A bottleneck in the plenum

This work is licensed under the Creative Commons | © Nicholas Harkness. ISSN 2049-1115 (Online). DOI: http://dx.doi.org/10.14318/hau6.1.025


A bottleneck in the plenum

Nicholas HARKNESS, Harvard University

Comment on Keane, Webb. 2016. Ethical life: Its natural and social histories. Princeton, NJ: Princeton University Press.

Webb Keane’s Ethical life (2016) begins with a first principle of ethnomethodology: If there is order in the plenum of the social, then it is to be found in human interaction. Keane’s masterful synthesis of psychology, philosophy, and anthropology draws upon this insight to argue that ethical life is afforded by interaction. Human beings everywhere not only engage in interaction but also reflect upon and judge interaction. They produce ethical knowledge through interaction. Human interaction is both ethicizing and ethicized. Because not all readers of Ethical life may be equally familiar with certain embedded assumptions of Keane’s semiotic framework, it may be useful to bring these assumptions to the surface, and briefly to essay their consequences for his argument.

First: Does interaction as affordance also imply interaction as limit? On the one hand, the multiple semiotic modalities of interaction afford first-, second-, and third-person perspectives, which appear predenotationally in children’s “proto-conversations” and then denotationally as person deictics (e.g., I and you) in speech (Rumsey 2003). On the other hand, the complex mediation of interaction by speech suggests that the same semiotic processes that afford ethical order might also constrain it. From the assertion that the natural and social histories of ethical life are to be found in the semiotics of interaction follows the question as to whether ethical life more generally is investigable in any coherent way without a serious and sophisticated consideration of semiosis and specifically language. And thus a central provocation of Ethical life: Can we take the foundational problem of language in culture—the semiotic processes through which social groups form and transform in relation to signification and value—to be avoidable or unavoidable in any anthropological analysis of ethics? In no way does this question suggest [464]that ethical life is reducible to language. Rather, Keane’s thesis seems to point to language as a formidable bottleneck through which any theory of socio-cultural “meaning” (ethical or otherwise) must eventually pass. How does this work?

As Keane notes, interaction is prior to denotation, both logically and empirically. And he claims that this is important for theorizing ethics. Linguistic anthropologists and others have recognized for some time the fact that, in semiotic terms, pragmatics, as the domain of indexicality, is prior to and encompasses denotation, as the domain of the semantico-referential or “symbolic” function of language. (The poly-functional concept of pragmatics here should not be confused with the crassly mono-functional pragmatics of goal-directed behavior suggested by Malinowski.) Both of these concepts—the broadly pragmatic and the narrowly denotational—allow us to explain how certain elements of speech can refer reciprocally to changing interactional roles in the form of semantically “empty” first- and second-person participant deictics (I and you or their equivalents in any language), a class of what Roman Jakobson ([1957] 1984), following Otto Jespersen, called “shifters.” More generally, by viewing pragmatics as prior to and encompassing denotation, we also are in a position to explain speech as social action, e.g., in John Austin’s (1962) sense of the “performative” capabilities of speech. Recall that Austin began with a list of English words and then found out that under certain grammatical and contextual conditions, some of these words appeared to do or “perform” exactly what they described in the very moment of describing it. The Austinian approach deals with performative action as a “perlocutionary” product of (and thus logically subsequent to) denotation under certain conditions. The Austinian approach is thus forced to explain (unsatisfactorily) how and why some words do things and others don’t. After all, in Lecture XII of How to do things with words, Austin revealed that he had gone through the dictionary to run the test of the first-person-singular present-indicative-active form on English verbs listed there to come up with some general classes of speech acts. There is a better way. By viewing the pragmatics of interaction as logically and empirically prior to (and affording) denotation as Keane has suggested, we can more easily and clearly explain Austin’s “discovery” by showing that words and other elements of language indexically always “do” and “perform” social things in cultural context; and moreover that in certain languages, under certain conditions, the ongoing social action (which might remain implicit, i.e., “primary,” rather than “explicit”) is brought to awareness by coinciding in utterance form with a description or “typification” (Agha 2007) of that action. The basic point here is that an approach that situates denotation within a broader semiotic pragmatics of interaction is able coherently and empirically to explain the effectiveness of speech (in whatever functional mode) as action. And as Keane has argued, this is precisely why interaction is a necessary starting point for the study of ethics.

As Keane explains, the openness of the reciprocal first- and second-person perspectives afforded by interaction can be anchored by third-person perspectives in the form of “specific ideas expressed in words such as the English condescension, the Sumbanese dewa, El Barrio’s respect, the Chewong punén, and the Inuit ihumaquqtuuq” (2016: 243), which “convey their meanings into an indefinite number of possible contexts, regardless of who ‘I’ or ‘you’ may be” (244). The introduction of what Keane calls the “third-person perspective” is significant. As Émile Benveniste ([1956] 1971) pointed out in “The nature of pronouns,” there is a substantial [465]functional shift from reciprocal first- and second-person pronouns to third-person pronouns. Unlike first- and second-person pronouns, third-person pronouns do not automatically refer to participant roles in a speech event, such as “speaker” or “addressee” (they can do, e.g., as tropic distantiation found in certain honorific forms, but they don’t have to). Rather, they are “endophoric” substitutes for persons referred to in speech or “exophoric” substitutes for namable persons (or entities) outside of the reciprocal interactional space of “Is” and “Yous.” So even though, for speakers of Standard Average European languages, “the regularity of the formal structure . . . produces the impression of three coordinated persons” (Benveniste [1956] 1971: 221), there actually is a profound semiotic break or rupture between reciprocal first- and second-person perspectives and third-person perspectives.

This is significant for Keane’s argument for two central reasons. First, the profound semiotic break between first- and second-person perspectives on the one hand, and third-person perspectives on the other, allows us to see how ethical order is produced by invoking third-person perspectives through the semiotic resources of what Mikhail Bakhtin (1981), using a discursive lens, called “voicing” and “heteroglossia,” and what Erving Goffman (1979), using an interactional lens, called “footing,” the “participant framework,” and the “production format.” These perspectives can be invoked vaguely and subtly or they can manifest clearly as enregistered characterological types or authorized figures of personhood. As captured in Keane’s reading of Jane Hill’s (1995) paper, “The voices of Don Gabriel,” third-person perspectives break through into interaction all of the time via speech that can be authorizing (e.g., as recognizable stances on condescension, dewa, respect, punén, or ihumaquqtuuq) or undermining (through various kinds of “clashes” or “dysfluency”). These pragmatic processes of authorizing and undermining are made possible by socio-cultural frameworks of value that often lie beyond speakers’ metapragmatic awareness.

Second, this profound semiotic break allows us to see how explicit “morality systems,” qua organized conceptual frameworks for evaluating human behavior, are formulated by linguistic elements (such as condescension, dewa, respect, punén, or ihumaquqtuuq) that can be situated along what Michael Silverstein (1987) has called a “referential hierarchy.” This hierarchy—really a dimensionalized space of referential possibilities—begins with interaction, namely the indexical functioning of reference for first- and second-person deictics. It then moves from the inherently interactional to third-person endophoric and exophoric deictics, to proper names (which, as Kripke, Putnam, and others have pointed out, are contingent upon the knowledge and status of the speaker and audience of a sociohistorical “causal” chain of reference from a “baptismal” zero-point), to culturally stipulated socio-centric kinand status-terms, to more general “things in the world” that sociolinguistic communities refer to by relying on prototypes and principles of stereotypic “likeness,” and finally to the class of semantico-syntactic “formal” noun phrases that depend most centrally on grammatical structures and processes of lexical derivation (e.g., English nouns ending with formatives like -ness, -tion, -ance, etc.). Along this hierarchy, we can observe a continuous, categorially (not categorically) organized shift from a relatively “microcontextual” situation in which the referring expression and its characterization are aligned in the same signal and coincide (e.g., I and you), across the semiotic break from interaction to third-person perspectives, at [466]the most extreme boundary of which lies a relatively “macrocontextual” frame in which the referring expression is associated with its conceptualization in so far as its structural form implies a larger structure of grammar (e.g., condescend, condescending, condescension, etc.).

I realize that this discussion must sound very technical and complicated. There is a good reason for this. In the communicative and conceptual space between the ethical affordances of interaction and the verbalized frames of explicit morality systems, there is the vast and complex problem of language in culture. When we accept the magnitude of this problem, as I think Keane has implied that we should, then we have to deal with the span of indexical functioning in reference and predication. This span goes from relatively semantically empty referential forms requiring saturated contextualization in order to refer to anything at all, to referential forms corresponding to culturally saturated concepts that can be invoked morally and ethically to evaluate situated human behavior while being construed ideologically as universal, essential, or at least stable across the widest imaginable scope of socio-historical contextualization. This is the linguistic bottleneck foregrounded by Ethical life.

Ethical order emerges clearly in speech when the two kinds of third-person perspectives discussed here coincide: figures of personhood that suffuse the “Is” and “Yous” of interaction, and lexicalized cultural concepts that distinguish these figures according to their actions. Simply put, this happens when “principled” people talk about “principles.” A member of a holy order speaks of holiness. A respected youth speaks of respect. A pious person speaks of piety. And so on. In such instances—which form a great deal of our ethnographic data, by the way—the microcontext of indexical underdeterminacy in interaction and the macrocontext of explicitly conceptualized morality systems meet in a ritual moment of pragmatic performativity that paints a metapragmatic picture of ethical order. Drawing from the vastly complex plenum of social life, semiotic values condense into ethical values, and culture squeezes through the bottleneck of language.


Agha, Asif. 2007. Language and social relations. Cambridge: Cambridge University Press.

Austin, John. 1962. How to do things with words. Cambridge, MA: Harvard University Press.

Bakhtin, Mikhail M. 1981. The dialogic imagination: Four essays. Austin, TX: University of Austin Press.

Benveniste, Émile. (1956) 1971. “The Nature of Pronouns.” In Problems in general linguistics, 217–22. Coral Gables, FL: University of Miami Press.

Goffman, Erving. 1979. “Footing.” Semiotica 25 (1–2): 1–29.

Hill, Jane. 1995. “The voices of Don Gabriel: Responsibility and self in a modern Mexicano narrative.” In The dialogic emergence of culture, edited by Dennis Tedlock and Bruce Mannheim, 97–147. Urbana: University of Illinois Press.[467]

Jakobson, Roman. (1957) 1984. “Shifters, verbal categories, and the Russian verb.” In Russian and Slavic grammar: Studies, 1931–1981, edited by L. Waugh, 386–92. Berlin: Mouton.

Keane, Webb. 2016. Ethical life: Its natural and social histories. Princeton, NJ: Princeton University Press.

Rumsey, Alan. 2003. “Language, Desire, and the Ontogenesis of Intersubjectivity.” Language & Communication 23 (2): 169–87.

Silverstein, Michael. 1987. “Cognitive implications of a referential hierarchy.” In Social and functional approaches to language and thought, edited by Maya Hickman, 125–64. Orlando: Academic Press.


Nicholas Harkness
Department of Anthropology
Harvard University
21 Divinity Avenue
Cambridge, MA 02138