Translating worlds

This work is licensed under the Creative Commons | © William F. Hanks and Carlo Severi. ISSN 2049-1115 (Online). DOI: http://dx.doi.org/10.14318/hau4.2.001


Translating worlds

The epistemological space of translation

William F. HANKS, University of California, Berkeley

Carlo SEVERI, École des Hautes Études en Sciences Sociales

Translation has played an important but equivocal role in the history of anthropology and linguistics. At least since Saussure and Boas, languages have been seen as systems whose differences make precise translation exceedingly difficult, if not impossible. More recently, Quine has argued that, in purely abstract terms, reference is ultimately inscrutable and translation between languages is in principle indeterminate. From a Kuhn-inspired point of view, we argue, on the contrary, that the challenge posed by the constant confrontation of “incommensurable” (yet translated) paradigms may become a field for ethnographical inquiry. This approach can provide a new anthropological way to define translation, not only as a key technique for understanding ethnography, but also as a general epistemological principle. Social anthropology would be thus defined not only as the study of cultural differences, but also and simultaneously as a science of translation: the study of the empirical processes and theoretical principles of cultural translation.

Keywords: translation, incommensurability, culture, ethnography, epistemology

Why translation?

Translation has played an important but equivocal role in the history of anthropology and linguistics. Linguists perform multiple translations, usually starting from an acoustic image of speech, or a visual image of a sign, which is transcribed in more or less phonetic detail and subjected to morphological, syntactic, semantic, or pragmatic analysis, depending upon the empirical focus and theoretical framing of the work. In the course of analysis, the object language is translated into the formalism of linguistic description. Even when not explicitly comparative, all of linguistics is virtually comparative insofar as formalisms are assumed to be applicable to many or all languages (Benveniste 1974, 2012). This also implies translating object languages and their grammars into typological categories. The dynamic field of linguistic typology is an outgrowth of this, with much to teach us about translation. Contemporary social and cultural anthropology are of course less formal than linguistics, but no less engaged in translation and comparison. When field notes are recorded, social institutions and discourses are analyzed (such as kinship, residence patterns, exchange or ritual practices), and ethnographic descriptions are crafted, translation is present in every step. And even when not overtly comparative, social and cultural anthropology inevitably involve comparison, and this means translation into some set of terms and concepts that can mediate between the differences among societies and best capture their particular dynamics. The importance of translation is therefore that it is through its multiple varieties that both disciplines constitute their objects and formulate generalizations.

The equivocal status of translating has several sources. First, with the exception of well-known debates in British social anthropology in the 1960s and 1970s, relatively few anthropologists would describe their own craft as a kind of translation—although the issue is coming back to the fore in the recent literature: see Pym (2010), Asad (2010), and Viveiros de Castro’s (2004) provocative argument that translation rests on a kind of “controlled equivocation.” For a working linguist, very careful attention is paid to analysis, but this analysis is rarely understood as a form of translation. For both fields, cross-language glossing, say from Kuna or Maya into French or English, is a mere heuristic. The load-bearing evidence comes not from a gloss, but from the social or linguistic practices themselves, and it is relative to their own social contexts that utterances, actions, or events gain their meaning. For these reasons, the previous paragraph may appear contentious to some readers. It would seem that we translate from the original mostly in order to abandon the translation in favor of the original. We know in principle that any translation is selective, which implies loss of features from the original, and that any translation also adds in supplementary features absent from the original (Benjamin [1923] 2004; Berman 2008). Ironically, the process of successive failed translation may be our best tool in discerning what is specific to any object society or to any “original.” In other words, it becomes a method, as we will show in the next section. Translation is both how we constitute our objects and how we make claims about them. This equivocal duality surely raises the risk of circularity or at least incorrigibility.

A second source of equivocation lies in the fact that fully accurate translation is exceedingly difficult, if not impossible, and yet translation is ubiquitous in social life. We do it all the time. Not only experts translate, but ordinary speakers do too, in the course of everyday activities. Bi-or multilingualism, code switching, blending, crossing, paraphrasing, reported speech, and giving accounts are all well-established sociolinguistic phenomena, and all may involve the same key elements as canonical translation. The fact that they are part of everyday practice, and not only of social science research, is a good reason to pay close attention to translation as a process endogenous to social life.

A related source of equivocation is that while we think of translation as operating across languages or social worlds, it is a robust feature of any individual social world, even in monolingual or monocultural societies (if such actually exist). There is a strong line of argument to the effect that understanding is itself a matter of translation: the object understood is translated into some variety of interpretant or representation on the part of the understander. Following Peirce (1955), this can be thought of as a mental representation, a corporeal response to the object, or a variety of other kinds of sign, but in each case, the interpretant can be said to translate its object into an understanding of it. We expand on this point in the light of American pragmatism in the next section. The key point for now is that in the passage of meaning from one language or society to another, translation does not come into play only after the translator has understood the original. It is not an ancillary rerendering or glossing, but is itself the basis for understanding. Translation in one or another variety is always already in play, long before the overt act of rerendering some social object into a foreign language. This is why we speak of an epistemological space of translation. At stake is what we can know, how we can know it, and how we can make it known.

Translation at a general level is too widespread and the concept is too powerful to let it run loose. Our aim in these papers is to relate the anthropological issues of commensurability, description, and understanding to the linguistic issues of determining and redescribing meaning, at whatever level. We start from the conviction that while different varieties of translation raise different questions, there are important commonalities. By combining, rather than isolating, linguistic and social analysis, we can improve both and point the way to a better theory of translation at all levels. This, we believe, will open a horizon for research in both fields.

From relativity to indeterminacy and incommensurability

At least since Saussure ([1916] 2006), Boas (1989), Whorf (1956) and Sapir (1985), languages have been seen as systems whose differences make precise translation exceedingly difficult, if not impossible. In fact, cross-linguistic differences of nearly equivalent expressions are Saussure’s preferred evidence for his signal concept of arbitrariness. The pairing of sound with meaning is arbitrary in that it is conventional, and the best evidence of this conventionality is cross-linguistic differences. Given what is ostensibly the same statement translated into, say, French, English, and Kwakiutl, each language expresses it in a unique way, according to its own way of “cutting up reality” into linguistic elements.

For Boas, this function of categorization was central, and he argued convincingly that it has consequences for how speakers of different languages perceive the world. His argument was not that language limits perception, but that the routine expressive patterns of one’s native language, especially the obligatory categories, render automatic or unreflective certain features of the worlds we describe. Marking of person, number, tense, deixis, noun classes, and phonology provide well-known examples in which the native speaker is induced to attend to the corresponding features of the scenes (s)he describes. The relativity effect is not about what a native actor can express or understand, but what (s)he usually does express or understand. This gave rise to what is known as classic linguistic relativity in the writings of Edward Sapir and Benjamin Lee Whorf. The former expanded the Boasian focus on categories to include the major grammatical systems of the language, drawing heavily on analogy, while the latter emphasized “habitual ways of speaking” in which languages are used. For both, as for Boas, the twin facts of cross-language difference and intralanguage norms of expression combine to guide or channel the perceptual and expressive habits of speakers. Although it has proven very difficult to demonstrate relativity effects conclusively, there has been a resurgence of research in the area over the last two decades (Hill and Mannheim 1992; Lucy 1992; Levinson 2003; Leavitt 2010; Enfield and Levinson 2006; Enfield and Sidnell 2012) as well as a large literature in cognitive and psychologically oriented linguistics (see, e.g., Gentner and Goldin-Meadow 2003). Like the translator, the relativity theorist must know at least two languages, contrasting them in order to better understand the specificity of each. Unlike translation, however, relativity theory, at least in it mainstream variants, has been based on the model of the monolingual speaker. When Whorf describes how Hopi speakers conceptualize time, or when Boas makes claims about Kwakiutl, they are imagining the native speaker caught in the grips of the native grammar. But what if the Hopi or Kwakiutl speaker is also a fluent English speaker? For a classical relativist, (s)he would be caught in a sort of parallax in which competing constraints vie for causal impact on her or his expressive and perceptual habits. Alternatively, the bilingual speaker might be subject to both systems and their respective habitual patterns of expression, with the dominant role at any moment played by whichever language the speaker is currently speaking. But this too is confounded by any variety of language mixture or blending in which single utterances may contain elements from two or more languages. The scope of relativity would become a pragmatic problem, not a semantic or grammatical one. And of course, bilingual speakers often translate, restate, and paraphrase both between and within their languages. As an analytic method then, translation underwrites relativity by providing evidence of cross-linguistic difference. As an endogenous social practice, though, it undercuts relativity by weakening the grip of any one language on the expressive habits of the bilingual speaker.

Another line of intersection between semantics, pragmatics, and translation emerges in the work of the analytic philosopher Willard Van Orman Quine. In his well-known book Word and object (1960), and in a series of related papers, Quine argued that reference is ultimately inscrutable and translation between languages is in principle indeterminate. Distinguishing observation sentences, like “this is a rabbit,” from standing sentences, like “democracy is a social good,” Quine shows that even the former are ultimately impossible to define with precision. The more general standing sentences are, by extension, also impossible to pin down. In his famous thought experiment, the “radical linguist” (not to be confused with an actual linguist) confronts an entirely unknown language, without the benefit of a mutually understood contact language. Trying to determine the precise meanings of words in this other language, Quine’s “linguist” finds it impossible to determine through ostension alone what terms like “gavagai” mean (apparently “rabbit,” but perhaps better rendered “undetached rabbit part,” “rabbit phase,” etc.). Quine presses this dilemma, and the critique of empiricism it implies, from the limits on intelligibility of an unknown foreign language all the way to limits on understanding our own language, at which point reference truly goes “inscrutable.” Philosophers, linguists, and some anthropologists were quick to pick up on the significance of his argument, which has remained a major position in the field today.

Quine’s thought experiment, and the very idea of radical translation, are explicitly distinguished from the actual practices of linguists and anthropologists in the field. Nevertheless, there are certain distortions that call for comment because they put in question the applicability of his conclusions to actual translation. First, Quine posits a monolingual native speaker who is evidently incapable of formulating meaning statements in his own language. That is, the native informant is never given the opportunity to state in his own language what the term “gavagai” means, nor is the radical linguistic given the opportunity to ask him, simply, “What does gavagai mean?” It is clear that for Quine metalinguistic questions and responses inherit the very same limitations as the term they would seek to define. The problem with this from a linguistic perspective is that it is empirically false: metalinguistic statement may be multiplied, covaried with variations in the object term, and used to establish an array of contrasting terms which, in the aggregate, severely constrain the possible semantic range of the object term. The result may be a range of possible translations, but surely not an indefinite range. In a sense, Quine’s native speaks a language in which neither paradigmatic nor syntagmatic contrasts can be deployed in order to discern the semantic boundaries of a term. Yet these contrasts are part of what any human language provides, and they are central both empirically and theoretically.

Second, Quine stipulates that in order to overcome indeterminacy, the linguist must derive an exact meaning from ostensive reference alone, and, moreover, that choice among alternative possible translations will ultimately turn on the linguist’s assessment of what is most “natural,” in the sense of corresponding best to his own (alien) sense of naturalness. Thus the linguist is, as it were, at sea without an independent point of reference. Neither of these problems is insuperable in linguistic fieldwork. The linguist gathers a wide variety of evidence for meaning hypotheses, including usage, metalinguistic commentaries, the ways in which the target term combines with others terms in syntactic constructions, analysis of the internal structure of the form (including morphology, compounding, etc.), and grammatical evidence of oppositions between the form and other forms in the language. Similarly, it is not what appears natural in the linguist’s native language that guides his or her choice among alternative translations, but rather all that is known of universals of language, and the possible arrays of distinctions that are encoded in lexical forms. Quine is very clear that radical translation is a philosophical thought experiment and not a description of empirical research, but the point is that the problems that arise in his thought experiment arise from a deeply distorted set of assumptions about both languages and how they are analyzed. A third critique from linguistics would observe that for many expressions, it is not the semantic boundaries of words that distinguish them, but their focal or most prototypical meaning (as in the extensive literature on color terms). The implication of this is that the apparent imprecision of some semantic boundaries is to be expected, and need not reflect inherent limitations on what the foreign linguist can discern. Finally, Quine shifts inconsistently between “the anthropologist” and the “radical linguist” in his experiment, yet the two disciplines have quite different views of language and their methodologies are correspondingly distinct. One of the main goals of this special issue is precisely to bring the two disciplines together.

Focusing on the difficulty of translating technical terms from ancient science into the language of modern science (and by extension between any two scientific paradigms separated by revolution), Thomas Kuhn (2000, 2012; Kuhn et al. 2000; see also Hallen and Sodipo 1997) famously developed the concept of incommensurability. Kuhn, a historian-philosopher, is more careful than Quine to acknowledge the relation between translation, language learning, bilingualism, and the actual practices of translators. Unlike Quine, Kuhn rejects the equation of meaning with reference, insisting on the importance of style, nuance, and the difference between translation and interpretation. In his later work, he puts translation—its limits, potentials, and unavoidability—at the center of his views, and he proposes to view scientific communities as speech communities (Kuhn 2000: 166) where many forms of translation are constantly carried on, despite the theoretical incommensurability of paradigms. Translation ceases to be defined as an abstract impossibility. The challenge posed by the constant confrontation of “incommensurable” (yet translated) paradigms becomes in itself, on the contrary, a field for ethnographical inquiry.

Ethnography as translation

For most field linguists and ethnographers, translation is of limited utility. The empirical evidence for a social or linguistic category or practice must always be from the native language, not from a translation. It is common sense that if you do not work in the language you are trying to describe, you are missing your object. Interlinear glosses, explanations, and such in a European metalanguage are heuristic devices for the reader to follow distinctions made in the object language (e.g., a description of Kuna [object] in French or Spanish [metalanguage of description]).

However, it is also clear that ethnography, from a theoretical point of view, is unimaginable without translation. As Conklin has remarked, “The problems of ethnography are in the largest sense those of translation” (1968: 12). This is true not only because almost any ethnographer faces the task of translating words and concepts from one language to another, but also because to “do ethnography” is to make descriptions, judgments, actions, and theories proper to a specific culture understood in the language of social anthropology (a scientific community in Kuhn’s sense). Seen from this point of view, translation ceases to designate only a linguistic technique. It becomes the definition of the core strategy of social anthropology itself.

Clearly, translation is a multidimensional phenomenon. A first dimension naturally concerns language. Jakobson (1959) and Benveniste (1974), in dialogue with Peirce (1955), both distinguished standard linguistic translation from one language to another from intralinguistic translation (restatement, gloss in the same language), and eventually from cross-modal translation (e.g., words to gestures, or verbal description to pictorial blueprint, or vice versa). However, even in a relatively language-centric view of translation, there are massive questions regarding the criteria under which some expression may be considered a “translation” of some other. In the last fifty years of research, linguistic anthropologists have made enormous strides toward understanding speech practices, developing socially embedded pragmatics (in the Anglo-Saxon sense based heavily on Austin [1975] and Grice [1989]), the ethnography of speaking, the study of indexicality, metalanguage and reflexive language, metaphor, style, conversation analysis, and increasingly the study of multimodal relations between speech, gesture, and material setting. As ever more aspects of communicative situations have been shown to frame talk, the question of what a source text even means has only become more subtle, let alone which aspects of meaning should be conveyed (under what circumstances) in a valid translation. If the translation is into language, as opposed to some other medium, then it too has all these elements of meaning. The simple pairing of head terms in one language with translations in another, as in bilingual dictionaries, obscures the fact that even the simplest lexical translation is multidimensional. When we move to pragmatically enriched utterances, the task becomes astronomically difficult. Therefore we must ask which elements of a putative meaning (in the source text) need to be present in a valid translation (the target text). Once we have answered that, we must decide which aspects of the translation itself must carry the critical information. Thus we confront the irony that valid translation gets ever more difficult to conceive the more we know about languages.

What we can know as anthropologists, how we can know it, what counts as warrant for knowledge, how we can express it in the language of our discipline, the sources and limits of anthropological knowledge—all of these questions engage translation either in principle or as a matter of fact in the practice of anthropology. The problems of inscrutability, indeterminacy, and incommensurability raised by Quine and Kuhn, for instance, all presume that the problem lies in capturing in one language meanings expressed in another. But when we look at translation as a historical practice, as Hanks (2010) does for the translation of Catholic doctrine into colonial Maya, we find a very different set of factors. For one thing, the target language is altered in the process of translation, which is pervasively and systematically neologistic. For another, translation is one part of a much broader colonial process involving religious conversion, conversion from hieroglyphic to alphabetic writing, and the reorganization of the political geography. Here not only is translation required on our part to capture this social world, but translation in the colonial setting produces the objects of our historical knowledge.

An analogous problem arises when we consider intracultural translation between significantly different registers of the same language. For example, can the language of a shaman in performance, or a ritual performer more generally, be rendered accurately in the ordinary nonritual versions of the language? Can it be translated into ordinary talk? Or more broadly, can the nonspecialist, say the patient, understand what the shaman is doing, and at what level of understanding? In fact there are significant differences between what the shaman in his own terms is doing in performance, on the one hand, and what the patient thinks he is doing from an everyday frame of reference, on the other. The asymmetries in their respective knowledge of what is going on are great, but they do not cause breakdown in communication. Hanks (2006, forthcoming) argues that the asymmetry of knowledge between participants, and the constraints on translation between the esoteric language of shamanism and ordinary Maya, are actually resources for shaman-patient interaction, not impediments. This in turn suggests that meaningful and consequential interaction can proceed in the absence of mutual translatability, or perhaps even intelligibility.

Translating worlds: Cognition, ontology, and the science of translation

All we have seen until now suggests that there is more to translation than language. In cultural practices, translation constantly goes beyond it, for at least two reasons. First, this is because the concept of translation implies all forms of “social traductions.” It designates the exchange not only of words, but also of values, theories, and artifacts from one culture to another, for instance in such processes as religious conversion, cultural mimesis, or messianic movements (Severi 2002, 2004).

The second reason translation exceeds language is because nonlinguistic forms of translation are constantly present in cultural traditions (Severi 2012). Words are translated into images, music into words, and gestures into objects. Furthermore, even within a single culture, translation processes enable the passage from one context of communication to another. Virtually everywhere, such formal contexts of the expression of meaning as ritual action, play, and other forms of performance generate their specific “ontologies.” Things, artifacts, and living beings may then crucially change their nature, as in the famous “qualitative analogy” which transforms a cucumber into an ox in Evans-Pritchard’s analysis of the Nuer sacrifice (1940). In these cases, the interpretation of such formal contexts of cultural representation transforms translation into a way to translate “worlds” (defined as “oriented contexts for the apprehension of reality”), not just words, or other ways to express meaning.

Translation is still, most of the time (see, e.g., Sammons and Sherzer 2000; Rubel and Rosman 2003; Silverstein 2003), with few exceptions (e.g., Keesing 1985), discussed only in technical terms. As an epistemological principle, it seems almost absent from the contemporary epistemological debate in the discipline. Two dominant trends in particular, cognitivism and ontologism, seem unable to understand what an epistemology of translation could be. For many cognitive anthropologists, and particularly for those adopting what Levinson (2003) has called “simple nativism” (e.g., Sperber 1996; Bloch 1998; Baumard, Boyer, and Sperber, 2010), this lack of understanding is a consequence of their denial of the epistemological import of cultural variations. If cultural differences have no fundamental influence on human cognition, where concepts (such as “continuity, solidity, gravity and inertia” [Spelke et al., 1992]) already exist independently from language, then translation is merely an ability to express a preexisting mental representation in the “phonetic clothing” (Levinson 2003: 28) of a specific language. It has, in itself, no cognitive relevance. As, for instance, Pinker puts it, “Knowing a language . . . is knowing how to translate mentalese into strings of words and vice versa” (1994: 82). Levinson has recently pointed to the two main difficulties generated by this approach: “First it is impossible to reconcile with the facts of variation across languages. Second, it is a theory of innate (thus biological) endowment outside biology” (2003: 26).

Ontologism raises other questions. Descola ([2005] 2013), for instance, distinguishes animism, totemism, analogism, and naturalism as implying different ontologies—different ways of going beyond any simple construal of nature vs. culture vs. other. In perspectivism, as developed by Viveiros de Castro, translation emerges as a sort of “controlled equivocation.” So if the Jaguar offers manioc beer and what you find before you is human blood, thick and frothy, you have a perspectivist difference on the world of objects (Viveiros 1998, 2004). But what about the grey zones between perspectives or between modes of existence? What of the blending or grading or switching between systems, which surely occurs over the history of colonialism in the Americas, for example? Any typological schema will face the question of blends, and if we use translation to name the process(es) that cross between types, then we will have to deal with blending, and the emergence of new types or varieties. How, then, shall we translate ontologies, given that they are emergent and not static, blended and not pure?

We think the proper starting point of a new theory is neither preemptive universalism (which is almost always Eurocentric) nor typological divisions (which bound off individual types of ontology, even as they pluralize their space), but instead the study of processes and principles of translation. Given the scope of translational practices and the history of our field, the epistemological stakes of this (late-)Kuhn-inspired approach are high.

We should consider the study of these processes and principles of translation not only as an important way to improve a number of key technical operations for the interpretation of ethnography, but also as a new way to reformulate the general epistemology of our discipline.

The contributions

In this issue, we have gathered a number of papers where these questions are treated from different points of view. In his paper, William Hanks argues that intracultural translation plays a constitutive role in the social life of any human group, and not only in mediating between different groups and languages. This is evident in all varieties of reported speech, paraphrase, commentary, and exegesis. These share with translation two features that distinguish it from other kinds of interpretation: a translation both refers to and paraphrases its source text. Hanks argues that it is the target language into which one translates that ultimately constrains the process. An adequate target language must be functionally capable of self-interpretation through metalanguage. Cross-linguistic translation presupposes intralinguistic translation. Historical examples of languages changing through intertranslation abound in (post)colonial contexts in which authoritative texts in a dominant language are translated into a subordinated language. This process inevitably alters the semantics and pragmatics of the subordinate language. The direction, scope, and depth of change are historically variable. Examples are adduced from modern and colonial Yucatec Maya and Spanish.

Carlo Severi is concerned with the relationship between translation and thought processes. He argues that forms of thought, from what Lévi-Strauss called the “systematization [of] what is immediately presented to the senses” to the causal theories studied by Evans-Pritchard in witchcraft, have generally been interpreted as an expression of a specific language or “culture.” In this paper, he discusses this way of defining thought. Three classic objections are examined: (1) Societies sharing the same “system of thought” may speak different languages, and vice versa. (2) If a relation between language and thought exists, it is an indirect and controversial one, and we should never take it for granted (or infer qualities of thought from language structures) without further investigation. (3) The languages that we use to qualify different kinds of thought are constantly translated. Through a discussion of the context of translation, Severi argues that instead of seeing the possibility of translation as a theoretical difficulty for defining thought, we could, on the contrary, consider the ethnography of translation as a chance to observe the dynamics and structure of thought processes, and to study how they operate in different cultural contexts. Using three Amazonian examples, Severi describes the kind of cognition involved in the form of translation that Jakobson calls transmutation. From this ethnographic analysis, we can derive not only a better (both wider and more precise) idea of some, rarely studied, cultural translation processes, but also draw from it a new way to define the concept of “cultural ontology,” both for Amazonian cultures and in more general terms.

Rupert Stasch presents a case study in the historically common phenomenon of a contact community between persons lacking any common language, such that their linguistic interactions are focused on linguistic otherness as such, or are mediated by a miniscule number of translation specialists. Stasch first explores contrasts in how international tourists and Korowai of Papua each take up the other’s difference of language as a figure around which to express primordial definitions of their relation. He then examines how mutual incomprehension is valued as a resource for staying separate. Finally, he analyzes how the role of tour guide, qua translator, embodies a political model of authoritative speakerhood that is antipathetic to Korowai egalitarianism, but nonetheless fosters egalitarianism-oriented paths of engagement with the quite different political formation of tourists’ home social orders.

Anne-Christine Taylor’s paper analyzes a series of intra- and intercultural translations involved in the shamanic practices of the northern Jivaroan Achuar. First, it shows how certain states of suffering, experienced as an unwanted metamorphosis of selfhood, are reframed in the course of shamanic healing rituals as the symptoms of an insidious process of disempowerment and “whitening” unleashed by other, enemy Jivaroans. The curing session conflates the victim’s sickness and the history of interethnic relations, construed as a painful process of involuntary qualitative change. A further series of translations come into being when the cure fails and the patient abandons his Jivaroan identity and moves into a lowland Quichua identity; this involves mapping the implicit autobiography of a Jivaroan moving from illness toward recovered health and social agency onto Quichua narratives of their own history. However, owing to increasing closure of ethnic groups, Jivaroans nowadays have to deal directly with the spoken and written words of the Whites, and this involves new forms of translation evoked in the final part of the paper.

Alan Rumsey’s paper deals with two kinds of translation among Ku Waru people in the New Guinea Highlands: (1) translation between the local language and the national lingua franca within everyday interactions between young children and their caregivers; (2) intercultural translation between the story world of a local genre of sung tales and the contemporary lived world of Highlands PNG as practiced by skilled composer-performers of the genre. Although these two kinds of translation take place on very different planes, they both operate in terms of a well-developed set of procedures establishing equivalence, between words and worlds, respectively. On both planes a key role is played by parallelism, suggesting a connection between equivalence in the ordinary sense of the word and in the specific sense of it that was developed by Jakobson with respect to parallelism—a connection which is significant for the understanding of translation in general.

In his paper, Adam Yuet Chau starts from a couple of questions: How would one translate the word “menu” (i.e., restaurant menu) into the native language of an (imaginary) tribal people (with no writing and no restaurants)? And how would you explain to them how ordering from the menu works? It quickly becomes clear that translating the word “menu” entails translating not only the world of restaurant-going and ordering from the menu but also our (i.e., ideal-typically Western) very conceptual and social world, which is another way to say that what seems to be a humble piece of paper listing a certain number of dishes is itself made by the world in which it is found and in turn contributes in a significant way to making that world. In this paper, Chau examines the restaurant menu as a world-making social and translocutional/transinscriptional technology (the menu as menu-logic and cosmo-menu). As a kind of text act that is situated at but one of many “iterative/inscriptional stations” along an indeterminate and continuous chain of translocutions and transinscriptions, the menu highlights the temporal dimension of all kinds of translations (translingual, intralingual, transmodal, transcultural, etc.).

In their paper, Emmanuel de Vienne and Carlos Fausto focus on a specific case of translation that was attempted in 2006 by a Kalapalo Indian (from Mato Grosso, Brazil). This man in his forties created a radically new liturgy and cosmology by combining elements borrowed from local shamanism and mythology, Christianity and TV shows, among other sources. He thus managed to convince entire villages to take part in spectacular healing ceremonies. Since one of these rituals was filmed by two Kuikuro filmmakers, it is possible to examine the precise mechanisms of this cultural innovation, and therefore address with fresh data and methodology the old issue of Amerindian prophetism. They propose the concept of translating acts as a means to describe this native practice of translation, which consists as much of gestures and ritual actions as of linguistic expressions, emphasizes practical effects more than the negotiation of semantic equivalences, and is subject to constant reorientations in the course of interaction.

In the Colloquia section, John Leavitt argues that the idea of translating worlds depends on the possibility that there are worlds to translate between. This has not always been the case in translation theory. This paper traces out some key moments in the history of translation theory in the West, which has shown an oscillation between what have come to be called “domesticating” and “foreignizing” approaches. The former seeks to present the referential meaning of the original work in an easily recognized and absorbed form for the reader. The latter seeks to preserve elements of the original work, and by implication its world, forcing the reader to work to reorient him- or herself, to cross a boundary into what is potentially another world, initially another language-world. The paper concludes with some examples drawn from Central Himalayan oral traditions.

In the Forum section, G. E. R. Lloyd acknowledges that the issues of translation and of translatability are general and concern the possibility of mutual intelligibility in many registers, including within a single natural language. Both anthropologists and ancient historians are faced with such problems, where the historians are at a disadvantage in not being able to check their understandings with those whom they are seeking to understand. But faced with seemingly paradoxical statements, beliefs, or practices, we must and can avoid the apparent dilemma (either those statements must be rendered in or reduced to our terms or we must admit they are strictly incomprehensible) by insisting on the revisability of our existing conceptual framework, especially in relation to such key terms as personhood, agency, causation, and nature. Instead of insisting on the dichotomy of literal and metaphorical, we should allow that any term may exhibit what is here called semantic stretch. Moreover, if we accept that the phenomena described are multidimensional, then the goal of a single definitive translation is a mirage. The open-endedness of translation is no threat to mutual intelligibility, but its precondition.


As we have seen, social anthropology mobilizes translation at many levels, from ethnography to comparative analysis, to the formulation of general theories. The analysis of these different processes of translation of means of expression and context of communication can enable us to account for what both cognitivists and ontologists do not see: the essential plurality both of mental operations and of “ontologies” which always exist within a culture, as well as in different cultures. From this perspective, the foundations of social anthropology (and, more specifically, the ground for comparison between cultures) are no more to be found in “a” universal cognitive endowment, which would exist independently from any cultural phenomenon. Nor should we look for the foundations of our discipline in a number of ontological “modes of inference,” which would define the essence of a group of cultures, separated from the others. Our proposal is that, in order to understand “cultures” (and the kind of mental operations that the representation of cultural knowledge imply), we should focus not only on “differences,” but also on the constant work of translation of languages, nonlinguistic codes, contexts of communication, and different traditions, which constitutes the field of “cultural knowledge,” both within a single tradition and in different societies.

The analysis of these processes can provide for a new way to define translation, not only as a key technique for understanding ethnography, but also as a general epistemological principle. Since Boas, Sapir, and Whorf, anthropologists have defended the idea that every language elaborates the world in its own way. In this perspective, translation has been considered, at best, as an artificial and difficult process, a way to struggle against the constitutive differences that distinguish each language from others. In this way, however, a general and important fact has passed unnoticed: every language and every culture are not only different from each other; they are also translatable into each other. No untranslatable language, or culture, has ever existed. This quality of being translatable is inherent in all forms of human communication, as well as in the generation of cultural differences.

The recognition of the universality of translation as a principle can provide for the basis of a new way to look at cultural cognition, which would no longer be founded on an ideal (postulated) unity of the human mind, but rather on the empirical study of the cognitive processes involved in the various forms of translation of languages, means of expression, and contexts of communication of cultural phenomena. In this way, we could pass from a conception of cognition founded on a sort of universal cognitive grammar of human culture (a kind of logical form, postulated in a Platonic-Chomskyan perspective, which prevails today in the field of social cognition) to a Wittgenstein-inspired (1958) universality of cognition, conceived as an epistemological principle of translatability of language games, nonlinguistic codes, contexts of communication, and different ontologies. The concept of ontology would no longer refer to “conceptions of the world” linked to different languages, but to a plural and unsystematic way of constantly activating different forms of thought.

In this new perspective, social anthropology would be defined not only as the study of cultural differences, but also and simultaneously as a science of translation: the study of the empirical processes and theoretical principles of cultural translation.


This collective volume originates from the symposium on “Cognition and Cultural Translation” held at the Fyssen Foundation, Paris, March 20–21, 2014. We want to express our gratitude to the Foundation for its help and support.


Asad, Talal. 2010. “The concept of cultural translation in British anthropology.” In Writing culture: The poetics and politics of ethnography, edited by James Clifford and George E. Marcus, 141–64. Berkeley: University of California Press.

Austin, J. L. 1975. How to do things with words. Oxford: Oxford University Press.

Baumard, Nicolas, Pascal Boyer, and Dan Sperber. 2010. “Evolution of fairness: Cultural variability.” Science 329 (5990): 388.

Benjamin, Walter. (1923) 2004. “The task of the translator.” Translated by Harry Zohn. In The translation studies reader, edited by Lawrence Venuti, 75–85. Second edition. London: Routledge.

———. 1974. Problèmes de linguistique générale, Vol. 2. Paris: Gallimard.

Benveniste, Émile. 2012. Dernières lecons, Collège de France 1968–1969. Edited by Jean-Claude Coquet and Irène Fenoglio. Paris: EHESS/Gallimard/Seuil.

Berman, Antoine. 2008. L’âge de la traduction: “La tâche du traducteur” de Walter Benjamin, un commentaire. Vincennes: Presse Universitaires de Vincennes.

Bloch, Maurice E. F. 1998. How we think they think: Anthropological approaches to cognition, memory, and literacy. Oxford: Westview Press.

Boas, Franz. 1989. A Franz Boas reader: The shaping of American anthropology, 1883–1911. Edited by George W. Stocking, Jr. Chicago: University of Chicago Press.

Conklin, Harold. 1968. “Ethnography.” International Encyclopedia of the Social Sciences 5: 115–208.

Descola, Philippe. (2005) 2013. Beyond nature and culture. Translated by Janet Lloyd. Chicago: University of Chicago Press.

Enfield, Nicholas J., and Stephen C. Levinson. 2006. Roots of human sociality. New York: Berg.

Enfield, Nicholas J., and Jack Sidnell. 2012. “Language diversity and social action.” Current Anthropology 53 (3): 302–33.

Evans-Pritchard, E. E. 1940. The Nuer. Oxford: Clarendon Press.

Gentner, Dedre, and Susan Goldin-Meadow, eds. 2003. Language in mind: Advances in the study of language and thought. Cambridge, MA: MIT Press.

Grice, Paul. 1989. Studies in the way of words. Cambridge, MA: Harvard University Press.

Hallen, Barry, and J. Olubi Sodipo. 1997. Knowledge, belief and witchcraft: Analytic experiments in African ethnography. Stanford: Stanford University Press.

Hanks, William F. 2006. “Conviction and common ground in a ritual event.” In Roots of human sociality: Cognition, culture and interaction, edited by Nicholas J. Enfield and Stephen C. Levinson, 299–328. Oxford: Berg.

———. 2010. Converting words: Maya in the age of the cross. Berkeley: University of California Press.

Hanks, William F. Forthcoming. “Counterparts: Co-presence and ritual intersubjectivity.” Special issue “Intersubjectivity across cultures,” edited by Eve Danziger and Alan Rumsey, Language and Communication.

Hill, Jane H., and Bruce Mannheim. 1992. “Language and world view.” Annual Review of Anthropology 21: 381–406.

Jakobson, Roman. 1959. “On linguistic aspects of translation.” On Translation 3: 30–39.

Keesing, Roger. 1985. “Conventional metaphors and anthropological metaphysics: The problematic of cultural translation.” Journal of Anthropological Research 41 (2): 201–17.

Kuhn, Thomas S. 2000. The road since Structure. Edited by James Conant and John Haugeland. Chicago: University of Chicago Press.

———. 2012. The structure of scientific revolutions. Chicago: University of Chicago Press.

Kuhn, Thomas S., Aristides Baltas, Kostas Gavroglu, and Vassiliki Kindi. 2000. “A discussion with Thomas S. Kuhn.” In The road since Structure. Edited by James Conant and John Haugeland, 255–323. Chicago: University of Chicago Press.

Leavitt, John. 2010. Linguistic relativities: Language diversity and modern thought. Cambridge: Cambridge University Press.

Levinson, Stephen C. 2003. “Language and mind: Let’s get the issues straight!” In Language and mind: Advances in the study of language and thought, edited by Dedre Gentner and Susan Goldin-Meadow, 25–46. Cambridge, MA: MIT Press.

Lucy, John. A. 1992. Language diversity and thought: A reformulation of the linguistic relativity hypothesis. Cambridge: Cambridge University Press.

Peirce, Charles Sanders. 1955. Philosophical writings of Peirce. Edited by Justus Buchler. New York: Courier Dover Publications

Pinker, Steven. 1994. The language instinct: The new science of language and mind. Harmondsworth: Penguin.

Pym, Anthony. 2010. Exploring translation theories. New York: Routledge.

Quine, Willard Van Orman. 1960. Word and object. Cambridge, MA: MIT Press.

Rubel, Paula G., and Abraham, eds. 2003. Translating cultures: Perspectives on translation and anthropology. Oxford: Berg.

Sammons, Kay, and Joel Sherzer, eds. 2000. Translating Native American verbal art. Washington, DC: Smithsonian Institution Press.

Sapir, Edward. 1985. Selected writings of Edward Sapir in language, culture and personality. Berkeley: University of California Press.

Saussure, Ferdinand de. (1916) 2006. Course in general linguistics. Translated by Roy Harris. Chicago: Open Court.

Severi, Carlo. 2002. “Memory, reflexivity and belief : Reflections on the ritual use of language.” Social Anthropology 10 (1): 23–40.

———. 2004. “Capturing imagination: A cognitive approach to cultural complexity.” Journal of the Royal Anthropological Institute (N.S.) 10 (4): 815–38.

———. 2012. “The arts of memory : Anthropology of a mental artifact.” HAU: Journal of Ethnographic Theory 2 (2): 451–85.

Silverstein, Michael. 2003. Translation, transduction, transformation: Skating “glossando” on thin semiotic ice. In Translating cultures: Perspectives on translation and anthropology, edited by Paula G. Rubel and Abraham Rosman, 75–109. Oxford: Berg.

Sperber, Dan. 1996. Explaining culture: A naturalistic approach. Oxford: Blackwell.

Spelke, Elisabeth., Karen Breinlinger, Janet Macomber, and Kristen. Jacobson. 1992. “Origins of knowledge.” Psychological Review 99 (4): 605–32.

Viveiros de Castro, Eduardo. 1998. “Cosmological deixis and Amerindian perspectivism.” Journal of the Royal Anthropological Institute (N.S.) 4 (3): 469–88.

———. 2004. “Perspectival anthropology and the method of controlled equivocation.” Tipiti: Journal of the Society for the Anthropology of Lowland South America 2 (1): 1–22.

Whorf, Benjamin Lee. 1956. Language, thought, and reality: Selected writings, Vol. 5. Edited by John B. Carroll. Cambridge, MA: MIT Press.

Wittgenstein, Ludwig. 1958. Philosophical investigations. Translated by G. E. M. Anscombe. Oxford: Blackwell.

Traduire les mondes: L’espace épistémologique de la traduction

Résumé : La traduction a joué un rôle à la fois crucial et ambigu dans l’histoire de la linguistique et de l’anthropologie sociale. Au moins depuis Saussure et Boas, on a insisté sur le fait que les différences entre les langues (et les cultures) rendent la traduction idéale extrêmement difficile. Dans le même esprit, Quine a pu soutenir que la traduction, si on la considère en termes abstraits, ne peut être que vague, indéterminée, et, en dernière analyse, impossible. Nous prenons ici un point de vue différent. En nous référant aux travaux de Thomas Kuhn, nous soutenons au contraire que la confrontation constante entre des paradigmes théoriquement incommensurables (mais, de fait, constamment adaptés, modifiés et traduits) peut devenir un nouveau terrain d’enquête pour l’anthropologie. Dans cette perspective, les défis posés par la pratique de la traduction deviennent un espace épistémologique nouveau où on peut repenser à la fois les techniques d’analyse et les principes d’ articulation entre Linguistique et Anthropologie. L’anthropologie sociale, quant à elle, se définirait non plus comme l’étude des différences culturelles, mais aussi et simultanément comme l’analyse des processus empiriques et des principes de la traduction culturelle.

William F. HANKS received the Joint Ph.D. in Linguistics and Anthropology from the University of Chicago in 1983 and currently teaches in the Department of Anthropology, University of California, Berkeley, where he is also Director of Social Science Matrix, a cross-disciplinary research institute. He investigates the relation between grammar and communicative practices, colonial history, and contemporary shamanism among the Maya of Yucatán, Mexico.

William F. Hanks
Berkeley Distinguished Chair in Linguistic Anthropology
Professor of Anthropology
Affiliated Professor of Linguistics
Department of Anthropology
Kroeber Hall 232
University of California
Berkeley 94720

Carlo SEVERI is Professor (Directeur d’études) at the École des Hautes études en Sciences Sociales and Director of Research (Directeur de recherche) at CNRS. A member of the Laboratoire d’Anthropologie Sociale of the Collège de France since 1985, he has been a Getty Scholar at the Getty Institute for the History of Art and the Humanities in Los Angeles (1994–95), a Fellow of the Wissenschaftskolleg in Berlin (2002–2003), and a Visiting Fellow at Cambridge (1990, 2012). He is the author of La memoria rituale (La Nuova Italia, 1993), Naven or the other self (with Michael Houseman, Brill, 1998; French edition: CNRS Éditions, 1994), and The chimera principle (HAU Books, forthcoming; French edition: Rue d’Ulm-Musée du Quai Branly, 2007).

Carlo Severi
Laboratoire d’Anthropologie Sociale
Collège de France
52 Rue du Cardinal Lemoine
75005 Paris