Abstract
‘Ordinary ethics’ suggests that everyday discursive interaction – interaction mediated by actual language use – has tacit ethical dimensions. This line of inquiry is productive for the anthropology of ethics and has the potential to reframe long-standing language-based research on everything from conversational turn-taking to politeness displays, but what does it mean to speak of discursive practice as a locus for ethical life? To what extent is the ethical inscribed in the ground-rules of interaction, or conditioned from below (e.g. biologically-based cooperative predispositions) or from without (e.g. culturally-institutionalized moralities)? The presumption that ethics is immanent in practice continues to distract from the problem of how to narrate, and theorize, the entanglements of discourse and ethics.
In Michael Lambek’s (2010) provocative volume Ordinary Ethics, ‘ordinary’ means an ethics that is ‘relatively tacit, grounded in agreement rather than rule, in practice rather than knowledge or belief, and happening without calling undue attention to itself’ (Lambek 2010: 2). Echoing the ‘ordinary language’ philosophy of old (see also Das 2012), whose contributors included Austin and the later Wittgenstein, this work also inspires a turn toward actual language use, toward discourse. Austin, as Jack Sidnell (2010) reminds us, tried to get moral philosophers to see that even mundane acts like excuses in conversation could serve as a window onto the ethical. Like Lambek, Veena Das (2012: 134) has tried to effect a turn from ‘thinking of the ethical as made up of judgments we arrive at when we stand away from our ordinary practices to that of thinking of the ethical as a dimension of everyday life’. 1
Ordinary ethics is an important turn. It gets us to stop privileging overt rational reflection and ‘choice’ and encourages us to look beneath highly visible institutionalized discourses, such as codes of bioethics and religious conduct. It invites us take interaction more seriously, since, after all, ‘one does not develop morality all by oneself’ (Keane 2010: 74; n.d.). In doing both, it spotlights practice and stimulates commerce with fields like linguistic anthropology and sociolinguistics that specialize in discursive interaction but mostly haven’t given ethics its due.
I want to contribute to this turn by registering a concern about the curious ease with which ordinary ethics locates ethics in practice and in discursive practice in particular. What could it mean to say that ethics is, as Lambek (2010: 1), Das, and their colleagues aver, ‘intrinsic to speech and action’ (emphasis mine)? I want to argue that whatever else the ‘ordinary’ in ordinary ethics might mean, it cannot mean a fluent, effortless immanence. Ordinary ethics is not ordinary in this sense of being unproblematically there, findable, as if locating it were as easy as turning over a stone.
The performative contingency of ordinary ethics – the fact that ethical events require communicative labor to happen and are hence precarious achievements – complicates the very notion that the ethical is intrinsic to practice. Consider Das, who narrates with eloquence how ‘small acts’ (Das 2012: 139) can do big things, from care to harm, without ever announcing what they do (Das 2007, 2010a, 2010b). Drawing on her work in low-income neighborhoods in Delhi, she recounts, for instance, how she learned to appreciate the ‘delicacy of maintaining regard for others through the minutest of gestures,’ as in cases where ‘women would refrain from sweeping the floor right after a guest had left because that might suggest that “We think that guest is just trash” ’ (Das 2012: 135). She contrasts such seemingly small, tacit acts with ‘dramatic enactments’ of ethical value, the latter including everything from the multi-party histrionics of public ritual to legal-juridical rulings on right and wrong. Das relies on a rhetoric of immanence in her appeal for ordinary ethics, though she does complicate this immanence in limited ways. She sensitizes us to effort – to degrees of moral ‘striving’ (Das 2010b, 2012) that occur within the everyday and are pronounced under conditions of abject poverty and violence. This exposes the fragility of ethical events, but once we scrutinize real-time ethical events with recordings and transcripts – as researchers on interaction do – we can see more vividly just how precarious ethical events are. I want to dwell on this precariousness and argue that the study of ordinary ethics could do more to illuminate the labor and methods through which actors strain to make the ethical not just effective but intersubjectively evident.
To the extent that ethical activity is kindled through discursive interaction rather than being an ever-present quality of it, we must remain alive to performative felicity, and failure. A virtue of the small but growing literature on morality in interaction, which I sample below, is precisely its attentiveness to the fraught discursive ‘work’ (Drew 1998) of moralization. While this literature has its limitations for anthropology, it details discursive methods for making ethics matter in interaction. In an essay on complaints about misconduct aired during phone conversations, conversation analyst Paul Drew (1998: 322) is at pains to show that no behavior is ‘self-evidently, intrinsically, or inherently morally reprehensible’ and that ‘moral character is constituted through the descriptive practices associated with complaining’. More deeply, Drew reminds us that ‘[o]rdinary or mundane conversation is not of course pervasively about morality’ (1998: 296), that interactants have to do things to kindle moralization. Indeed, the ethical sometimes goes not just unremarked but unfelt. Making ethics matter in discursive interaction – making it intersubjectively recognizable and pragmatically consequential – can sometimes seem as impressive a feat as materializing an incorporeal agent, like a spirit or god whose presence depends on an awful lot of semiotic and interactional labor. Ordinary ethics may not be so ordinary after all. Its presence can’t be taken for granted.
There is a second sense in which immanence misleads. Just as the lure of immanence distracts us from noticing how actors invoke and register the ethical in practice, so, too, does it exaggerate the immediacy and localizability of ethical events. If everyday practice is a locus for ethical life – as the appeal to the ordinary often suggests – then we wonder: what kind of ethics appears in discursive interaction – interaction mediated by concrete instances of language use – and what relation, if any, might this have to the full-blown moral discourses and institutionalized doctrines that ordinary ethics says it brackets? Ethics may seem to lurk in everything from conversational turn-taking to verbal politeness to embodied acts of considerateness – refilling a cup, holding open a door – but how do we as analysts, to say nothing of the social actors we study through copious recordings and faithful transcripts, ‘locate’ (Lambek 2010: 39) and specify the ethical with respect to the everyday behavior in which it is enacted? Is it adventitious or inherent, exogenous or endogenous, parallel or orthogonal, rhizomic, entangled, or enmeshed? Might ethics even be hewn into the foundation of interaction, where it supplies the tacit, taken-for-granted ground rules of the ‘interaction order’ (Goffman 1983)?
Declaring ethics to be immanent in discursive practice waves off such questions, even if its stress on immanence ironically makes them that much more pressing, as we shall see. But let us first recall how familiar this restless questioning over ethics’ whereabouts is. Ethics has long seemed the fugitive, for many chase it to its presumptive source, winding their way to ‘cultural,’ ‘social’ or ‘biological’ provenance. (One need only glance at the entrancing oscillations of the nature-nurture see-saw, where some, for example, pursue morality’s evolutionary roots in nonhuman primate cognition while others declare the roots shallow, entwined only in the soft contingencies of history or culture.) The very search for ethics’ source betrays disciplinary commitments to objects of knowledge to which deference is still owed, and duly paid, thereby reassuring those in these fields that they still have something solid and perhaps even epistemologically irreducible under their feet. Others claim to have given up the chase and rest content with knowledge of what is now something of an anthropological commonplace: that ethics isn’t a separate ‘domain’ but is rather so entangled in other practices and domains of life that it makes little sense to try to stake out its boundaries or trace its roots. Advocates of ordinary ethics would concur, to the extent that they view ethics as an immanent dimension of everyday life. I want to argue that this does not and cannot end the chase, however, and that we need to examine with more care how the communicative labor used to invoke ethical events does not so much expose a latent dimension of practice as much as cobble together or ‘assemble’ (cf. Latour 2005) the ethical from diverse and often far-flung materials, creating entanglements from which ‘ethics’ can neither be wrested nor said to be anywhere in particular.
In this essay I explore the consequences for ordinary ethics of taking seriously this communicative labor. I sample discourse-centered literature, ranging from work on politeness to moral indirectness to ritualization, in order to suggest how the assumption of ethics’ immanence in discursive practice has come at the expense of an appreciation of ethical contingency and entanglement. The literature cited spans fields and is by necessity a very small sampling. I highlight a few tendencies that are instructive for what they tell us about the state of the problem as a whole and for where they point in terms of future research in ordinary ethics.
1. Is morality immanent ‘in’ interaction?
The idea that ethics is intrinsic to discursive practice would seem well supported by research on interaction. In Ordinary Ethics, conversation analyst Jack Sidnell (2010) pages back to Austin’s (1961) prescient essay ‘A Plea for Excuses’ which, despite its introspective methods and fixation on words rather than stretches of discourse, usefully suggested that moral philosophy turn for its evidence to everyday talk. Everyday conversational practices seem to presuppose ethical principles. As conversation analysis has demonstrated, rarely do we brusquely decline an ‘offer’ or ‘invitation,’ for instance; instead, we temper our refusals with such behavior as palliatives (‘that’s awfully kind of you’), speech delays (filled or unfilled pauses) and dysfluencies (cut-off speech), and even ‘accounts’ where we state reasons – excuses – why we can’t accept. 2 Such conspicuous accounting for our non-acceptance, together with the delicate manner in which we exhibit discomfort and express appreciation for our interlocutor, seem to rest on a sense of moral obligation, on what we owe others.
Or consider the bare fact that we can get someone to do something for us just by expressing how we feel (e.g. a ‘response cry’ [Goffman 1978] like brrr) or by assessing some external state of affairs (e.g. ‘it’s chilly in here’.). Linguistics research in pragmatics marveled at first at the roundabout manner in which these expressions seemed to convey their meaning, treating them as ‘indirect’ speech acts (Searle 1975), though we may also wonder why a speaker’s alleged discomfort should cue assistance at all. Does some principle of altruism, cultural or natural, mediate the response to such expressions, and, if so, what additional data-points and evidence must one include in order to illuminate this mediation?
Teasing out such principles from discourse seems no mean feat, but, then again, perhaps we don’t need to. Perhaps interaction has its ‘own’ ethics. Erving Goffman (1959, 1967) suggested this when he spoke of interaction as a moral order. He wrote with elegance and unnerving acuity about the thousand little obligations humans tacitly uphold when they interact. The labor of ‘mutual monitoring’ (Goffman 1966) in which speakers in their ‘eye-to-eye ecological huddles’ (Goffman 1966: 95) busily scan their interlocutor’s face and body and speech for signs of attention and response, and the way hearers, for their part, offer backchannels such as head nods and minimal responses like ‘uh huh,’ to signal their alleged attunement, suggests a deeply collaborative venture. At root would seem to be a mutual commitment to certain ground rules of interaction, rules designed, as Goffman saw it, for the protection of something sacred: the self (Goffman 1959, 1967). 3 (For Garfinkel, with whom Goffman is often compared, social interaction is more about the co-construction of mutual intelligibility than the production of and care for self, but they converge on their sensitivity to the reciprocal and cooperative nature of interaction [see, for example, Rawls 2010].)
To what extent does this moral order derive its morality from ‘outside’ interaction? 4 Some noted that Goffman saw interaction’s obligations more as ‘etiquette’ than ‘ethics’ proper, for he was preoccupied with the way interactants strain to preserve a semblance of propriety in ‘ceremonial’ fashion, irrespective of whatever ‘substantive’ moral sentiments may – or may not – drive this labor (Bovone 1993). In The Presentation of Self in Everyday Life (1959), he deemed interactants ‘merchants of morality,’ for, ‘qua performers, individuals are concerned not with the moral issue of realizing these [moral] standards, but with the amoral issue of engineering a convincing impression that these standards are being realized’ (p. 251). Ann Rawls (1987) observed that Goffman was neither clear nor consistent here. At times morality seemed to mean an obligation to institutionally defined roles, which implied that one’s obligation ultimately was to social structure and that one’s role inhabitance dutifully served this order, reproducing it in the process. At other times – arguably more often – these obligations seemed like strictly ‘involvement obligations’ (Rawls 1987) to a sui generis ‘interaction order’ (Goffman 1983). That is, the obligations were to interaction itself and to the joint labor of face-work that occurs within its fold. The joint labor subserves the mission of protecting an ever vulnerable sacred self, but all this labor is specific to a spatiotemporally delimited, diminutively scaled ‘order’.
In this view – and it is one that Rawls (2010, 2011) herself maintains – the interaction order is its own ‘constitutive’ moral order, and one cannot assume that this morality reinforces or even has any simple relationship to institutional moralities outside that of the face-to-face. At one extreme (again, not an extreme that Goffman himself always occupied), then, the ethics of interaction is specific to interaction. Interaction is a ‘constitutive’ social order (Rawls 2010) whose ethics does not depend on institutionalized moralities, even though it may articulate with those moralities in varying degrees and diverse ways. 5
‘Interaction’ is imagined to be hived off, its morality endogenous. Fast forward to a recent volume on The Morality of Knowledge in Conversation (Stivers et al. 2011) and we can see a similar sensibility toward morality in interaction – at least in terms of anchoring morality in the interaction order. The volume details how all sorts of unstated normativities affect the expression of knowledge in conversation, which, at their most sweeping, include heuristics – tacit Gricean-style rules-of-thumb – such as, Don’t tell people things they already know (‘speakers should not inform knowing recipients about some state of affairs’) and Don’t say things you don’t know (‘avoid making claims for which they have an insufficient degree of access’ (Stivers et al. 2011: 10). Stivers (2011) looks at English ‘of course’-replies to tag problematic questions. To the query, ‘So did you introduce her?’ for instance, an ‘of course’ reply questions the question’s ‘askability’ (Sacks 1987) – whether it should have even been posed; it tends to be used when ‘something morally problematic may be the case and when questioners have epistemic access to the answer either from interactional history or from general knowledge’ (Stivers 2011: 88).
Stivers has actually shown us more than just the manner in which ‘interactants hold each other accountable for the rights and responsibilities associated with epistemic access, primacy and responsibility’ (Stivers et al. 2011: 19), for is this not also a method used to call out floutings as floutings? Likewise, a chapter by Heinemann et al. explores how adverbs jo (Danish) and ju (Swedish) can, in response to certain questions, signal ‘that the questioner failed to take into account shared knowledge’ (Heinemann et al. 2011: 107). Quoting Heritage (1988: 182) on Garfinkel’s (1967) ethnomethodological breaching experiments, the authors remind us that ‘there is no quicker way, it appears, of provoking moral outrage than by not using background knowledge to make sense of other people’s actions’ (Heritage 1988: 182; cited in Heinemann et al. 2011: 107). This may be, but is the content of such moral outrage just as substantive as the ‘morality’ that was there before the breach was fingered for repair? Keane (2010: 69 et passim) raises a similar issue when he notes the need to distinguish acts of justification and their content from interactional events that incite such justification. These are distinct events to be pried apart analytically, even though the two may be entangled in specific ways (e.g. the first may be felt to ‘cause’ the second; the second may point to, cite, or typify the first). Let us therefore say that what these volume-contributors have pointed out are reflexive acts that result in moralization events in subsequent phases of interaction, ‘justification’ being but one event among many. 6
This is variation here, too. Gentle though reflexive engagements like ‘of course’-replies may be, they can be set on a continuum whose extremes include full-tilt moral ‘breakdowns’ (Zigon 2008) in which everything screeches to a halt and the reflexivity that erupts is acute and sustained. With video-recorded interactions of four- and five-year-old children, volume contributor Sidnell (2011) notes a couple of minor meltdowns in his chapter on the epistemics of make-believe. In individual play, children use baldly asserted ‘stipulations’ to transfigure materials-at-hand such as mundane plastic blocks into a ‘Rubik’s cube’ or wooden planks into a ‘swimming pool’. In joint play these stipulative transformations shouldn’t be settled by fiat. They should be ‘proposed,’ left ‘open to negotiation’ (Sidnell 2011: 144), and to this end kids may append a delicate ‘okay?’ tag-question or frame their proposal with a diplomatic ‘let’s pretend’. Which helps explain why little Sean cries foul – ‘No that not part of the ga::me’ – when playmate Andy summarily shoots their imagined horse without his consent: ‘An- and we shoo:t the horse. Ba::ng. He’s dead’ (Sidnell 2011: 144). In crying foul, remedial action may follow, such as apologies or even reasoning about how one is ‘supposed’ to play. Again, interactants sometimes reflexively direct attention to and address breaches in expected sequence organization (cf. Drew 1998; Zigon 2008: 141–2). In this case the kids sometimes end up citing rules of conduct.
Though Morality of Knowledge in Conversation does not make this distinction, its findings invite us (minimally) to distinguish between (a) ‘morality’ qua the tacit normativities of discursive interaction and (b) ‘moralization’ events that occur in the wake of reflexive engagements with this first-order normativity, engagements that include the targeting of breaches for repair. (I do not intend this as an all-encompassing distinction but as a reminder that distinct moments of semiosis are being collapsed here, a collapse that exaggerates ethics’ immanence by telescoping events of moralization into the behavior that occasioned it.)
A distinction between tacit, unproblematic normativity and reflexively problematized normativity is a familiar one, paralleling others in the literature. Zigon (2008, 2009), for instance, wants to distinguish ‘morality’ as enacted normativity from ‘ethics’ as moral questioning and choice, just as (mutatis mutandis) Robbins (2007, 2009) contrasts a Durkheimian ‘morality of reproduction’ with a ‘morality of choice’. As Zigon suggests, these distinctions correlate with two major approaches to the contemporary anthropology of morality: those who, in a neo-Aristotelian or Foucauldian vein, approach ethics as embodied in practice, and those who, in the spirit of Kant, see ethics as acts of standing apart from the world and reasoning about it, such as from particular to general, as the categorical imperative would have it. Apart from questions about the utility of these dichotomies – some say it’s impossible to sustain such a distinction between ethics and morality – what emerges from these exchanges is the recognition that reflexivity is a critical parameter of variation.
‘Sometimes we are in the midst of action,’ writes Keane (2010: 69) in his contribution to Ordinary Ethics, and ‘sometimes we seem to stand apart from it’. This ‘standing apart’ – this reflexivity – is indeed vital and is more heterogeneous than analytic captions like ‘choice’ let on and comes in many more degrees than a trope like moral ‘breakdown’ (Zigon 2008) allows. In the linguistic anthropology literature, ‘reflexivity’ tends to be used more expansively to mean ‘activities in which communicative signs are used to typify other perceivable signs’ (Agha 2007: 16). It needn’t mean ‘choice’ or require self-consciousness (see also Robbins 2009: 278), nor is it limited to what people explicitly say ‘about’ language, language use, or language users. Distinct, denotationally implicit varieties of reflexivity exist, which can often be more pragmatically important precisely because they are less easily reportable (Silverstein 1981, 1992; Agha 2007). Indeed, this variation can be sensed in the Morality of Knowledge volume itself, since its studies illustrate how responses to breaches vary a lot, from subtle reactions of unease (e.g. dysfluencies like cut-off speech, markedly long pauses) to conversational repairs in which an interlocutor calls attention to prior speech as problematic, to in-your-face indictments that spell out what’s wrong. All such responses involve reflexivity but vary along several dimensions (e.g. degree of denotational ‘explicitness,’ degree of conventionalization, range of semiotic modalities used, scope of participants involved and targeted, etc.). 7
In addition to appreciating the heterogeneity of reflexivity, ordinary ethics should also consider which kinds of violations tend to draw attention in the first place. The presumption of ethics’ immanence in interaction too often coincides with a diffuse and underspecified sense of the ethical. As Robbins (2007) recounts, citing Laidlaw’s (2002) influential essay, Durkheim ironically sounded the death knell for the study of morality when he made it an object of knowledge in sociology, for in wresting morality from moral philosophy he ended up reducing it to the social. In a sense Goffman (and Garfinkel) did to the morality of interaction what Durkheim did to the anthropology of morality: he ‘spread morality too thinly over society, making it everywhere present but almost invisible in its role in shaping social life’ (Robbins 2007: 294). This legacy haunts works like The Morality of Knowledge in Conversation. While admirably faithful to the specificities of interactional normativity, the volume never tells us what makes any of this normativity distinctly ‘moral’. Morality is left undifferentiated and co-extensive with rights and responsibilities of knowledge expression tout court. Under certain conditions but not others, don’t we expect interactants to appeal to and cite cultural discourses on morality more easily and extensively? Perhaps a controversial issue has been broached or moral quandary presented.
Consider those normativities felt to cluster into metapragmatic skills, such as, ‘how to listen,’ ‘how to share the floor,’ which are culturally elaborated, inculcated, and policed through regimes of etiquette and language socialization. In terms of listenership, say you find yourself a hearer as a speaker shares a self-narrative of woe. Distracted, you succumb to ‘misinvolvement’ Goffman (1957) and neglect to emit backchannels and the occasional response cries of empathy or surprise (‘jeeze,’ ‘oh my god’) expected at critical narrative junctures; and so on. Or consider an event in which symmetrical, ‘dialogical’ turn-taking – exchanging speaker and hearer roles with relative parity – matters a lot, as in some styles of political debate and conflict mediation. Would violations in turn-taking and floor-sharing here be just as likely to be targeted for repair as, say, the norms for bodily comportment such as how to manage one’s elbows? Clearly some things are more likely than others to be targets of repair, and figuring out which may allow us to infer how the moral is distributed in a given context.
The act of reflexively targeting for repair some bit of conduct is but one method to make moralization happen. A journal issue dedicated to ‘morality in discourse’ (Bergmann and Linell 1998) opened up a discussion of the discursive methods used to incite moralization. As co-editor Jörg Bergmann remarks in his introduction, these methods include the obvious – lexical resources, such as names for social types like drunkard – and the less obvious, including prosody and bodily comportment. It isn’t long before Bergmann (1998: 281) bursts open the proverbial flood gates: ‘An inconspicuous body twist, a slight rise of the eyebrows, a certain tone of voice will suffice to give a statement a moral tinge’. So widespread and motley are these signs that Bergmann reaches a conclusion that sounds both familiar and unsettling: ‘Obviously, morality is omnipresent in everyday life,’ and ‘it is so deeply intertwined with everyday discourse that the interlocutors hardly ever recognize their doings as moral business’ (1998: 281). There is a slip by which ethical omnipresence – the capacity to see ethics everywhere if one looks hard enough – becomes an ontological claim about ethical immanence in practice, but for now let me expand on Bergmann’s astute comment that most events featuring moralization do not call attention to themselves as ‘moral’ (see Drew 1998); they are not, in a word, denotationally ‘explicit’.
Literature in linguistic anthropology and related fields is replete with illustrations of this point, even if this literature does not always understand itself in these terms. Keane’s (2011) revisiting of Jane Hill’s (1997) essay on the ‘voices’ of Don Gabriel, for instance, reminds us well of the implicit work of moralization. When Don Gabriel tells the poignant story of his son’s death, he makes choices at every turn. His choices among linguistic variables (e.g. Mexicano versus Spanish, prosodic and stylistic contrasts) become ‘moral’ choices, because these variants help fashion morally weighted figures – images of kinds of people that populate the storyworld – and express the narrator’s stance toward them. All this occurs without overt moralizing.
In fact where better to appreciate the labor of moralization than in more extreme cases of purposeful moral ‘indirectness’ (Lempert 2012b)? 8 Consider for a moment the stealth and care by which moral criticism is quite often aired. So-called ‘weapons of the weak’ (Scott 1985, 1990), which range from foot-dragging to backstage griping, do not merely express resistance but also frequently invite others to infer the source and type of injustice behind the behavior. ‘Teasing’ and ‘shaming’ are familiar methods for shaping moral dispositions through evaluation, evaluation that is quite often purposely indirect (see, for example, Schieffelin 1987, 1990; Miller 1987; Goodwin 1990, 2006; Lo and Fung 2012; Lempert 2012a, 2012b; in a different vein, see Foucault 1979; Drinan 2001; Keenan 2004).
Two brief examples. Consider the routinized indirectness of ‘moral irony’ constructions in Sakapultek, a Mayan language spoken in highland Guatemala (Shoaps 2009). These constructions denote stances that are modalized with a nonfactual clitic + t (sometimes with the irrealis ni) and negatively evaluated with xa'. They translate roughly as ‘as if p,’ where p is not a proposition but a propositional stance: e.g. ‘as if being a witch didn’t matter’. Expressions like this summon an unnamed figure, a hypothetical person other than the current speaker, who is depicted as staking out a morally inappropriate stance – here, that being a witch doesn’t matter, when, of course, everyone knows (or ought to) that it does! Without lifting a finger, the Sakapultek stance-taker has seemingly managed to drag moral truths into plain view. All she has done is invite her interlocutor to recall the moral system the figure presupposes in order to decipher what stance she – the speaker – ‘actually’ upholds. One interlocutor starts the work of moralization while the other, through the labor of inference, is made to finish it.
Consider a more elaborate case, an entire speech genre featuring moral indirectness. In his classic essay on the Western Apache genre of ‘speaking with names,’ the late Keith Basso (1988) detailed how Apache speakers could pronounce a placename and people would fall silent as they visited the place in their mind’s eye. These toponyms are hitched to moral-didactic narratives, stories of what the ancestors did. In recalling the place, hearers are to recall the associated story and grasp the allegory. Through parallels between the story and event, placenames are credited with bringing there-and-then ancestral wisdom to bear on the here-and-now.
‘Indirectness’ in such cases is but a caption for a deliberately distributed participation structure, where one shifts from speaker to hearer the burden of figuring out that ethics is relevant and who this ethics is ‘for’. All this contingent and distributed performativity, with its hints and innuendo and inferences, makes it easy to grasp that ethics is not unproblematically immanent, that interactants work to invoke and infer the ethical.
2. Ethical entanglement, not immanence
Immanence implies not only the effortless presence of ethics in practice but also its localizability – the very sense that ethics can be traced to a singular, determinate source. Latour (2005) proposed a series of exercises that erode the sense of the spatiotemporal localizability of interaction and can be extended to the localizability of ethics in interaction. No interaction is, for example, ‘isotopic,’ because ‘what is acting at the same moment in any place is coming from many other places, many distant materials, and many faraway actors’. No interaction is ‘synchronic,’ because the pieces it comprises did not all begin at the same time (Latour 2005: 200–2). These exercises – and there are others – suggest that interaction has no intrinsic scale (see Lempert 2009, 2012c). Actors also do things that deliberately make the interaction order more porous, and nowhere is this more apparent than with acts of ‘interdiscursivity,’ where people link or liken one discourse with another (Agha and Wortham 2005). Moralization events, whether incited by repairs or induced by ‘indirect’ means or some other method, can bring into play ideals of virtue and non-virtue, figures of good and bad people, notions of right and wrong, that weren’t ‘there’ in the conversation until that moment and that are instead associated with spatiotemporally distinct domains and activities. Even when interlocutors pause and inspect what ‘just happened’ – a norm of conduct breached, a fraught ethical choice made, a virtue or non-virtue enacted – such apparent discoveries ferry in as much as they find in situ. When the kids in Sidnell’s (2011) study experience a breakdown in make-believe play and cite rules, are they not delicately calibrating here-and-now discursive events with authoritative discourse they’ve heard elsewhere? In these and other ways events of moralization routinely transgress the limits of the interaction order, encouraging us to peer past it.
Cultural routes of ‘politeness’ in interaction
If ethics is not immanent in the interaction order, than surely it must be found elsewhere, we feel. Let us entertain this instinct and do what most do, and that is to trace a course from interaction to some distant, constitutive ethics, be it ‘culture’ or ‘society’ above or ‘biology’ below. I will begin by following a few such routes that depart from the landmark of ‘politeness’ behavior in interaction, with some paths leading right back to the interaction order itself while others stretch far toward the firmament of ‘culture’. The studies mentioned are only a handful but illustrate a broader tendency relevant for ordinary ethics, namely, the tendency to presume a constitutive source of ethics and hence imagine ethics as localizable, and the concomitant neglect of cross-domain and cross-scale entanglements that are often an artifact of discursive practice. 9
For many, if not most, it is a truism that ‘politeness’ – whatever moral facets it may have – is an ideological instrument par excellence. ‘Politeness,’ ‘respect’ behavior and the like, whether codified in etiquette manuals or inculcated tacitly in institutional sites such as homes, classrooms, and workplaces, seem designed to reproduce asymmetries like those of class and gender (Bourdieu 1984; Langford 1989; Hemphill 1999; Davidson 2004). A chasm separates such expansive studies from much of the language-based studies of politeness and related phenomena (e.g. ‘deference,’ ‘respect’), however. The most industrious field here has been sociolinguistics, which has been heavily influenced by Brown and Levinson’s (1996) classic but oft-challenged work, Politeness: Some Universals in Language Usage. Brown and Levinson’s argument was strongly interactionist and largely bracketed the relevance of culture. A cornerstone in their theory was the purportedly universal notion of self-esteem or ‘face,’ a notion inspired by Goffman (1967). ‘Face’ is something interactants supposedly take for granted when they interact, something to which they feel both entitled and obliged to safeguard in others, so that if we wish to do a ‘face-threatening act’ (e.g. an invitation or request that imposes on someone), it may inspire us to twist ourselves up and perform all sorts of florid politeness displays, from circumlocutions to hedges to hints, to mitigate the threat. Brown and Levinson (1996: 253) resolved politeness into a polynomial expression consisting of contextual variables including ‘P’ for relative ‘power’ between speaker and hearer and ‘R’ for the weight of the face threat.
While the exigencies of interaction are said to explain politeness, Brown and Levinson did admit a measure of cultural ‘skewing,’ holding that these universal variables could be weighted differently as a function of cultural and even subcultural factors. (A Tamil proscription against asking where someone’s going, for example, is said to result in increased R for that category of speech act [Brown and Levinson 1996: 12].) Sociolinguistic work inspired by Brown and Levinson ventured further into culture. Sifianou (1992: 224), for example, found that Greek ‘off-record’ requests (requests that are not explicitly reportable as requests and hence afford the speaker plausible deniability) did not seem crafted to avoid imposition, as Brown and Levinson’s theory would predict. ‘Instead,’ she argued, ‘they provide addressees with the opportunity to express their generosity and solicitude for the interlocutor by offering’ (Sifianou 1992: 224). Sifianou linked this politeness behavior to virtues of hospitality and generosity, and, more deeply, to morally inflected conceptions of personhood in Greece.
Stronger culturalist accounts of politeness derive ethical content from wholly outside the interaction order. In his study of interethnic service encounters between Koreans and African-Americans in Los Angeles, Bailey (2001) suggested that Korean storeowners and cashiers tended to adopt a taciturn demeanor that was part of a broader style of ‘negative’ or ‘restraint’ politeness. Much like Sifianou, Bailey fanned out from his transcript-centered observations. He cited the Korean pragmatic notion of nunch'i – which may be glossed as ‘perceptiveness,’ ‘studying one’s face,’ ‘sensitivity with eyes’ – and even suggested that this ‘ideal, of communicating and understanding without talk, is present in the two most important religio-philosophical traditions of India – Confucianism and Buddhism’ (Bailey 2001: 128). Bailey saw culture in this taciturnity, and taciturnity in culture, creating, in effect, a closed circuit of interpretation. His integrative instincts led him to move across diverse data points and evidence types, but he seems to treat these materials as if they were all of a piece. Apparent convergence among his materials (viz. among ‘taciturnicity’ in service encounters, a reportable ethnometapragmatic norm of ‘perceptiveness,’ institutionalized religio-philosophical frameworks that valorize silent communication) is read through and taken as evidence of the unifying concept of ‘culture’. So much so that it would seem that beneath discourse lurks a generative framework of cultural meaning and value – a view of culture that has been charged as problematic for quite some time in anthropology but persists in research on ‘crosscultural’ and ‘interethnic’ (mis)communication (Lempert 2012b: 193–5). (Indeed, a limitation of cross-linguistic research on politeness is that it tends to exaggerate cultural wholeness, boundedness, and internal coherence and uses this framework to read discursive behavior.) In contrast, most contemporary work in sociocultural and linguistic anthropology tries to pry apart such data points and appreciate the disjunctures. In linguistic anthropology, one follows this methodology in part by distinguishing among types of reflexive activities and by situating these reflexive activities in social space (Agha 2007) – which includes detailing the institutional sites in which reflexive events occur (Silverstein 1998), the communicative media and modalities used, the categories of social actor involved, the stakes and fallout of such encounters. Rather than explain discursive behavior by fashioning synoptic, view-from-nowhere portraits of culturally defined moral systems, to leap, in other words, from discourse to the abstraction of ‘culture’ – as if the latter were an independent variable – contemporary anthropology aspires toward processual accounts of cultural production and circulation.
This problem never even arises in strongly culturalist-styled arguments on politeness, because they tend to treat group-relative moralities as if they were automatically expressed and reflected in discursive behavior, and tend to speak as if participants effortlessly noticed all this. Provenance and perception get conflated, in the sense that one fails to distinguish between practices that may have arisen once-upon-a-time out of ‘ethical’ considerations and those practices whose ethics are perceived, if only tacitly, as ‘ethical’ in real-time interaction. In Bailey’s work on service encounters, we wonder, to what extent did the Korean storeowners and their clients really feel that communicative taciturnity was culturally ethical – that it instantiated morally inflected norms and perhaps even venerable Buddhist and Confucian ideals? Or was it just that the ethnometapragmatic notion of nunch'i (‘perceptiveness’) surfaced during an interview and was then grafted back onto the service encounters, where it was assumed to have served as a heuristic – a way for participants (and analysts) to make sense of ‘taciturnicity’? The whole problem of how some among the many cultural presuppositions become relevant in interaction tends to get neglected in strongly culturalist readings. Methodological difficulties aside, the basic hermeneutic is clear: politeness behavior in interaction is read by tracing it back to a monolithically imagined culture.
Deep routes of ‘dialogue’: Downward reductions, ritual entanglements
If, for some culturalists, ethics is ferried into interaction from without, as we can see from some work on politeness, others find ethics by delving below the surface. Through a type of downward reduction interaction’s ethics is discovered below, such as in our biological inheritance. Research in psychology, primatology, and biological anthropology has found ontogenetic and phylogenetic evidence for what many consider to be ethical predispositions in humans. Tomasello (2009), for instance, has drawn attention to altruistic and cooperative behavior in humans that does not appear to be shared by our nearest primate relatives. Only humans, for example, seem to inform each another altruistically, such as by referring to things in the environment through pointing gestures (Tomasello 2008, 2009; Call and Tomasello 2007). Chimps sometimes do what looks like a deictic, finger-pointing gesture when they interact with other apes, but, unlike humans, they point only as a directive, not to help other apes find things. Ontogenetically, between the ages of 14 and 18 months, human infants have been shown to act helpfully (e.g. spontaneously picking up things that were dropped), and they do this not just with primary caregivers but also with strangers. For humans, recall, again, how even voicing a need or desire is often enough to spur others to action (Tomasello 2009: 6–13).
The question here – and it is not one that Tomasello addresses – is that of how these rudimentary and perhaps even hardwired habits ‘inform,’ ‘motivate,’ ‘afford,’ ‘structure,’ ‘condition’ ethical behavior, reflection, and discourse. (I enclose these terms in scare quotes to flag the way they suggest different architectures of ethical development, despite the fact that these are seldom elucidated.) A cooperative predisposition may be necessary for something we would want to call ethics, but it isn’t sufficient, and we need to specify what more is needed, how we get from such elementary predispositions to full-fledged ‘ethics,’ and what relations obtain among these levels (see Keane n.d.). Even if we grant that a domain such as the biological is constitutive of ethics in some respect or capacity, this tells us nothing specific about the relation between this constitutive ethics and second-order manifestations of ethical behavior and reflection, including institutionalized moralities inscribed in normative discourses and inculcated, for example, through language socialization. Cross-level relations are often far more orderly than a mere undifferentiated entanglement or enmeshment (cf. Ingold 2008, 2010) of ethics and discourse, and these relations are often produced through semiotic processes that merit analysis.
A vivid example of this orderliness – and the need to study it with more care – can be seen in the influential work of Jürgen Habermas, who argued for the immanence of ethics in dialogic interaction and tried to ground this ethics in universal pragmatic presuppositions. In critically revisiting claims from works like his Theory of Communicative Action (1984), we will see how Habermas unwittingly relied on some of the same methods used to ‘ritualize’ interaction. This is a telling reliance, because, as I will then suggest, cases of ritualization – to caption it crudely – illustrate well how invocations of ethics in practice involve cross-domain entanglements rather than immanence, even if the impression of ethical ‘immanence’ is one of the ritual effects.
Habermas defended the Enlightenment faculty of rationality against a league of critics, but first he had to rework it. He re-imagined this hallowed faculty as a capacity for dialogic communication in which humans seek consensus on criticizable validity claims, claims not just about the ‘true’ but also the ‘good’ (the domain of ethics) and the ‘beautiful’ (the domain of esthetics). For Habermas communicative rationality could be seen in everything from everyday disputes to formal debates, but it was all said to stem from a small bundle of underlying pragmatic presuppositions called the ‘ideal speech situation’. This bundle was not the result of nurture but rather a universal, species-level competence – a claim that amounts to a downward reduction of sorts, though not one made in evolutionary and biological terms. This bundle of expectations supplies speakers with a kind of tacit ethics of communication, an ethics that also smuggled in, as we shall see, morally inflected liberal-democratic ideals such as autonomy, rights, disinterestedness, and civility. Liberal ideals and the ideal speech situation converge, for it was in this very convergence that Habermas could make it seem as if liberal ideals were natural outgrowths of instincts about how we ought to communicate.
Habermas never meant the ideal speech situation as rules or constraints on discursive behavior but rather as counterfactual expectations that are orthogonal to the way people actually talk. His argument resembled that of philosopher of language H.P. Grice (1975). Grice posited a ‘cooperative principle’ that human actors presume whenever they engage in rational and efficient communication. Under this principle were ‘maxims’ of Quantity, Quality, Relation and Manner. The ‘Manner’ maxim, for instance, dictates ‘be perspicuous’. When flouted – and it frequently is, because maxims are not rules for or constraints on behavior – the violation attracts attention and triggers inferences (called ‘implicatures’) that invite hearers to move from ‘what is said’ to ‘what is done’. If you choose to be ‘vague’ or ‘ambiguous’ you prompt hearers to fish for the intention of your departure from expectations; the floutings are said to be recognized as floutings and used to ferret out intentions. Despite protests from anthropologists who have argued otherwise, Grice insisted that his cooperative principle had no positive cultural content, that it shouldn’t be confused with, say, ‘helpfulness’ or ‘altruism’. Habermas, too, was keen to keep his core pragmatic presuppositions pristine and insulated from culture. He didn’t succeed in this, and couldn't, because if he did then he would not have then been able to argue that liberal-democratic ideals were inscribed in tacit presuppositions that inform discursive practice. Through what may seem like a semiotic sleight-of-hand, Habermas ended up investing the ideal speech situation with cultural and ideological value.
It is instructive to recall how he did this. Habermas argued that the ideal speech situation’s cardinal principle was symmetry – symmetry, for instance, in terms of the speech acts we carry out, the roles we inhabit when we interact, even the telos of the dialogic event – the shared goal being to arrive at intersubjective consensus. Instincts dictate that everyone should enjoy being speaker and shouldn’t be stuck in the role of hearer, for example. And no one type of speech act ought to be allowed for some but disallowed for others.
Why symmetry? Habermas never tells us why, but it hardly takes a semiotician to see that mutually reinforcing symmetries might help project ‘equivalences’ between actors, which, in turn, can invite us to imagine them as ‘equals,’ perhaps even possessed of individual ‘equality’ and ‘rights’. Act out these symmetries in interaction and the ideal speech situation supplies iconic backing for liberal-democratic principles, so that Habermas, in effect, could bolster Western liberal institutions from below, as if these were not ‘mere’ contingent institutions belonging to the West. From the early 1980s Habermas extended his reflections on communicative rationality into ‘discourse ethics’. He rejected the monological Kantian exercise of the categorical imperative, where one deliberates in solitude about which norms are generalizable to all, in favor of a dialogical exercise of argumentation, where one wrangles with others over competing normative claims (Habermas 1995, 1999; for a useful critique, see Ferrara 1990).
What Habermas did is not unlike what’s involved in designing a ritual, where ‘poetics’ matters crucially. As the semiotic and linguistic anthropological literature has shown, emergent poetic structures in discursive interaction, especially ritualized interaction, can help figurate social relations and imbue them with value. In the spirit of the Russian Formalist literary critics, Jakobson (1960: 356) defined the ‘poetic function’ as the foregrounding of message form, a ‘focus on the message for its own sake’. (Think of how rhyme elevates form over content.) Poetic structure may be most salient in the world’s metrical poetries, but forms of felt repetition and parallelism crop up in all kinds of discourse, from nursery rhymes to spells to conversational talk (see Fleming and Lempert, in press). Poetic structures can function reflexively in discourse, helping parse the stream of speech into like-and-unlike units that can then motivate pragmatic effects. In ritual contexts, dense cross-modal poetic structures typically serve as scaffolding for the construction of intricate diagrams of social and cultural order, including those that exhibit moral ideals.
For example, in Urban’s (1990) discussions of South American ceremonial dialogues, poetic patterns of repetition and symmetry helped participants act out a cultural ideology of solidarity. Of the Shokleng origin-telling myth style of ceremonial dialogue termed wãñeklèn, Urban (1990: 101–2) described how … two men sit opposite one another in the middle of the plaza, their legs entwined.… One interlocutor leads, uttering the first syllable of the origin myth. The respondent repeats that syllable, after which the first speaker utters the second syllable, and so forth, in rapid-fire succession. Speakers move their heads and upper torsos rhythmically in time with the syllables, which are shouted with extreme laryngeal and pharyngeal constriction.
Rituals like this invite participants to act out the desired action, as if through pantomime (Silverstein 2003b). Through a superabundance of mutually reinforcing signs –discursive, prosodic, bodily – participants trace out a diagram of what the ritual as a whole hopes to effectuate. Shokleng speakers, in effect, enact a multi-dimensional ‘model of and for coordination more generally, this coordination in turn representing a fundamental building block of social solidarity’ (Urban 1990: 106). The cross-modal poetics of this dialogic ritual helps speakers materialize and publicly exhibit cultural ideals, and some take this to be a characteristic of ritual in general. In a recent review Stasch (2011: 162) aptly notes that ritual ‘intensifies features common to human activity at large’ and is hence parasitic on ‘ordinary’ interaction. Ritual reflexively foregrounds and caricatures lower-order features of communication and relies on the poetic function to do this.
So observe the fate of largely tacit, everyday interactional patterns, like the way humans ‘ordinarily’ swap speaker and hearer roles when they converse. These regularities undergo moralization as they are foregrounded, stylized, and incorporated into an extra-ordinary ritual genre. Ritualization is one way (not the only way, to be sure) in which everyday communicative behavior can be sourced and repurposed. The reflexive sourcing of everyday interaction here creates an entanglement in which distinct orders are thereby distinguished, orders that now seem to rest on a foundation of ‘ordinary’ interaction.
It’s important to trace out these entanglements. In Ordinary Ethics, Lambek (2010: 3) writes that the volume’s ‘focus is less on special cases, unusual circumstances, new horizons, professional rationalizations, or contested forms of authorization than on everyday comportment and understanding’. Such matters are sometimes broached, but ‘precisely to show how they are drawn into or drawn from the ordinary’ (2010: 3). In this case of dialogic ritualization, we can say what it means to be ‘drawn from the ordinary’. Lower-order patterns of discursive-interactional ‘symmetry’ (e.g. the reciprocal, metricalized alternation of participant-role inhabitance across turns of talk), which may be informed by their own normativity, are transfigured into a second-order moral diagram of social relations. Through this foregrounding of ‘ordinary’ interactional normativities, a second-order institutionalized morality can then be made public, made recognizable to others. 10
Returning to Habermas, he concealed the labor by which he created the impression of liberal immanence in dialogic interaction. In one moment he teased out pragmatic presuppositions from below interaction, painstakingly separating this stratum from a surface world of actual dialogue in which the light of liberal ideals spreads. In the next moment he flattened the two worlds, making it seem as if liberal ideals were immanent in these tacit heuristics – a naturalizing move central to his argument’s persuasive force. This is not unlike the way strong culturalists create and then collapse levels of analysis by exaggerating the convergence across data points, so that they end up seeing discursive behavior like politeness as shadows cast by culture. This is not unlike the way some interactionists narrow and perimeterize the interaction order till they can seem to derive morality from within it and not bother with entanglements at all. This is not unlike the way some track ethics to biologically-based cooperative and altruistic predispositions, then leave us stranded there, so that we cannot find our way back. Neglected in each case, mutatis mutandis, are the specific entanglements of the ethical in relation to everyday discursive practice and the means through which entanglements are created. Merely dropping a purified domain-based view of ethics and embracing its allegedly messy entanglements – a view that revels in rather than explains the way ethics is caught up in this and that – will not do, since these entanglements have empirical specificity that can be analytically reconstructed. The specificity of this entanglement should be narrated, and theorized, in any anthropology of ethics, ordinary or not.
As public rituals that deviate from the everyday, practices like South American ceremonial dialogues are anything but ordinary, so why include them here? The reason for considering strongly ritualized interactions is that they illustrate well one kind of ethical entanglement and hint at the existence of many others. The larger point is that we should take care when segregating ordinary ethics not to sever the entanglements that make it seem quotidian in the first place. Discursive practice may seem ‘small,’ ‘everyday,’ ‘tacit,’ and so on – there’s no denying that, that’s how participants themselves often feel – but those are practical accomplishments that need explanation. Ordinary ethics is at its best when it expands the range of behaviors that can serve as sites for ethical practice and when it encourages us to look across them, not when it hardens dichotomies of small and large, micro and macro, implicit and explicit, and makes it seem like we have to choose.
Might it be a symptom of such dichotomies that explanations of ethical entanglements often seem thin and wooden? Das (2012: 138), for example, sometimes says that dramatic enactments of ethical value – as in public rituals – are ‘grounded within the normative practices of everyday life,’ but one isn’t entirely sure what this grounding entails. Bergmann (1998: 283) uses ‘proto-morality’ for underlying universals of interaction that serve as ‘preconditions’ for morality but are not themselves fully-formed morality, but how exactly does one move from proto-morality to morality, and should we expect the same for all species of moralization, from enactments of virtue to moral deliberation? These examples could be multiplied, and they are meant only as a reminder of what should be apparent, that there are many conceivable ethical entanglements and that we have at our disposal a wealth of explanatory resources, from causal logics to the ecological notion of ‘affordance’ (Gibson 1979) to the semiotic notion of ‘motivation’. While some may want to settle how moralities arise and are distributed across our disciplined objects of knowledge and domains of study, we should pause and treat ethical practice and its entanglements with more care.
‘Where is the ethical located?’ asks Lambek (2010: 39) in the lead line of his essay, ‘Toward an Ethics of the Act’. Musings on the where of ethics can lead us astray. More than where is the question of when, and how. Apart from moral breakdowns and self-conscious deliberation and the histrionics of public ritual, it is not always clear how ethics even matters for interactants (cf. Keane 2011: 169). It takes labor to make ethics matter in discursive practice, and this labor frequently results in entanglements, such that ethics cannot be said to be located anywhere in particular.
And yet the chimera of an immanent ethics, of an ethics always already being there – or at least being traceable to some determinate source – remains. It remains both an allure and obstacle for ordinary ethics. Perhaps this immanence is the product of a certain imaginary that is preoccupied with the hoary question of where ethics comes from, nature or nurture, a transcendent God or contingent custom. Perhaps it is the result of antinomies that set the unreflective immediacy of ‘practice’ against self-conscious ‘reflection’. Perhaps this immanence is even aspirational, in the sense that, as ethical subjects-in-becoming (Faubion 2011), it surfaces especially when we exaggerate, for purposes of self-cultivation, the breadth and depth and fluency of our own present moral grounding. Whatever the tangle of motivations may be, the ‘ordinary’ in ordinary ethics should not beguile us with its achieved immanence but inspire us to study the communicative methods and labor – discreet to dramatic, implicit to explicit, improvised to institutionalized – through which actors make ethics recognizable and effective in discursive interaction.
Footnotes
Acknowledgements
Thanks to Webb Keane and Charles Zuckerman for sharp comments, to the doctoral students who participated in a characteristically spirited session of our Linguistic Anthropology Lab at the University of Michigan, to Ann Rawls for fielding questions, and, last but not least, to Anthropological Theory’s astute reviewers and editors Jonathan Friedman and Joel Robbins, who pushed me to hone and clarify my argument.