary form of judgment. Its pragmatics of truth-hopefulness, he intuits, are anchored on a form of moral reasoning based on the individual: The personal and the moral interlink in the activity of writing F/LOSS code. Pursuing one’s craft is also to behave ethically and to have a positive influence in society. The moral is presented as if it were merely an outcome of writing source code and fulfilling one’s personal ambitions. . . . Ethics and politics are pursued through making material objects which then in themselves, without the need for ‘social’ action, achieve political ends (64-65). Moreover, the culture of binarism flourishes on public and “flaming” tournaments of arrogance and bravado. According to Leach, in this context, what the culture of free software liberates is for the most part flows of maleness—women being conspicuously underrepresented in F/LOSS communities. The body of the freedom coder, it seems, is a male one. Binarism, however, summons forth a second kind of relation: the Network/Machine as a social relationship itself. Chris Kelty (2008) has offered what is perhaps the most comprehensive ethnographic account to date of the infrastructural and sociological dimensions of Free Software, and their entanglement in the production of what he calls “recursive publics.” What distinguishes a Free Software community from others, according to Kelty, is the work of “modulation” they lend to their own infrastructural capacities. They work on themselves by working on (writing code, editing, patching, compiling, improving) their infrastructure. Their capacities to act are therefore infrastructural capacities, in that it is the recursiveness of code as both a social and machine element that animates collective action. In Kelty’s words, “Geeks create andmodify and argue about licenses and source code and protocols and standards and revision control and ideologies of freedom and pragmatism” because these things “are reconstituting the relationship between liberty and knowledge in [our] technically and historically specific context” (2008: 10). Freedom, in other words, is sourced in the infrastructural equipment that furnishes the modern subject’s capabilities for social (network-mediated) action. This is of course an extendible or expansible notion of freedom, inasmuch as an agent’s capabilities for free action derive from the ongoing infrastructural transformation of the Net/Machine itself. With every “add-on,” freedom is recursively expanded. Our freedom depends on the Net’s own temporality as an infrastructure.10 Conclusion: Wide angle binoculars The entanglement of the Net/Machine’s temporality as an expandable horizon and our (net users) sense of freedom and enablement, throws into sharp relief, I would like to suggest, some aspects of our contemporary Euro-American conception of knowledge. Our sense of freedom as an epistemic enterprise delineates in one and the same move both the horizon of our social world and the political epistemology for its cognizance. We are defined as politically cognizant subjects by the Net/Machine’s own social relationality. What the Net intuits as potential freedom the subject interiorizes as epistemic knowledge. The ambit of our horizontal, netmediated exchanges ‘flows as an analogy’ of our vertical, autonomous sense of liberation / liberty. Knowledge is what we have in common. In the terms of the trompe l’oeil anthropology I have sketched above, what this implies is that it is the Net that perpendicularizes every knowledge-effect. It is the Net that mesmerizes us with its trompe l’oeil surface. The Net proliferates the points of view on/of knowledge. In this light, one is tempted to suggest that it is the Net/Machine, too, which deploys the image of knowledge as both an encompassing and encompassed object: it is the Network that tricks into view the allure that the Commons exercises over the imagination (its futures, promises and liberating potentials—its vanishing points), as well as the Network that affords concern over “narrower” issues, over bytes, licenses, privacy and rights of access (its objectives, properties and standards—its points of view). Of course this is the point made a while ago by Annelise Riles (2001)—that the Network may be “seen twice.” For Riles, the network refers “to a set of institutions, knowledge practices, and artifacts thereof that internally generate the effects of their own reality by reflecting on themselves” (3). The Network’s internal relationality (as an infrastructural network) defines the subject’s internal conception as a relational and social person. As she puts it, “networkers deploy the optical effect of Network form as a ‘fulcrum or lever’ that generates alternative inverted forms of sociality by projecting an image of each—Network and ‘personal relations’—from the point of view of the other (115-16, emphasis added). A baroque device like no other, the Network’s trompe l’oeil surface holds an alternation of points of view within a singular optics. Chris Kelty describes such optical effects alternatively by noting that the “singularity of the Internet is both an ontological and an epistemological fact” (2008: 306). My aim in this article has been to try to elucidate the internal organisation of the epistemic form that allows for such duplex descriptions. I have shown how the epistemological organisation of “perspective” is fuelled from within by what Hanneke Grootenboer calls the engine or form of baroque allegory (2005: 163). I have further traced such allegorical effects in the “axial,” “anamorphic” and “optical” heuristics employed by Wagner, Viveiros de Castro or Riles. An underlying concern throughout has been to understand what enables the allegorical form of perspective to travel as a modern anthropological device (see also Corsín Jiménez 2011). Recently Marilyn Strathern has analysed the duplex qualities of Euro-American knowledge and exposed the role played by “science’s relation” in the organisation of its epistemology: where the concept of the “relation” does double work as both discoverer of existing co-implications and inventor of new connections (2005: 3349). “The relation,” for Strathern, is the epistemological device that self-describes and transforms the modernist project. Here I have sought a similar yet not-quite-replicated effect, in that Euro-American epistemology is caught by surprise and entrapped in a pictorial/allegorical manoeuvre. This playful surprise has led me to unearth in turn a role for “perpendicularity” in the internal configuration of such epistemology. What anthropological form must knowledge take if it is to produce duplex/optical/ reversible effects? A tentative answer suggests that it might be a form of (Daribi) kinship at perpendicular angles. At perpendicular angles there is another quality of knowledge that is made visible, namely, knowledge’s own angular effects. I build here on a recent piece by Marilyn Strathern (2011), who has picked up on the binarism that characterizes Euro-American knowledge practices at large. It has not always been noted, Strathern observes, that in this epistemology “distinctions easily contain or lead to other distinctions” (91). Knowledge self-analogizes itself through ongoing bifurcations. Every new analogue of knowledge faces a wide open terrain wherein to analogize itself further: “there is a kind of arithmetic here that unitizes or individualizes the concepts that drive arguments—and any concept can be individualized—so that there is great latitude in what can be brought into conjunction” (ibid., emphasis added). With Strathern, I wish to bring attention to the latitude or wide angularity of our conceptual habitus: commons, freedom, publicity, choice. Concepts all that appear to carry within their own sense of amplitude. Wide concepts for a world wide open. It would appear that in this epistemology it is hard not to look at the world with wide-angle binoculars. Acknowledgments I am very grateful to HAU’s two anonymous reviewers for their close reading of an earlier version of the article. Their comments have help made the argument much stronger. I am also thankful to Stéphane Gros for his patient and encouraging editorial work. References Benkler, Yochai 2006. The wealth of networks: How social production transforms markets and freedom. New Haven: Yale University Press. Boyle, James. 2003. “Foreword: The opposite of property?” Law and Contemporary Problems 66:1–32. Coleman, Gabriella E. and Alex Golub 2008. “Hacker practice.” Anthropological Theory 8: 255–277. Cornes, Richard and Todd Sandler. 1996. The theory of externalities, public goods, and club goods, Vol. 2. Cambridge: Cambridge University Press. Corsín Jiménez, Alberto. 2008. “Introduction: Well-being’s re-proportioning of social thought.” In Culture and well-being: anthropological approaches to freedom and political ethics, edited by Alberto Corsín Jiménez, 1–32. London: Pluto Press. ———. 2011. The strabismic polity: An anthropological optics of political modernity http://sites.google.com/site/acorsinjimenez/articles/strab_polity_110516.pdf?attredirects=0. Crook, Tony. 2007. Anthropological knowledge, secrecy and Bolivip, Papua New Guinea: exchanging skin. Oxford: British Academy/Oxford University Press. David, Paul A. 2000. A tragedy of the public knowledge “commons”? Global science, intellectual property and the digital technology boomerang. Stanford Institute for Economic Policy Research Discussion Paper Series, 2. Stanford University. David, Paul A. and Dominique Foray 2003. “Economic fundamentals of the knowledge society.” Policy Futures in Education 1:20–49. Foray, Dominique. 2006. The economics of knowledge. Cambridge, Mass., and London: The MIT Press. Frischmann, Brett M. and Mark A. Lemley 2006. “Spillovers.” http://ssrn.com/abstract-898881. Grootenboer, Hanneke. 2005. The rhetoric of perspective: Realism and illusionism in seventeenth-century Dutch still-life painting. Chicago and London: The University of Chicago Press. Hess, Charlotte and Elinor Ostrom. 2003. “Ideas, artifacts, and facilities: information as a common-pool resource.” Law and Contemporary Problems 66: 111–145. Kelty, Christopher M. 2008. Two bits: the cultural significance of free software. Durham and London: Duke University Press. Leach, James, Dawn Nafus and Bernhard Krieger. 2009. “Freedom imagined: morality and aesthetics in open source software design.” Ethnos 74:51–71. Riles, Annelise. 2001. The network inside out. Ann Arbor: The University of Michigan Press. Stiglitz, Joseph E. 1999. “Knowledge as a global public good.” In Global public goods: International cooperation in the 21st century, edited by Inge Kaul, Isabelle Grunberg, and Marc Stern, 308–324. New York: Oxford University Press. doi.org/10.1093/0195130529.001.0001. Strathern, Marilyn. 2005. Kinship, law and the unexpected: relatives are always a surprise. New York: Cambridge University Press. ———. 2006. “Intellectual property and rights: an anthropological perspective.” In Handbook of material culture, edited by Christopher Tilley, Webb Keane, Susanne Küchler, Michael Rowlands and Patricia Spyer, 447–462. London, Thousand Oaks and New Delhi: Sage Publications. ———. 2011. “Binary license.” Common Knowledge 17:87–103. Viveiros de Castro, Eduardo. 2010. Metafísicas caníbales. Líneas de antropología postestructural (Spanish translation of Métaphysiques cannibales. Lignes d’anthropologie post-structurale.) Buenos Aires and Madrid: Katz Editores. Wagner, Roy. 1977. “Analogic kinship: a Daribi example.” American Ethnologist 4:623–642. La parenté Daribi sous des angles perpendiculaires : une anthropologie en trompe l’œil Résumé : Quels sont les contours d’un « savoir » qui servirait à la fois de bien public ou commun et de source de prise de pouvoir ainsi que de liberté individuelle ? Cet article offre une analyse de l’organisation épistémologique d’une « économie de la connaissance » en passant un travail ethnographique sur le Free Software à travers un trompe l’œil espiègle, par le biais de l’œuvre classique de Roy Wagner sur la parenté Daribi. Il propose un modèle préliminaire pour penser la connaissance euro-américaine en tant que dispositif trompe l’œil elle-même. Alberto Corsín Jiménez is Senior Scientist at the Spanish National Research Council. Previously he was Dean at Spain’s School for Industrial Organization (EOI) and University Lecturer in the Anthropology of Organizations at the University of Manchester. He is the editor of Culture and well-being: Anthropological approaches to freedom and political ethics (Pluto, 2008) and The anthropology of organisations (Ashgate, 2007), and is currently revising a book manuscript for publication, provisionally titled The strabismic polity: An anthropological essay on the political optics of modernity. His areas of interest lie in the history and anthropological theory of knowledge practices, and in particular their contemporary expression in science/management/public encounters. ___________________ 1. Wagner’s piece opens thus: “Let us begin with the proposition that all human relationships are analogous to one another” (Wagner 1977: 623). As I hope to show in the rest of the article, the terms of the paraphrase, and the substitutions employed therein, are offered in the spirit of Wagner’s original investigation into the work of “analogy.” I am interested in what may internally configure “analogy” as an enabling epistemological device itself. I thank one of the reviewers of the article for prompting me to clarify the artifice of my opening analogy. 2. Leach reminds us that “enablement” is a self-descriptor used in F/LOSS (Leach et al. 2009: 55 and 69, fn. 5). 3. Here I follow Marilyn Strathern’s useful recent convention, where she distinguishes between “perspectivalism” and “perspectivism.” The former “rests on a mathematics that seeks to resolve the world into discrete entities,” whereas the latter ‘implies an ontology of many worlds” (Strathern 2011: 92). The former enacts a plural world, the latter multiple worlds. See note 4 below for some additional comments on the “perspectivalist” vs. “perspectivist” distinction. 4. Although I cannot offer a full elaboration of the epistemology and heuristics of “perspective” here (see Corsin Jimenez 2011), I shall briefly point out that as a general epistemic device, the work of perspective does not accommodate itself to a simple dichotomy between classical Renaissance or Cartesian “perspectivalism” and Amerindian or, indeed, Daribi “perspectivism.” For a start, the modern tradition carries a split within: the work for example of Baroque (anamorphic, trompe l’oeil) perspective is very different from that of Cartesian perspective. Indeed there are arguments, and Hanneke Grootenboer’s wonderful book is an example, that suggest that the whole structure of “perspective” is fuelled from within by a rhetorical or allegorical power (Grootenboer 2005: 163). It is the baroque form or expression of perspective that shows that all perspective is in fact allegorical. What merits attention, then, is how this allegorical power travels when adopted as a heuristic by recent anthropological theory—what enables the anthropological form of “perspective” to be internally organized as allegory? 5. An anamorphic projection may not require an exact 90° (perpendicular) adjustment of our point of view to bring the image back into its proper shape. 6. Note that the substitution of “axes” for “perspectives” is already insinuated in Wagner’s piece, when he observes that kawa and pagekamine “amount to the same thing seen, as it were, from different angles” (Wagner 1977: 628). The word “angle” invokes the perpendicularity between the horizontal and the vertical. At the same time, however, these axes suddenly emerge as simultaneously enabling situated points of view (“things seen from”). 7. What kind of epistemic objects fulfil the conceptual imagination of public knowledge as an externality? Mathematics is the example preferred by economists (for example, Stiglitz (1999: 308) and Foray (2006: 114-15).) Once a mathematical theorem is made public, no one can be directly excluded from learning it, nor can they be taken the knowledge away once they have learned it. (An underlying assumption here is that people have the skills to learn the theorem. The assumption is problematic.) The knowledge is “external,” has escaped its conditions of production, is free to travel, and is available also for anyone to consume. 8. After Wagner, who notes that “an (unrealizable) ideal of total analogy is detotalized and distributed over a range of partial realizations, each corresponding to a kind of relationship. A kind of relationship (designating the particular kinds of relatives proper to it) can then be considered as an analogue of relationship in general, diminished and restricted in certain dimensions so as to control and channel the flow of relational analogy” (Wagner 1977: 639). If the knowledge commons is our total (idealized and unrealizable) analogy, what are its partial realizations? 9. Tony Crook has described the internal equation of knowledge with the social in terms that I find particularly illuminating: “One root of the problem here is a peculiar euro-american habit of eliding the solidity of a social relationship with the solidity of knowledge—and then taking one as a measure of the other” (Crook 2007: 10). Note that what is reciprocally “owned” here is “the social” as an epistemic form. We might say that the flow of knowledge lends the social a certain tangibility. The result is not to be confounded with what Marilyn Strathern calls a transactional regime (not unlike the organisation of Daribi exchanges): “Transactions thereby create a form of ‘multiple ownership’. In so far as they determine what is transactable they can also determine the parties to the transaction. Thus ‘groups’ may ‘emerge’ at the time of the interchange itself, identified as parties to the transaction. The gaining and disposal of things do not extinguish the parties’ mutual interest in the things or in one another: in ‘owning’ the flow of items they ‘own’, as an intangible thing, the relationship between them” (Strathern 2006: 454, emphasis added). 10. Kelty thus observes that the question of Free Software’s “modifiability” (its source code being available for reuse and transformation) “raises a very specific and important question about finality. When is something (software, a film, music, culture) finished? How long does it remain finished? Who decides? Or more generally, what does its temporality look like, and how does that temporality restructure political relationships?” (Kelty 2008: 11).