Confirming Allusions: Toward an Empirical Account of Action

by Emanuel A. Schegloff
Citation
Title:
Confirming Allusions: Toward an Empirical Account of Action
Author:
Emanuel A. Schegloff
Year: 
1996
Publication: 
The American Journal of Sociology
Volume: 
102
Issue: 
1
Start Page: 
161
End Page: 
216
Publisher: 
Language: 
English
URL: 
Select license: 
Select License
DOI: 
PMID: 
ISSN: 
Abstract:

Confirming Allusions: Toward an Empirical Account of Action1

Emanuel A. Schegloff

University of California, Los Angeles

As part of a larger effort to develop an empirically grounded theory of action, this article describes a previously undescribed action that occurs in talk-in-interaction. The practice of agreeing with another by repeating what they have said is shown to constitute the action of confirming an allusion-that is, confirming both its "content" and its prior inexplicit conveyance. The author reviews the past treatment of "action" in sociology and the key constraints on undertaking an empirically grounded account. The account of "confirming allu- sions" is offered to exemplify what this undertaking will involve: sev- eral instances of an unremarkable usage in conversation are displayed and used to formulate a puzzle, a database is developed for the explo- ration of the target usage, and a candidate solution to the puzzle is formulated, exemplified, and defended through a range of analytic techniques. The linkage between the practice and the action that it implements is analytically sketched by examining other uses of repeti- tion in talk-in-interaction. In conclusion, the significance of both the theme and the analysis for studies of interaction and culture and for sociological theory is discussed.

INTRODUCTION

The primary goal of this article is to provide an account of a previously undescribed action that parties to talk-in-interaction may do. Ideally, it would begin by launching directly into an account of this action, but,

' Earlier versions of the present text have been presented in a number of venues, including the Linguistic Institute, Stanford University (1987), the Nordic Council Summer Course on Spoken Interaction, Mullsjo, Sweden (1987), the nineteenth con- gress of the Applied Linguistics Association of Australia, Melbourne (1994), and as the Brittingham Lecture, Department of Sociology, University of Wisconsin-Madison (November 1991). My thanks for useful comments on earlier drafts or presentations of this material to Jack Bilmes, Steven Clayman, Fran~oise Brun-Cottan, Susanna Cum- ming, Paul Drew, John Heritage, Robert Hopper, Lena Jayussi, Gene Lerner, Douglas Maynard, Elinor Ochs, and Sandra Thompson. I am especially indebted to Paul Drew for drawing my attention to relevant British data. The paper which appears here is quite different from the one I meant to publish, in large measure because of the input of the AJS referees, whose impact continues to be substantial, even if only anonymously ac- knowledged. Direct correspondence to Emanuel A. Schegloff, Department of Sociology, University of California, Los Angeles, California 90095- 155 1.

O 1996 by The University of Chicago. All rights reserved.

0002-9602197/10201-0005$01.50

AJS Volume 102 Number 1 (July 1996): 161-216 161 perhaps more than is ordinarily the case, it is first necessary to set the undertaking into a theoretical context relevant to sociologists. The kind of introduction that turns out to be in point for the project-one relevant to its presentation rather than its conception and execution-is informative about both the topic and about sociology's development and current situation as a discipline.

This article is conversation-analytic in orientation. In common with many such papers, it invites reading not primarily for its bearing on so-called sociolinguistic themes but as a contribution to both basic socio- logical theory and appropriate methods for developing it. Past preoccupa- tions have put the economy and the polity, religion or culture, or person- ality, character and "human nature" (as in theorizing inspired by Freud) at the center and foundation of social and sociological theory. In spite of a recurrently expressed recognition by sociologists from Durkheim and Mauss to Bourdieu (to cite only the French lineage-the German one might be from Simmel to Habermas, the American one, from Mead or Parsons to Goffman or Garfinkel) that language figures centrally in the organization of social life, it has remained peripheral to the main thrusts of the discipline. Several strands of contemporary theorizing (most nota- bly those associated with the names of Bourdieu and Habermas) have sought to put language, communication, and "practice" in a position of comparable theoretical "gravity"; still, none has yet provided a clear depiction and exemplar of how the prima facie, observable embodiment of sociality-action, activity, and conduct in interaction-as effectuated through the deployment of language and the body can be put at the center of theorizing about the social and can be grounded and elaborated in detailed, empirical analysis of that conduct.

The need to establish the relevance of "describing a previously undescribed action" as a sociological undertaking reflects the absence from sociology's past of a phase of naturalistic observation-one given over to the accumulation of accounts of the basic inventory or repertoire of social actions and forms of social conduct which constitute the primary observable referents of the sociality of this social species (the cognate for interaction to the database supplied to comparative-historical inquiry by the work of historians, to demographers and other quantitative students of social structure by the gathering of official social statistics, etc.).' Per-

'When George Homans (1961) undertook to offer a synthetic account of "Social Behavior: Its Elementary Forms," it was not actions that he described. I should note that I use here a different sense of "naturalistic" than the one that informs the discussion in Matza (1969), which, e.g., includes ethnography-both participatory and interview based-within naturalism, whereas I would treat ethnography as an alternative.

haps it is because sociology never went through such a phase as a disci- pline that it did not quite know what to do with Erving Goffman who, more than any other sociologist, tried to draw the discipline's attention to aspects of this raw material and its relevance to the discipline, though I will suggest below that his effort in this regard was itself problematic (see also Schegloff 1988~).

Goffman had better luck with anthropologists. In part this was because anthropologists had had good reason over the course of the development of their discipline to wonder what some "natives" were "doing" by conducting themselves in a certain way, for example, by talking in a certain way. Both linguistic obstacles and so-called "cultural differences" could pose quite sharply the problem of "recognizing actions," and then the analytic problems of describing what those actions were and how they were done or accomplished. Aside from the encounter with actions that had no cognate in their own culture, there were also "indigenous" ways of doing actions that did (appear to) have a place in the field-worker's culture but whose enactment was different (see, e.g., Frake 1964). Both the practical exigencies of fieldwork in an "exotic" setting and anthropology's principled disciplinary commitment to expli- cate the detailed diversity and specificity of human culture funneled at- tention to "the different." By the mid-1960s to early 1970s anthropolo- gists converged from different routes (some of which subsequently diverged again) on an orientation to "culture" as the ways of doing things-whether as linguistic or cognitive procedures (as in Goodenough 1957, 1964), ways of meaning (Geertz 1973) or ways of communicating (Gumperz and Hymes 1964)-under rubrics such as "ethnoscience," "componential analysis," "ethnosemantics," "ethnography of communication," and/or simply "cultural anthropology."

When my late colleague Harvey Sacks offered an undergraduate class in social science in 1965, his characterization of a culture as "an appara- tus for generating recognizable actions" (1992, 1:226), this was not out of keeping with the theoretical zeitgeist of the anthropology of the time. And when he proposed as an analytic task that we provide an account for how "describing" was done (1972, p. 332; see also the transcript of a lecture in the spring of 1966 [Sacks 1992, 1:236-37]), this could be understood as a confluence between ethnomethodology's interest in prac- tical theorizing and commonsense interpretation on the one hand (Gar- finkel 1967), and, on the other, a then-current anthropological interest in describing how actions are done, albeit in this case actions from one's own culture.

But as the persistent, bemused popularity of Horace Miner's classic paper (1956) "Body Ritual among the Nacirema" shows, the same sort of attention to, and description of, actions in one's own culture never had quite the same resonance, even in anthropology. The actions were "transparent" to comembers of the culture; even naming them by action names might appear a bit arch and scholastic. Even more so was this treatment accorded accounts of the practices, rules, or mechanisms by which these actions were done. They just "were" invitations, requests, promises, insults, and so on. For mainstream sociology, and even for most of its new tributaries of the time, this sort of orientation to this society's actions apparently constituted an unassimilable-or uninteresting-proposal and undertaking.

Perhaps the development of a serious cultural sociology can change that, if by "cultural sociology" we understand something that includes not only high and popular culture, and not only that "anthropological" sense of culture that features values, beliefs, taste, fashion, and so on, but one that focuses on the "repertoire" of actions and practices out of which the quotidian life of the members of a social species is fashioned. To the degree that this ambition can be realized, cultural sociology serves simply as a key part of a serious general sociology, and not just theoreti- cally (as it was in Parsonian theory) but empirically. It is this ambition that this article pursues.

After a brief prCcis of largely conceptual modalities of addressing "ac- tion" within sociology and without, the article reviews several empirical accounts of the practices for doing particular actions, as background to the present undertaking, which is to describe a new type of action and the mode of inquiry by which it was come upon.

ACTION AND ACTIONS

With all the commitment of major segments of the social science commu- nity, and sociology in particular, to "action" as a pivotal component of social life and its disciplined study, actual action-and particular ac- tions-have been curiously absent from sociological inquiry and dis- c~urse.~

At times they appear to be virtually epiphenomena1 expressions of "underlying" factors, processes, and variables-norms, rationality, conformity, power, system functions, and the like, but important largely as the public face, the accessible display and indicator, of those underly- ing forces. Nowhere has the contrast between conceptual centrality and

As will become obvious in the ensuing text, the "action" or "social action" under discussion here is not "political action" or "social action" in the interests of social change. These have, of course, figured centrally in a wide range of sociological under- takings over the historical career of sociology as a discipline, and with special intensity in the last 35 years or so. The focus here is rather on the "action" composing the moment-to-moment flow of daily life in-and outside-interaction.

empirical dismissal been more marked than in the Parsonian "theory of

action," which played so prominent a role in the sociology of the middle

decades of this century. The "action" that figured so centrally for Par-

sons was a conceptual or analytic unit-the "unit act," with theoretically

conjectured components such as the inescapable facets of action-actor,

goals, means and conditions, and so forth (Parsons 1937). Actual actions

were "merely" concrete and empirical and of interest only insofar as

they could be analytically decomposed into the then-current terms of

theorizing (cf. Heritage 1984b, pp. 103-34).

Paradoxically, the main alternative has been to presume actions cen- tral, only then to look elsewhere for the "unofficial"-and therefore more "realn-agenda, as in much of Goffman's writing. Goffman, whose microethnography might have been thought to cast him as the most likely observer and analyst of social actions, more often took the tack that whatever else ordinary actors might in the first instance appear to be doing-walking, calling the doctor, offering medication, sanctioning rule violations, expressing intimacy by holding hands, serving a meal, waiting, and so on-what was sociologically key was that they were managing impressions and information (Goffman 1959, 1963b), at- tending to considerations of face (Goffman 195 5, 1967), organizing the allocation of attention and involvement (Goffman 1963a), and so forth. First-order actions here are cast as the official or ostensible version of what is going on, a veil whose penetration and revelation is the true calling of the sophisticated, perspicuous sociologist or anthrop~logist.~ But in these quarters no more attention was given than elsewhere to whatever persons might be thought-or observed-in the $first instance to be doing. Ordinary social actions have thus fallen between the several sociological cracks; they are not to be found in action theory, in the microethnography of everyday interaction, or in the study of the underly- ing organizing social processes that drive social life.

Different as they may appear, these "approaches" come to the materi- als of everyday life with a theoretical filter that separates the sociological wheat from the chaff, the gold from the dross, the important from the trivial, the real from the apparent, the enduring from the transient. Gen- erally this has resulted in accounts that formulate the actions being char- acterized not by reference to the projects of the actors who enacted them

They may also have figured for Goffman-like the allocation of opportunities to participate, resources to assure mutual understanding or remedy its breakdown, etc.-as a kind of underlying set of "system requirements," a proper object of atten- tion for "communications engineering" (or perhaps, in this case, social engineering), but set in contrast to the real social science interest in interaction, which addressed itself to "ritual requirements" (see Goffman 1976, pp. 265-70). For a fuller discussion, see Schegloff (1988a, pp. 93-100).

and were their recipients, but only by reference to the terms of the theory that was "processing" them, the only terms by reference to which they were taken to matter-as in Bales's (1950) reduction of all actions to 12 characterizations, formulated by reference to the then-current stage of Parsonian action theory. With all the differences between the various sociological treatments, actual, empirical, naturally occurring garden- variety actions have been found wanting in their own terms, from what- ever direction the approaches approached (cf. Drew and Heritage [1992, pp. 10-161, who find virtually no writing by sociologists on the actions done by talk).

Aside from the theoretically motivated or presuppositionally grounded sources of this treatment of ordinary actions, there was a likely methodological source as well. Methodologically no one fully exploited the use of recording technology (audio or video) to provide the opportunity for freezing the object of inquiry-the stream of conduct of which social settings are composed-for the repeated de- tailed examination that might allow the analyst to shed, at least partially, the relentless interference of vernacular familiarity in the analytic depiction of actions. For example, although much of Bales's data was recorded on audiotape after the challenge of coding the stream of behavior in real time was recognized as problematic, the audiotape was erased for reuse as soon as the "acts" had been coded. The results of the coding were taken to be "the data," leaving actual, recorded conduct as a kind of scientific detritus.

Vernacular familiarity was very likely one obstacle to the passage of the problem of recognizable action from anthropology to sociology. By contrast with work in "strange" settings, the action-import of usages in one's native or well-assimilated language or culture may fade into invisi- bility by virtue of its very familiarity. The same commonsense knowledge of the culture, and the semiotics, pragmatics, and discourse structure of a language that helps to constitute our cultural and linguistic competence blinds us and impedes our capacity to get at the constitution of action technically. Indeed, the combination of native knowledge and the ab- sence of recorded specimens of the quotidian can obscure for us that there is a distinctive project to be undertaken here. We need to exploit the newly available technologies not only to do better-more "precisely" or compellingly-what has previously been done by observation in real time or through ethnographic elicitation; we need to ask what new ana- lytic possibilities are made available by this new technology, and these may follow on the new observational possibilities. The increasingly so- phisticated recording technologies of the last three decades may thus be to this area of inquiry what the microscope was to the life sciences earlier.

This is by no means a new idea. To realize its possible benefits, how- ever, we need to press inquiries into what speakers can do-do do-with language and the other resources deployed in interaction. And we need to press those inquiries especially with materials to which we bring native competence and cultural membership. If this undertaking falls to any discipline in the current map of the social sciences, it falls to sociology. The prospect is the development of an account of action in interaction grounded in the observable details of conduct in naturally occurring so- cial settings, which can be juxtaposed to the abstract and theoretical accounts of past and current work, accounts which were grounded in conceptual and normative considerations and invoking the common cul- ture's stipulated understanding of typical courses of human conduct that are taken to be shared by the writers and readers of this professional literature.

Some years ago, Sacks (1992, 2:419-20) began a course of lectures by underscoring the troubles of studying talk-in-interaction with imagined or typified data, but his point extends to other nonnaturally occurring data as well.

What I want to argue is that if a researcher uses hypotheticalized or hypo- theticalized-typicalized versions of the world, then, however rich his imagi- nation is, he is constrained by reference to what an audience, an audience of professionals, can accept as reasonable. . . . One is then debarred from using these kinds of materials [hypotheticals which are not credible to an audience]. And that debarring of lots of things that actually occur, at least presumably affects the character of social science very strongly.

By contrast, Sacks continued,

a base for using close looking at the world for theorizing about it is that from close looking at the world you can find things that we couldn't, by imagination, assert were there: One wouldn't know that they were typical, one might not know that they ever happened, and even if one supposed that they did one couldn't say it because an audience wouldn't believe it. Where, then, if we can add to the stock of things that can be theorized about we will have done something more or less important-if the things that we've added have any import to them.

Thirty years later, this remains a robust ideal and a useful measure of many research projects. Indeed, one point of this article is of just this sort. It is to explore something that speakers of language in interaction do observably do with it, and to add thereby to an inventory of sorts, a catalog of recognizable social actions in this culture, in this case one which language is used to accomplish in interaction. The aim is to find it and provide an account of it empirically and precisely, not imagina- tively or typically or hypothetically or conjecturally or experimentally, and to use actual, situated occurrences of it in naturally occurring social settings to control its description. Whatever the value of previous philo- sophical, conceptual, and text-based theories of action may turn out to have been, there is a separate and distinct place for an empirical theory or account of action, or actions. What might it look like? How might we arrive at it? For an empirically grounded account, we need to deal with the latter question first, and the first step in addressing it is to assemble at least a moderate set of descriptions of particular actions-a corpus to which an empirical theory of action can be answerable and by which it can be constrained.

DESCRIBING ACTIONS

But what does a description of an action look like? Before proceeding with the account of the action that is the focus of this analysis it may be useful to sketch a few accounts of actions and the practices for accomp- lishing them produced in past work.'

Even quite elementary procedural accounts of the practice(s) for doing some action ("elementary" both in the sense of "simple" and in the sense of "explicating the elements"), grounded as they are in the empirical detail and specificity of actual instances, open the door to empirical examination and, where appropriate, critique and modifi- cation. For example, once Sacks ([1964-651 1992, 1:97) had proposed that greetings are done by putting some greeting term in first position in a conversation (". . . something is a 'greeting' only if it's a 'greeting item' in a 'greeting place'"), it was possible to examine the "hello" with which most American "private" telephone conversations begin to assess whether it is the action "greeting" that these "greeting items in greeting places" are employed to accomplish, and are taken to accomplish by their recipients, as the basis for subsequent action. This was not a conceptual undertaking and critique but an empirical one (Schegloff 1967, 1968, 1986), and it yielded the proposal that at

'Space limitations preclude an account of what the description of an action looks like in the main academic enterprise in this domain-speech act theory, especially of the sort introduced by Austin (1962, 1970) and developed in a distinctive direction by Searle (1965, 1969, 1976). Speech act theory might be of special interest to (some) sociologists because it has been adopted as a theoretically central element of Ha- bermas's theoretical project from the early stages (1970) until its recent major state- ments (1984-87), in spite of its atomistic, asocial, ahistorical, and decontexted char- acter. For a discussion contrasting speech act theory with conversation-analytic approaches to describing actions, see Schegloff (1992a, pp. xxiv-xxvii). For some obstacles to applying speech act theory to actual utterances, see Schegloff (1988~). Efforts to develop a more empirical form of speech act theory grounded, e.g., in linguistics (see, e.g., Labov and Fanshel 1977) have met with only qualified success and have not proven generative of further work (see the critical discussion in Levinson [1983, pp. 226-831).

least these apparent greeting items in first position in a conversation were understood by parties to the incipient conversation to be doing not greeting, but answering the summons embodied in the phone's ring (with consequences such as the subsequent occurrence of an exchange of greeting terms which were doing greeting).

So one form that an account of an action can take is a characterization of some form or practice of talking and some characterization of the place or location in which that practice is employed-as in articulating a greeting term in an initial slot or exchange in a conversation. Relatively few actions have their procedural basis exhausted by an account of this sort, and in accounts of this sort relatively few can invoke the availability of a limited set of "dedicated" linguistic resources such as "greeting terms." But "greeting" is not the only action we can explicate in this way. In an exchange with an "artificial intelligence" computer scientist at a conference in which "compliments" were being discussed, I asked for an account of how a computer might be programmed to "do" a compliment and, very likely taking this as a request for a "speech-act" type of account, my colleague declined even to try. But one can begin with something as simple as "Select a positive assessment term for some feature of the recipient/interlocutor or interlocutor's just prior conduct." Although more is clearly necessary (as is a specification of even this much), a simple account of the constructional elements and their position- ing is often a good start on an at least partial account.

One important consequence of undertaking such an account is that it can yield empirically grounded results at variance with our commonsense intuitions about how some action is accomplished or what action some utterance is to be understood to have accomplished. For example, in her account of assessments in conversation, Pomerantz (1984) focuses on responses to assessments, specifically on agreements and disagreements with them. The "position" component of her account is specified as the turn following an initial assessment, and one question she entertains is how "agreeing" and "disagreeing," respectively, are done in response to the initial assessment. Commonsense intuition might suggest that offer- ing in the next turn an assessment of the same object that has just been assessed and offering an assessment of the same class or valence (i.e., positive if the prior assessment was positive, negative if it was negative) would go a long way toward describing "agreeing with an assessment." Consider the following exchanges (taken from Pomerantz [1984, pp. 68- 691):

Excerpt a (GJ : 1)

A: She's a fox!

L: + Yeh, she's a pretty girl.

Excerpt b (NB :VII : 2)

E : e-that -Pa: t -isn'she a do : [:ll?

M: -t [iYeh isn't she pretty,

Excerpt c (AP: 1)

G : That's fantastic.

B: -t Isn't that good.

As readers may already have gathered from the framing with which I presented these fragments, the responses in these exchanges are of the "same sort" (in these cases all positive); but they contrast with instances such as the following (Pomerantz 1984, p. 65):

Excerpt d (JS :I1: 28)

J: T's- tsuh beautiful day out, isn't it?

L: + Yeh it's just gorgeous . . .

Excerpt e (MC: 1)

A: Isn't he cute

B: -t 0::hhe::sa::DORable

These "same sort" assessments are "upgraded," (Pomerantz 1984, p. 65) and do indeed accomplish agreement. But the second assessments in excerpts a-c are "downgraded" (p. 68), and although they are of the "same sort," they are taken as disagreements by the first speaker and are followed by reassertions of a "higher" assessment (pp. 68-69):

Excerpt a (G J : 1)

A: She's a fox!

L: Yeh, she's a pretty girl.

A: + Oh, she'sgorgeous!

Excerpt b (NB : VII :2)

E : e-that -Pa : t -isn'she a do : [ :11?

M: [iYeh isn't she pretty,

(.)

E: + -Oh : she's a beautiful girl.

Excerpt c (AP : 1)

G :   That's fantastic.
B :   Isn't that good.
G: -t That's marvelous

So the description of the action of agreeing with an assessment turns out to involve more than "same object, same valence in next turn," but must include (because the data show that the parties do it that way) practices of upgrading, downgrading, and the like. Pomerantz goes on to describe other practices in the talk that are elements of the accomplish- ment of agreeing and disagreeing with assessments as well, ones that would almost certainly elude an "untechnical" examination-one not informed by repeated and detailed examination of actual occurrences. (The centrality of action itself aside, sociologists may register that other key features of a social order are implicated in such exchanges-assessments being one key locus of social and cultural norms regarding "the good, the true, and the beautiful," and agreementldisagreement being one interactional locus of consensus and conflict. These little ex- changes are, then, the mundane sites in which the grand, macrotheoreti- cal themes about norms and values and cultural capital are played out.)

Another form which accounts of actions can take shows them to be the outcome of specifiable practices of talking, or classes of such prac- tices. One class of such practices may be termed "formulating," and a number of formulating practices have been shown to serve as ways of accomplishing determinate actions. For example, Sacks (from a lecture given in the spring of 1966 [1992, 1:300-3051; and another in the fall of 1968 [1992, 2:71-831) shows how the practices of formulating for a new arrival to a conversation what was being talked about before his arrival ("We were in an automobile discussion") can serve as a possible invita- tion for the person to join in or not, as the case may be. Pomerantz (1980) shows how speakers' formulations of their own "limited access" to a setting or occurrence as a way of talking about it ("Your line's been busy," "I saw you drive by last night," "You were in room 252 a long time this afternoon," etc.) can serve to solicit their interlocutors' account of that setting or occurrence without explicitly requesting it. Elsewhere (Schegloff 1988a, pp. 120-28) I link "negative observations" or the for- mulation of an occurrence as an absence ("You didn't get an ice cream sandwich") to the doing of complaining. And so forth.

Another type of finding shows how alternative vehicles for implement- ing a facet of the talk-for example, doing it with or without delay (cf. Pomerantz 1984; Sacks [I9731 1987; Schegloff 1988b)-serve to accom- plish additional actions as well, such as aligning with or against, agreeing or disagreeing, encouraging or discouraging.

The upshot of this discussion is that the description of an action and how it is done can extend in many, not necessarily anticipatable direc- tions. The stance toward their object that formulations in the talk em- body, the timing and delivery of the talk, the selection of implementing linguistic and body-behavioral resources-all have entered into past ac- counts. And what can count as "actions" in the first place has also involved a mixture of the familiar and unfamiliar. Jefferson (1984, [I9831 1993) shows that choice of interpolation by a current recipient of the talk (e.g., as between "uh huh" and "yeah") can display either projected continuing recipientship on their part or an incipient move to assume speakership, possibly with a change in topic. Heritage and Sorjonen (1994) show that starting a question with LLand" can do the activity of making the question-to-follow sustain an already ongoing activity rather than constituting a new departure in the talk. I too have shown that some apparently paradoxical utterances ("Can I ask you a question?") can serve to mark what directly follows as a "parenthetical" preliminary to something else that will follow it (Schegloff 1980).

Virtually all of these results emerge from an "unmotivated" examina- tion of naturally occurring interactional materials-that is, an examina- tion not prompted by prespecified analytic goals (not even that it be the characterization of an action), but by "noticings" of initially unremark- able features of the talk or of other conduct. The trajectory of such analyses may begin with a noticing of the action being done and be pursued by specifying what about the talk or other conduct-in its con- text-serves as the practice for accomplishing that action. Or it may begin (as did the research reported below) with the noticing of some feature of the talk and be pursued by asking what-if anything-such a practice of talking has as its outcome.

Whatever trajectory of inquiry is followed, to support an empirically grounded account of action at least three distinct elements ought ideally to enter into an account of the action that some utterance implements.

  1. First, the account requires a formulation of what action or actions are being accomplished, with compelling exemplifications in displays of data and analysis, including ways of "testing" the claim via confronta- tion of problematic instances and apparent "deviant cases," if possible. (The next six sections below are in various ways addressed to this task.)
  2. Second, there must be a grounding of this formulation in the "real- ity" of the participants. Here the investigator undertakes to establish that the formulation is not an academically analytic imposition on con- duct that may have been quite differently understood and experienced by the participants. This requires some demonstration that the inter- locutors in the data being examined have understood the utterances (or other conduct) in question to be possibly doing the proposed action(s) or that they are oriented to that possibility-a demonstration ordinarily grounded in the interlocutors' subsequent talk or conduct (see Sacks, Schegloff, and Jefferson [1974, pp. 728-291 on proof procedures; also Heritage 19843, pp. 254-60). This immediately subsequent talk, being appropriate to-or even responsive to-what preceded it, ordinarily dis- plays an understanding of what that preceding talk was "doing." (This issue is addressed intermittently in three sections below, from "A Can-

didate Solution" through "Evidence from Avoidance and Nonoccur- rence.")

3. A third element of a proper account of an action is an explication and analysis of what it is about the observed talk or other conduct or the practices embodied in it, which makes the enactment of that talk/ conduct possibly an instance of the proposed action, and makes it analyz- able by the coparticipants as an instance of that action, that is, why or how that practice can yield that action. It is not enough to show that some utterance was understood by its recipient to implement a particular action (and, indeed, sometimes this is not possible because it was not so understood, though it "might have been"). In order to provide analyti- cally the grounds for the possibility of such an understanding, an account must be offered of what about the production of that talklconduct pro- vided for its recognizability as such an action; that is, what were the methodical, or procedural, or "practice-d" grounds of its production. Once explicated and established, this serves as part of the account of the utterancelaction, whether or not it was so understood by its recipient on any particular occa~ion.~

If the account of action in which we are inter- ested is to be understood as informiag the deployment and uptake of particular acts by participants in singular contexts in real time, then it must include a specification of the methodic basis for the construction, deployment, and recognition of this action. (This matter is addressed below in "The Fit between the Practice and the Action.")

Although not each of these elements can always be provided in full and compelling form, whether for analytical or editorial reasons, these three elements ought to provide the minimum points of reference for a minimally

This paragraph should make apparent the misunderstanding embodied in some "interpretivist" and "constructionist" treatments, according to which a recipient's understanding of some utterance is definitive of its import and the utterance itself has no "objective" import. This view (and its attribution to conversation analysis) is mistaken on many counts, not least of which is its total subversion of the possibility of analytically specifiable "misunderstanding"; for if a recipient's understanding is definitive, what leverage is there for claiming it to be a misunderstanding? What basis is there for the claim that a recipient failed to understand the import of an utterance? Care must be taken, then, to distinguish between, on the one hand, the professional analyst's undertaking to establish the understanding of some utterance in some interac- tion-which should indeed seek to ground itself in the recipient's displayed under- standing, if possible, and on the other hand the recipient's undertaking to understand the import of some utterance, which clearly cannot be so grounded, for that would presume its own outcome. The recipient's conduct requires engagement with the methods or practices that inform the production of the talk being "analyzed," and the professional analyst's undertaking must make reference to these as well if it is to address both the speaker's and the recipient's parts in bringing off an accomplished, recognizable action. These considerations are critical to an empirical account of action.

satisfactory analysis. The ensuing sections of this article offer, instead of further discursive or programmatic elaboration, possible exemplars of what such a grounding for an empirical account of action might look like.

To revert, then, to where we began, the primary goal of this article is to describe a previously undescribed action that parties to talk-in- interaction may do and something of its provenance and uses. It is a contribution to addressing a lacuna in what should be a central area of inquiry for sociological theory and analysis-the database without which there cannot be an empirically grounded theory of action(s). In this re- gard, I am reporting on what might merit the term "discovery" in this corner of the world of inquiry into the foundations of sociality-I want to describe a possible discovery of an action and something of how that came about.

I will proceed developmentally, from (1) some initial observations and a puzzle that they posed to (2) the collection of a set of potentially relevant materials and the isolation within it of a core collection of fragments of talk bearing on the initiating observations to (3) an initial statement of a candidate solution to the originating puzzle-the phenomenon I mean to be introducing. That phenomenon is then further elaborated with respect to underlying practice and provenance.

INITIAL NOTICINGS AND A PUZZLE

In the course of examining with a seminar (whose contribution I hereby acknowledge) a telephone conversation between two sisters-in-law in their 60s, the two fragments of talk shown in excerpts 1 and 2 were encountered several minutes and topic shifts apart. (The notational con- ventions most important for reading the transcribed data in this article may be found in the appendix. I cannot urge strongly enough that the reader examine the data excerpts closely, and not "read around them." It is not possible to understand the claims and arguments of the text properly without doing so.)

Excerpt 1 (Berkeley 11: 103- 114; simplified) Evelyn has been called to the phone.

1 Evelyn: =Hi: Rita
2 Rita: -Hi: Evelyn : .How [are y'
3 Evelyn:   [I hadda come in another room
4 Rita: Oh:.Uhhuh.=  
5 Evelyn: =Ifee-:1 a bi :ssel verschickert.
6   (0.2)  

7 Rita: W-why's 'a: t,

-

-

8 (0.4) 9 Rita: -t uh you've had sump'n t' drAink. = 10 Evelyn: S =Ihad sump'n t' dri :nk. 11 Rita: Uh huh.

Excerpt 2 (Berkeley 11: 157-168; simplified)

1 Rita: . . . y'know we -: went to the mg :vies. =  
2   =We went to -: -uh :m. uh to -: (m) u(h) :  
3   What's *the name of ( ) ((off-line))
4 Evelyn: (Sh'beau)?    
5 Rita: (** **) (Millbrae) ((off-line))
6 -Millbrae. which is over past Burlingame.  
7 (0.2)  
8 Evelyn: -+ Yeah th[at's 'n far away.  
9 Rita: [(and-)  

10 Rita: S That's far -away. And- there were -two good movies. . .

The initial observations concerned the arrowed turns in each of these excerpts. Two aspects of these turns invite attention and together prompt a puzzle. First, their speakers appear to be doing some sort of agreement; indeed, for reasons that remain to be explicated later, they appear to be doing the particular species of agreement that we ordinarily term "confirmation." Second, they do this confirmation by repeating that which is being agreed with.

The initial puzzle was, Is something special going on here? What are these repeats doing? Or, more precisely, what (if anything) are these repeats being distinctively used to do when employed as the means for agreeing or confirming?

One of the ingredients of this puzzle is that, while some sort of agreement or confirmation seems to be involved in these fragments, the most familiar forms for doing it are not being used, namely, (a) agreement tokens such as "yeah," "uh huh," and the like, (b) pro-termed versions of the talk being agreed with, such as "I did," or "It is," or upgraded or intensified versions of these, such as "It certainly is," or (c) forms that can be more specifically used and heard as ones for conjirmation, forms such as "that's right," as in excerpt

3. Here a teacher confirms a child's observation (at the single-headed arrow) with a change-of-state token "Oh" (Heritage 1984a), a confirmation form ("that's right"), and an intensified or underscored reasser- tion of what is being confirmed.

Excerpt 3 (Philadelphia preschool)

Jason is reporting the mispositioning of a felt marker, which indi-

cates the presence or absence of each child.

1 Jason: Somebody put mine right there.

2 ((gap))

3 Jason: There.

4 ((gap))

5 Jason: I didn'.

6 Teacher: I can't imagine why that ha :ppened Jason.

7 ((gap))

8 Teacher: But I'm really glad that you're here. = .hhh

9 Check and see if there's any down on the

10 bottom that people forgot to hang up.

11 Girl 1 : +That was Alison's job.

-12 Teacher: 3Oh that's right. It -is Alison's job. 13 Girll: A:lison! ((calling out for her))

14 Girl 2 : -Ma :rk is absent.

Whatever their actual relative frequencies, the question remains whether selection among these forms is a locus of order and constitutes a set of practices for differentially implementing particular actions. And, if so, exactly what might someone be doing by confirming in this way-that is, by repeating, and (to make explicit another feature of these two instances, excerpts 1 and 2) by repeating in the next turn, that which is being agreed with or confirmed?

ASSEMBLING A DATABASE

One way of proceeding is to assemble a collection of fragments of talk-in- interaction in which "repeats" are employed, in order to enrich the set of exemplars of whatever it is we are investigating (if, indeed, it is an "it" at all-something not known at the outset of the inquiry). We assemble the collection of repeats "generously," that is, with an ample and inclusive net, including types of repeats which prima facie appear diflerent from our target instances. This will allow us-indeed force us

'There are grounds in earlier work for considering such a possibility, e.g., the afore- mentioned argument by Jefferson (1984, 1993) that "yeah" and "uh huh" interpo- lated into another speaker's talk can do contrasting actions, or Schegloff's proposal (1986, pp. 1245, 1434) that "Did I wake you?" and "Are you awake?" asked by a caller just after the start of a telephone conversation, implement quite different actions in spite of their apparently identical (topical) content-the former a preapology, the latter a preannouncement of bad news.

later on, when we discard these instances, to make explicit just what it is which makes them different from our targets, and thereby potentially to specify progressively just what (if anything) is distinctively going on in the fragments which set us off. As a by-product, we may also be led to encounter other orderly forms, uses, and positionings of repeats-some of them familiar, some of them pointing to other new projects to be pursued.

There is not the space here to elaborate the major initial differentia- tions that discriminate the sorts of occurrences with which we began- the kernel of the "core collection"-from the several other varieties of repeats that an initial canvass of the data revealed. But in seeking ini- tially to decide only which other occurrences of repeats belong with our initial two instances, we are led to make the following discriminations- there will be seven of them-which I will only list and briefly explicate, but not exemplify or explore in any detail.

First, repeats may be done either by the speaker of the first saying or by some recipient of the first saying. Either sort of party may use repeats for a range of different uses, but those ranges of uses are different for the two classes of "repeater." Our core collection contains only repeats by someone other than the initial sayer of the repeated utterance.

Second, as noted in passing earlier, the repeat can be in the turn after the first saying or in any subsequent turn. The repeats that we are tracking are in next turn after the first saying. (The reader will by now have noticed from the first three data excerpts that initial sayings are marked by single-headed arrows, and the turns in which they are re- peated-or could have been-are marked by double-headed arrows.)

Third, repeats by a recipient in the turn after the first saying can occupy or enact at least three different types of sequential position. (1) a repeat may be used to initiate a sequence, most commonly (but not in- variably, cf. Jefferson [1972]) a repair sequence addressed to some prob- lem of hearing or understanding (Schegloff, Jefferson, and Sacks 1977), as in excerpt 4.8

Excerpt 4 (Shiffman :378)

1 S: Did you work for somebody before you worked for Zappa?
2 0: Yehmanymany
3   (3.0)
4 0: Canned heat for a year
5 S: Didya?

Note that excerpts 4-7 are intended to exemplify the sequential positions being discussed in the text, not the phenomenon being tracked in the inquiry or alternatives to it.

  1. 6 0: Poco for ayear 7 T: Ooh when they were good?
  2. 8 0: -+ Bangor-Flunt Madura fer a y- couple years 9 T: 3Bangor Flunt Madura?
  3. 10 0: Bangor Flying Circus 11 J: 0:h yeh I [remember Bangor Flying Circus 12 L: [Ye::ah 13 J: Played at Creation 14 L: Yeh

(2) a repeat may be used responsively, insecond position in a sequence, either in a complementary relationship to the prior utterance, for exam- ple, as an answer to a question (as in excerpt 5), or in a reciprocal or exchange relationship to the prior utterance, as in exchanges of such ritual tokens as greetings or farewells (e.g., the exchange of "bye byens in excerpt (6).

Excerpt 5 (BB gun)

1 Bonnie: C'd yuh bring it to the meeting?
2   (0.4)  
3 Bonnie: the lo :ngest one you [ havle.
4 Jim:   S [ure.]
5   (0.4)  
6 Bonnie:   [An 1
7 Jim:+   [The] longest -one?

8 Bonnie: 3The longest one.

Excerpt 6 (Pink book)
1 B : Okay Vaness. I'm (.) gonna get going.
2 V: Okay.
3 B: Alright?
4 V: Alright.
5 B : See you this evening.
6 V: +Okay -bye -bye. =

7 B: Z=Bye bye.

Or, (3), and finally for now, a repeat may be in, or may embody, a sequentially third position, where it may serve to receive or register an utterance that was produced in second position as a response to some first position utterance (as in excerpt 7).

Excerpt 7 (Upholstery shop)
1 Mike: W'11-what are yer religious beliefs.

2 Vic: +Uh, (0.7) S'preme being.

-

3 Mike: 3Supreme being.

Of the three positions I have mentioned, the repeats in our core collec- tion occur mostly in the second position-that is, as responses of some sort, a few in the third, but not as sequence initiators.

A fourth discrimination made in assembling the core collection of the target repeats is between "identical" and modified resayings of the prior utterance. Now what counts as "identical" turns on how the criteria are drawn; there will almost always be differences of prosody or intonation, for example, between initial and subsequent sayings. For our purposes, we will distinguish between those repeats which, on the one hand, resay all, rather than part, of the prior turn-constructional-unit or clause (often the whole of prior turn), and say it using the same words (except for transformations to satisfy the constraints of speaker-change and deixi~),~ and those, on the other hand, that modify the diction of the turn on its resaying. With respect then to lexical composition, the instances included in our core collection are identical repeats (with deictic and speaker- change adjustments), rather than transformations or paraphrases of the first saying.

Fifth, the repeats of prior turn can be all that occurs in the next turn or they can be combined with other talk. Even the two instances from the core collection with which we began show that either of these may be the case. However, if there is talk of another sort in the turn in addition to the repeat, it turns out that it follows the repeat.

A sixth, and related, feature of those repeats in sequentially second and third position that appear to be doing some form of agreement with prior utterance is that the turn in which the repeat is done may include other forms of agreement besides the repeat or may not, and if there are other forms of agreement, these may precede the repeat or follow it. Of the 60 or so instances in the core collection of what has emerged as the target phenomenon here, the vast majority have no other forms of

This may involve as well changes in verb tense, epistemic marking (such as adding a frame such as "I think" to an otherwise identical repeat), etc. What is excluded is that use of the term "repeat," which includes such transformations as paraphrase, as in Tannen (1987a, 1987b, 1989) and others (e.g., Norrick [I9871 and other contribu- tors to Johnstone [1987]) who employ this looser usage. From the point of view of practices of speaking, it would appear that paraphrase is specifically an alternative to repetition (and to quotation) and involves very different uses for, and procedures of, constructing an utterance. An appreciation of just such differences underlies Elinor Ochs Keenan's (1977) treatment of repetition among children and Heritage's (1985) treatment of news interviewer conduct. See also Gonzales (1995). Running paraphrase and repetition together risks underspecifying both practices and vitiates claims about their functions and uses.

agreement in the turn. Virtually all those that do include agreement tokens have them after the repeat and not before it. The significance of this (and of the preceding paragraph) is that the first such component of a turn can occasion the relevance of turn transfer (Sacks et al. 1974, p. 704) that, if acted upon, can block further components of the turn from being effectively produced. The highly regular initial positioning of these repeats within their turns displays an orientation to getting them said in their speakers' turns, even if nothing else gets said there.

From a quite broad collection of repeated utterances, we have progres- sively focused our attention to those that are repeated, virtually identi- cally, in next turn, by recipients of the first saying, with the repeat em- bodying a second or third position in its sequence. We are looking at repeats that are doing agreement. The repeats either are all of what is in the turn or are the first thing in the turn.

Recall that we have been trying to circumscribe the sort of practice instantiated by our initial instances. It is not that other values on each of the variables I have described do not occur; many of them do. But each of those combinations invites its own inquiry: for each we must inquire whether it defines a natural class of cases, a class of practices of talk, a particular way of doing some action(s). We cannot pursue those inquiries here; we are in search of what is getting done by the practice whose features we have been specifying.

A further, seventh, characteristic of the instances with which we began is useful in discriminating between other repeats meeting the specifica- tions so far adduced that belong in our core collection and those that do not. This characterization makes reference not to the repeat itself, but to the initial saying of the utterance involved-the one indicated by the single-headed arrow in excerpts 1 and 2. The feature is that the speaker of the initial saying, in saying it, is offering a candidate observation, interpretation, or understanding of the recipient's circumstances, current or past.'' And it is that recipient who repeats the utterance in next turn. It is this feature that lends the specific aspect of "confirmation" to what would otherwise be more broadly characterized as "agreements." It be- ing their circumstances that are formulated by the first saying of the utterance in question, they are presumed to be authoritatively informed about those circumstances, and their agreements are thus "confirmatory."

'O There is an element here of what the linguist William Labov described some years ago as "the rule of confirmation," namely "If A makes a statement about B events, then it is heard as a request for confirmation," where "B events" refers to "known to B [hearer], but not to A [speaker]"; see Labov and Fanshel (1977, p. 100). A complementary practice is described in Pomerantz (1980).

Now it may appear, especially with this last quite substantial constraint, that the class of forms that was the target has been exhaustively specified, if not virtually defined. But note that, even with all of these features or constraints, a participant in an occasion of talk-in-interaction who means to respond affirmatively to some interlocutor's proffered un- derstanding of their circumstances can still do so with a common agree- ment token (yes), as will be seen later in extracts (15) and (16) below.

What we have done so far, then, is not the provision of an account of what this practice of talk-in-interaction in particular is doing or is being used to do. The specification of these seven features has served to define a domain of occurrences-a universe of data to which an account of this practice (if it is a practice) should be adequate and against which it can be assessed. What remains is to develop and assess such an account: What (if anything) is someone doing by agreeing with such a prior utter- ance-confirming it-by the use of a repeat of that utterance?

A CANDIDATE SOLUTION: CONFIRMING ALLUSIONS

Let me offer an initial account, which I will then explicate on excerpt 2. In the instances in our core collection, we can note that the sense of what is being confirmed can be seen to have been "planted" (so to speak), or conveyed, in the preceding talk, but to have not been said "in so many words," or in any words "explicitly." The other participant, the one whose utterance is being repeated-the "first sayer," so to speak-has, by that utterance, formulated explicitly an understanding of what the recipient had conveyed without saying. By repeating that "explication," the "confirming party" not only confirms the sense that the utterance proposes is to be made of what preceded. In addition, the confirmer confirms that that sense had been "alluded to," had been conveyed without being said. The repeat confirms the allusion, and confirms it as an allusion.

I use the term "allusion" here very broadly, including diverse usages from "hinting" to such "nonliteral" tropes as metaphor, metonomy, and analogy. There seems little point in discriminating which literary usage some vernacular expression might be taken to exemplify. Nor is it the case that "allusions" can only be to other texts, though this is a very common form of literary allusion.

With respect to the notion of a nonliteral trope, I understand the cur- rent view held by some that no clear distinction can be drawn between the literal and the nonliteral. Although semiotically speaking it may well be that anything can mean anything, it is striking that in ordinary uses of the vernacular, participants do not behave that way. They use the talk, and certain resources in it, to do just these things, and they understand it as doing just these things. They wrest the ordinary from the indefinitely many possibilities and from the possibilities of indefiniteness. That in- cludes the literal and not, the direct and not.

How then do these repeats confirm allusions? In excerpt 2, for example, Rita reports that she and her husband went to the movies in a town whose name she first searches for and then consults her husband to re- trieve (thereby marking it as possibly unfamiliar). She then provides a possible basis for the "difficulty" of the reference, by characterizing the town as "over past Burlingame."

Excerpt 2 (Berkeley 11: 157-168; simplified)

1 Rita: . . . y'know we -: went to the mo -:vies. =  
2   =We went to -: -uh :m, uh to -: (m) u(h) :  
3   What's *the name of ( ) ((off-line))
4 Evelyn: (Sh'beau)?    
5 Rita: (** **) (Millbrae) ((off-line))
6 Millbrae.-which is over past Burlingame.  
7 (0.2)  
8 Evelyn: +Yeah th[at's 'n -far away.  
9 Rita: [(and-)  

10 Rita: 3That's far -away. And- there were two good movies. . .

-

Now there are various ways to "locate" a place geographically in talk-in-interaction, various ways to formulate where it is (Schegloff 1972). Even among place references that formulate location by reference to some landmark, a place that is "past X" can ordinarily also be "this side of Y." By characterizing Millbrae as "over past Burlingame," Rita can be heard to be characterizing it as a place "far away." (And indeed she later explains that she had wanted to see the movie in question when it was in Berkeley where she lives, but that "it was gone. so I knew it was the only place we could go to see it.") But she does not explicitly say "far away"; she does "formulating it as 'far away.'"

Evelyn shows her understanding of what is going on here in the next turn by characterizing the place that Rita has referred to as "far away." Rita's repeat of that, in contrast to "yeah" or "that's right," shows not only agreement with the proposition that Millbrae is in fact "far away," but that that is what she had been doing-conveying "far awaynessn-by the practice that she had employed in her preceding talk. She thereby also underscores that it is not that she is agreeing with Evelyn's characterization of the distance; but that Evelyn was agreeing with her depiction of the distance, conveyed by the form or practice of her loca- tional formulation. This matter of who is agreeing with whom may seem

petty, but parties to conversation do care about it, and it can be interac- tionally quite consequential, as we will see before we are done." In excerpt 8, this operation of the repeat is underscored by the con- firmer's subsequent remark:

Excerpt 8 (Shreve: 2)

Interview with Susan Shreve on National Public Radio concerning her

recent novel.

1 Edwards: Why do you write juvenile books.

-

(0.5)
Edwards: ['s that- b- (0.?) [havling [children? ]
Shreve: [Because I love child [ren] . [I really do :] =

= .hh I enjoy children : , .hh I started writing: (.) juvenile books fer entirely pra: ctical -reasons, .hh

(.)

Shreve: [u- u-
Edwards: +[Making money : : .

-

10 Shreve: 3 Making [money

-

11 Edwards: [yes (( +laughter))

12 Shreve: 3 that--that practical reason hhh

13

14 Shreve: I've been writing juvenile books for a lo :ng .

The interviewer has understood "entirely practical reasons" as an allusion to the need for earning money, and he formulates that as his understanding in the next turn. In repeating his formulation to confirm it, Shreve confirms as well the status of "practical reasons" as an allusion to making money (perhaps a more "tasteful" allusion to it). Her fol- low-up remark-"that practical reasonu-is addressed just to making this tie back to "practical reasons" as the allusive reference-source quite clear, and a bit of laughing together marks their coimplication in intro- ducing the "grubby" subject into a discussion of the arts (see Jefferson, Sacks, and Schegloff 1987).

This fragment, then, should serve not only as another instance in which a veiled or allusive reference, or a conveyed but not explicated sense or meaning, is brought "into the open." It also shows this process itself being overtly recognized and marked by the participants.

However, neither "self-conscious design" nor "intentional" deployment (whatever those terms are understood to mean) is criteria1 to this

l1 See the discussion of excerpt 14 below. See also Lerner (1987, pp. 125-26 nn.4, 5, and p. 2 12).

analysis. In offering as alternative formulations that something had been "planted," "alluded to," or "conveyed without being said," I mean to circumlocute the issue of intentionality or design. "Planted," "hinted at," and "alluded to" appear virtually inescapably to carry that implica- tion in their vernacular usage, and I will accordingly limit their use in favor of "conveyed without being said." This fundamentally common- sense (and in some views culturally specific; see Duranti 1988; Ochs 1988) usage of "intentionality" is of equivocal analytic relevance or use here and in any case is subject to a continuum of experience that belies bold contrasts.12 The following exchange may embody a limiting case-and an informative one. An "assistant professor type" is observed entering his department's main office, to be greeted by a staff member:

Excerpt 9 (EAS :FN)

1 She: ^Heythere! ((Highly animated))
2 He: ^Hey there! ((Echoed intonation))
3 She: +You're back in town?  
4 He: ZI'm back in town.  

One's physical presence is hardly an allusion; it is a palpable embodi- ment of presence. But it is nonetheless inexplicit-conveyed without be- ing said. One's presence has been presented, so to speak, but not formu- lated explicitly. So it is inexplicit conveyance that is being confirmed in this exchange, not any specific trope, whether literary or vernacular. Were the term "allusion" employed here, it would be as a term of conve- nience, to serve as an occasional relief from the more cumbersome "inex- plicit conveyance."

Although agreeing with another's explication of an inexplicit "mes- sage" may not always indicate a prior orientation to conveying it, agreeing by refleating may be a practice that does just that. That, at least, is the proposal being explored here. That proposal is that the repeat is designed to indicate a prior orientation to convey, but that may stand in an indeterminate relation to the orientation that informed the prior talk. The most proximate exchange in which the confirmatory repeat figures, then, can be a contingent part of the larger sequence in which it occurs and doing something in it, and not just an objective indicator of what happened earlier.

l2 Although taken from written discourse, the following characterization offered by the journalist Lance Morrow may well also be relevant to talk-in-interaction, mutatis mutandis: " 'I only realized I was making the allusions afterward. . . . It wasn't entirely advertent. But then I said, what the hell, leave them in' " (quoted in Safire 1991).

EXPLORING THE PRACTICE IN DIVERSE CONTEXTS

It will be useful to linger a bit with a number of additional fragments from the core collection assembled to explore this "candidate phenome- non"; they readily lend themselves to explication along the lines already suggested and give a sense of the robustness of this usage. They allow us to explore some of the forms of inexplicitness that get explicated in these episodes, and, relatedly, the uses parties to conversation may have for talking allusively or indirectly in the first place.

In excerpt 10, a separated couple are discussing the return of their teen- aged son to his father after having had a visit in another city with his mother. The father has called to find out when the son should be expected to arrive via car, only to be informed of a change of plan by the mother:

Excerpt 10 (MDE :MTRAC :60- 1 :2)

1 Marsha: -=He's flying.
    (0.2)
2 Marsha: En Ilene is going to meet im: ,Becuz the to :p wz ripped
3   off'v iz car which is tih ssay someb'ddy helped th'mselfs.
4 Tony: + -Stolen.
5   (0.4)
6 Marsha: 3 -Stolen. Right out in front of my house.

Marsha's initial account for the shift to "flying" is that "the top was ripped off'v his car" (the "top" involved is apparently a cloth top to a convertible model). Marsha herself catches the ambiguity potentially hearable in "ripped off," between the idiom for "robbery" and the physical act of tearing, and extends her turn in order to disambiguate it. The form she chooses is, however, still "nonliteral," a kind of irony-

"somebody helped themselves." The sense of what has happened has, then, been conveyed without being given its "common name." When Tony offers "stolen" as literally what happened, Marsha not only con- firms that that is what has happened, but does so with a form that confirms as well that that is what she was conveying in her prior talk (Schegloff 1988b, 1992b).

With this instance in hand (and the orientation it shows parties to have to idiomatic usages), we can return to excerpt 1 and understand it also to involve an ambiguity.

Excerpt 1 (Berkeley 11: 103- 1 14; simplified)

Evelyn has been called to the phone. 1 Evelyn: =Hi: Rita 2 Rita: Hi : Evelyn :.How [are y'

-

Evelyn: [I hadda come in another room. Rita: Oh:.Uhhuh.= Evelyn: =I fee-:1 a bi :ssel verschickert.

(0.2)
Rita: W-why's 'a: t,

-

-

(0.4)
Rita: +uh you've had sump'n t' drAink. =

Evelyn : 3=I had sump'n t' dri :nk.
Rita: Uh huh.

The ambiguity here turns on the Yiddish word verschickert. It can be used to refer both to mild intoxication roughly translatable as "tipsy," or to refer to a cognate state (if I may put it that way) arrived at without benefit of alcohol, roughly translatable as "punchy" or "groggy." ("A bissel," by the way, means "a little.") Rita apparently understands it in the second sense (possibly because Evelyn says, not that she is "a bissel verschickert," but that she feels that way). Offered as an answer to the question "How are you?" such a response invites the inquirer to pursue the matter and inquire after the causes of the reported state (Sacks 1975; Jefferson 1980; Schegloff 1986), typically via a "what happened" to produce that state. And this is what Rita does ("Why's that?"). But when an answer with an account is delayed, she appears to have reana- lyzed the prior response as intending the first of the senses suggested earlier ("tipsy" being the one actually more literally related to the root of the word, which means "drunk"), and she offers that as her under- standing ("You've had something to drink"). In confirming with a re- peat, Evelyn not only verifies the supposition as to how she came to be in the state she is in, but invokes the presence of that meaning in her own immediately prior talk.

The locus of inexplicitness, and the focus of explication, need not be concentrated in a specific expression, word or phrase. It can somewhat more diffusely inform a discursive unit such as a narrative. In excerpt (1 I), Vera is engaged in reporting on the visit of her children and grand- children, one consequence of which is making audiotape copies of various of her music records for them. (If this one "reads" differently, it may be because it is taken from British data, as are other excerpts whose identifying number is marked with an asterisk.)

Excerpt 11" (Rahman B :2: JV[14]2; #12)

1 Vera: Any rate ehv eh ah'm -tapin' fuh -Bill no :w:

2 etheh little devils with the ta: pes you kno :w

-

Jenny: Oh: they pull -the tap [es (puhhaps do they)]
Vera: [Ah:nd em:]
Jenny: Mhm, =
Vera: =Jean said -
oh I didn't knew- (.) know you had the

-

Tapestry reco :rd. of Carol King's which is a luvly

-reco [:rd Jenny] Jenny: [Ah : she fi] kes[thaht does she Vera: [So-Jenny: Mm,
Vera: I: s'd -why-::so: she s'd oo ahr James : -broke it.

(0.3)
Vera: -So ah s'd oh well ah'll do y'anothuh

-

one. yihkno[w, ahn also they wa: nt eh: m (.) =

-

-

Jenny: [Mm-:, Vera: =Neil -Diamond, -Bill listen'to it heahr 'n'e sid oh -thaht's

-smashin, =
Jenny: =A[h t h e y : [like [the Neil [Diamond, ]
Vera: ['n ah- -sai [dah- [.hh [ah'll tape] - it
  [fohr you] =          
Jenny: [Ye:s.]
-
         

Vera: =en John Denver they want doing. yih [know- ,I Jenny: [ehhhh]= +heh-so yuh -busy at it [again.] Vera: 3 [So-:ah] 'm -busy at it
agay: [n. yih [nuhhh

-Jenny: --

-

[.hhhhh[hu:h .uh

Initially Vera just reports that she is taping, but then she begins a series of brief stories, in each of which an indirect indication by her daughter or her son-in-law elicits an offer by Vera to make a tape. Note that Vera reports Jean's remarking on her having a certain record (lines 6-7)' which Vera shows herself to have treated as a "hint" of some sort by responding with "why" (line 12). And Bill calls another record "smashin" (lines 17-18), which prompts a separate offer to tape. Then there is a third artist that "they want doing" (line 23)' and the sense is conveyed of the cumulative burdensomeness of the undertaking. This sense is extracted and crystallized by Jenny ("so yuh busy at it again") after the third of the "assignments" has been mentioned, when they have become more than occasional-and is confirmed by Vera, both as to content and as to prior implication.

Excerpt 11allows us to see clearly that some telling may be constructed by its teller, and/or be taken by its recipient, to embody and/or to reveal a tack that the teller is taking to the tale, some stance being taken up, or some action being done. When a recipient makes that explicit in the uptake, the teller can confirm both the particulars of the uptake, its "propositional content" so to speak, and that he or she was engaged in such a "project." But note that by doing it in this manner, Vera can avoid any actual, explicit complaining and still have her burden regis- tered and appreciated. There can be, then, grounds for talking allusively and for practices that deal with such talk. (So also in excerpt 10 above is there a specific basis for talking allusively; that is a common practice for organizing the delivery of bad news, a theme discussed in Schegloff 1988b.)

Excerpt 12 offers another environment in which there are systematic grounds for allusive talk and accordingly systematic grounds for usages that are addressed to allusive talk.

Excerpt 12 (Brun-Cottan [1989, chap. 5, p. 21) M is returning F's call; first several utterances not recorded

1 M: ...y ou:::
2 F: -~'mwell:thank you.
3 M: . hh(d) you sound like you have a jo :b hh
4 F: -NoIdo:n(hh)(hh)'t. .hh
5 --M: +Oh you're not tha: t wel[l.
6 F: 3   [I'm not that well, . hh
7 M: Oh: goodness: : .  
8 F: I-before we get into anything else. . .

The women in this episode are part of a mutual-help network of women all looking for work as assistant producers in the Hollywood movie industry. Note that M is returning the call, which is to say that F called her in the first instance. F may, then, have had a "reason for the call," one important such warrant being "news to tell." When F responds to a "howaryou" with an emphatic "I'm well" (line 2), M hears in it a possible allusion to the occurrence of something that would make F emphatically well-a job.13

l3 It is true, as a number of people have commented, that individuals may differ in their deployments of the conventional responses to this inquiry. Still, personal styles appear to be grounded in their relationship to the common usages in which "fine," "okay," andlor "good" are the "no news" mild positive responses, with superlatives and disaster on either side. One may know about a particular other that, when they

As noted in the discussion of excerpt 1 above, this is a sequentially strategic place for such an occurrence. Answers to "howaryou"s in the openings of telephone conversations are specifically a position for pre- emptively launching first topics (Schegloff 1986), and particularly for first topics that announce good or bad news. Raising first topics of these sorts preemptively can be a way of avoiding awkward conjunctions of the good and bad news of the respective parties. Alluding to the news is a strategic way of proceeding in that it allows a recipient who has claimably higher priority news-either good or bad, but especially of the opposite type than otherwise seems about to be told-to decline to pursue the matter and initiate their own first topic instead.

But in the episode from which excerpt 12 is drawn, M does pursue the matter, and does so in a characteristic way-she offers (line 3) a candi- date understanding of what she takes it the other may have meant to convey. As it happens she's wrong, and F tells her so (line 4) by denying the formulation of her possible news. M then offers another candidate understanding (line S), this one designed to register a more careful cali- bration of the well-being that F's initial response to the inquiry conveyed and displaying that it was that utterance and its articulation that was the (inexplicit) source of her prior candidate understanding. Now F means not only to confirm the degree of well-ness that M has ventured, but also that this is what her "I'm well" had been designed to convey. Rather than doing the confirmation by a form, such as, for example, "No, I'm afraid not," she does the confirmation by repeating the pro- posed explication of what she had conveyed allusively earlier, "I'm not that well."

LESS TRANSPARENT INSTANCES

With some clear instances in hand, we can examine occurrences that may initially appear problematic. These, it turns out, are better examined not forward in time (from the inexplicit conveying to the explication) but in retrospect, starting with the repeat's suggestion that an inexplicit message had been conveyed and then seeking it out in the earlier talk.

As excerpt 13 shows, the line between what was said and what was conveyed, between the "explicit" and the "implied," can be drawn very finely indeed.

say "fine," they are "great," or they are "terrible," but the distinctiveness of this knowledge is based on the otherwise operative understandings of that response term. In any case, in excerpt 12, M shows that she hears F's response to be an "upper" by the terms of her pursuit of it, even though that hearing turns out to be incorrect.

Excerpt 13* (Drew: Student [A:I: 7 : 10-1 11) 1 B : but e- I mean he was in a -terr :ible state, because they- -W ( )

2 they took -away the roof of his mouth an' his cheek bone an'
3 all the inside of his nose.
4 (0.8)
5 B : an' his --one eye they've sewed ^up.

6 (0.5)
7 P: Ae..h....

.. ....

8 (0.5)
9 B: ((cough)) (0.4) yeah and th [en-

10 P: [so-he's in quite a state
11 [reall [y *is he.*
12 B: [Oh he [is, an' of course he --can't eat: you see,

13 or -swallow so he's just having these -fluids and things,
14 (0.9)

15 P: ^oo::h[:
16 B: [he's --so thin that I think 'tis really (.) -quite
17 unbelievable you know that anybody could be so thin.

18 P: O::^h:[:^
19 B: [and still sort of (0.3) sit u(h)p if (h) you kn(hh)ow
20 whatI mean.

21 P: ^o..h:.. : I:

.. nGod[ 22 B: [ 'tem] 23 24 P: +He can't eat any solid food at all [:then?

25 B: 3 ['e can't 26 3eat any solid food at --all at the moment. 27 P: oAo::hAmy:

Initial examination of this episode can use the exchange at the arrows to provide prima facie grounds for treating this as an instance of the phenomenon we have been trying to describe. P has offered a candidate understanding of the situation of the patient being described in gory detail by B, and B has confirmed with a repeat. On the account so far offered of such repeats, we are invited to see that B had earlier alluded to, or somehow otherwise inexplicitly conveyed, just the understanding that P here proffers.

When we try to do this, it may appear that excerpt 13 is not a good instance of the phenomenon, is perhaps counterevidence, or an exempli- fication of some other use of the confirmation-by-repeat practice. For it appears that B had not conveyed this sense inexplicitly earlier, but had said "can't eat" in so many words (lines 12-13). But we should then note that what B had said earlier was that he (the patient) "can't eat . . . or swallow . . . so he's just having these fluids." This could be taken literally, to mean that he cannot ingest orally, and "having fluids" then refers to IVs, or intravenous feeding. Alternatively, "eat" could be taken more narrowly, as a contrast to "drink," with "having fluids" then referring to that ability to drink and implying that "can't eat" refers only to "food"; note that in this hearing it is an implication that "can't eat" refers only to food. P's eventual candidate understanding is that he "can't eat any solid food at all," which makes explicit one understanding of what the reference to "having fluids" now turns out to have done inexplicitly earlier, which B's repeat confirms.

There are two points to be registered and stressed here. First, we must recognize that, for understanding the production of the talk, the relevant senses of "explicit" and "implicit" or "inexplicit" are not those of the professional student of logic, semantics, pragmatics, or rhetoric, but those of the participants themselves; and the level of detail worth working out is given by the concerns of the participants' lives, not those of profes- sional students. (It is their mutual acquaintance they are talking about, and the extent of his permanent disfigurement and incapacity.) We are here getting at the members' then-relevant sense and practice of explic- itness.

Second, looking at the data in this way (i.e., back from the confirma- tory repeat to the preceding talk to search out what is now being claimed to have been made allusively available earlier) can provide "profes- sional" analysts a methodological resource for the analysis of a class of especially challenging occurrences-those in which speakers design their talk, at least in part, to mask what they are doing from coparticipants. To the degree to which they succeed, they challenge analysis and compelling explication by professional analysts as well. Among these special chal- lenges to analysis are deception, lying, pretense, simulation or covering up. When done fully adequately, they will in the nature of the case appear to defy analysis for what they are (cf., e.g., Jefferson 1986).

Allusion provides a milder version of the same phenomenon. With it speakers do indeed avoid explicit saying or doing. At the same time, however, speakers do introduce the import of what they are about into the interactional occasion (as in Vera's story in excerpt 11 above). But professional analysis does not settle for implicit "senses" of what another has introduced, as "lay analysis" by interactional coparticipants may; it seeks explicit statement of what has occurred and how, and it is just this which an alluder, an inexplicit conveyor, has not done, has on occasion specifically not done. It is useful then to find a post hoc form of talk which appears to provide a sort of internal grounding for such analysis.

For, if it is the case that confirmation-repeats can do what I have pro- posed they do, they can serve for the occasions in which they occur as data-internal verification that inexplicit conveyance had been oriented to by the participants. The study of allusion and inexplicitness can thereby be rescued from allusiveness and inexplicitness of its own and can be grounded in a corpus of in situ participant-attested occurrences.

EVIDENCE FROM AVOIDANCE AND NONOCCURRENCE

I want to offer one last sort of evidence and display of the practice of agreeing-by-repeating for confirming allusions that builds from its nonde- ployment rather than from its deployment. One is exemplified in avoid- ance-an observable withholding of the practice; the other is an analyti- cally grounded claim of relevant nonoccurrence.

Avoidance

One often compelling sort of evidence for a claimed practice, orientation, or organization in talk-in-interaction is the eventfulness of its absence, or an orientation to avoiding it as well as achieving it. If some practice of talking is used to do some action, then there will be occasions on which a participant will undertake to avoid that action, and that will involve avoiding that practice of talking.

Excerpt 14 offers a possible instance of such an occurrence. Two stu- dent couples are having dinner together. Vivian and Shane have just told about a cheap lobster lunch that Shane had eaten that day, which he had not mentioned to his "friend" Michael. Michael goes on to recount (ironically, or half so) another incident of less than generous conduct on Shane's part.

The possible selfishness of Shane has just been (perhaps ironically)

alluded to Shane: [.huhhh u [hh

Excerpt 14 (CD: 11: 23-25)

1 Michael: --Shanethinks he's doin'me a big favor 'm riding th'bike 'ee
2   siz 'ey -I'm unnuh gittadurink.  
3   (0.9)  
4 Nancy: m- (h) mn -Oh (h) o [N-hn -  
5 Shane: --[heh heh  
6      
7 Michael: ha[hhlf yihkno[w? I give yih - I mea[n
8 ( 1: [eh ^heh [ [

-

Nancy: [*mm-hmm-hm hm ^hm h[m*

-

Michael: [^Su-:h ^'ere ee am ah'm -(.)

bikin' [a half -hour] [('n evrything) 'n I] =
Shane: [.hahhh heh] [h u h h -u h h u h]
Michael: =git -awv'm thristier'n shit,

(1.2)
(Shane): [(*hheh ih huh*)

-
Michael: [Spect this big --mo :njo drink [tih come yih]
=

Shane: [.hhh ^huh]=

Michael: =[know,]
Shane: =[h u]h
Nancy: Mm-h

-[m :
Shane: [.hhh

(0.7)
Michael : Fuckin -dixie cu [p .

Shane: [.-huhh .huh [.huhh .hu]

-

Michael : (This) [bi:g. yi] [h na: h? Nancy: [nuhh

(0.8)
Nancy: Sha(h)a(h)[:ne.

-

Shane: [.hahh-hhu-uh .hhh(h) 'At's w'd'ee put (it in) I(h) said wait -a minute. Vi :c. {nh .hh (.)/(0.6)) Isaid (ziss is) f 'r d'two peo -:ple .

(1.7) Shane: An (.) he s'd (weiy) muh man -at's w'tche -ge ::t,

Nancy: mm hm hm

(0.8)
Michael :*Eez gitt'n li'l cheap there. ha : [h.

Shane: 3 [Y:ah. I

-

=f think he's gett'n a li'l bit uh ti: ght .wid is eh -::p

(1.4)
Shane: B't I-had the big glass. . .

In Michael's story, he and Shane were exercising when Shane left to get something to drink, promising half to Michael on his return. Michael then reports exercising vigorously in anticipation of a substantial drink, only to have Shane reappear with a tiny cup of potable. At this point, Michael's companion, Nancy, chimes in with an admonishing tone (line 30).

In his response, Shane takes the tack of reporting "just the facts" (Pomerantz 1978, 1987). He offers no characterization of the parties or the events, attributes no blame, offers no assessments or excuses. He reports what the service person, Vic, served him, what he said to the service person on behalf of a larger serving, and the rebuff he received in return (lines 31-35). When, in response to this account, Michael says of Vic the service person (line 38), "He's getting a little cheap there," this is Michael's (the complainant's) blame attribution.

Were such a characterization offered by Shane, it could be seen as self-exculpating and self-serving. Its production by Michael is precisely a redirection of the blame-which he had appeared to be directing at Shane-now to Vic (the service person). This redirection of blame away from Shane as an action of Michael's would be subverted if Shane were understood to have himself planted this sense of Vic having been at fault for the episode, with Michael merely making it explicit, which is what a repeat-confirmation by Shane of Michael's characterization could be heard to claim.

Note then that Shane appears speci$cally to avoid the repetition that would implement that practice. His turn repeats the frame from Mi- chael's turn, "He's getting a little . . . , " but just where the descriptor "cheap" would go, there is first the insertion of the word "bit," then there is a moment of hesitation with a display of a bit of a search ("uh"), and finally the selection of a sense-retaining alternative to "cheap," namely "tight." Shane has come to the very verge of a possible allusion- confirmation before veering away.

There is little doubt that Shane's construction of his account is indeed built to provide for an analysis of Vic as cheap and as the culprit of the tale-most obviously by the "no quarter given" response that he attri- butes to Vic. And very likely this is reflected in his (Shane's) initiation of an utterance form that would confirm that Michael had picked up just what he-Shane-had meant to convey. But there are good grounds for the diversion of this utterance form before it is brought to fruition. The overt blaming of Vic is left as something that Michael has done and Shane has agreed with, not as something that Shane has planted for Michael to pick up. Here then is one way in which an issue mentioned much earlier in this account-the apparently petty "who is agreeing with whomn-can and does matter. And the talk is adjusted midcourse to achieve a particular stance on just that issue by avoiding an already-in- progress allusion confirmation (see also Mandelbaum 1993).

Nonoccurrence

In other episodes, we find sequential and interactional contexts with the features that elsewhere are found to occasion allusion-confirmations (arguably "environments of relevant possible occurrence"; see Schegloff 1993), but in which no such confirming repeats are produced. In such instances, there has been some inexplicit conveying or allusion, the recipi- ents of which subsequently make explicit their understanding of what had been conveyed. In the turn following that candidate understanding, the stage is set for a repeat that would confirm both the sense and the prior indirect conveyance of what is being checked. But in these exchanges the prior conveyor eschews the confirming repeat and employs some other ("single-barreled" so to speak) form of agreementlconfir- mation.

These instances might initially be taken as another version of the test-by-avoidance, explored in excerpt 14, but they are different and somewhat weaker in analytic warrant. In excerpt 14, a confirmatory repeat was apparently underway and was then suppressed at the last moment. There, one is warranted in speaking of its absence or avoid- ance in the strongest sense, because its relevance has already been introduced by the participants-by their very launching of its deploy- ment. But the practice of confirming allusions by repetition of their explication has not yet been shown to have that order of relevance for parties to talk-in-interaction that would make it relevantly, event- fully absent if not done at all. Still, in excerpts 15 and 16 below we can see that the conditions for embodying this practice were present by juxtaposing them to other, analytically comparable episodes in which confirmatory repeats were employed (a sort of "paired compari- son"). We can then ask if there is any interpretive or analytic relevance to the practice not having been invoked though "eligible," and we can ask whether this "eligibility" would have been available as an interpretive resource to the participants and thereby have made avail- able to them as a noticeable absence-and therefore an interpretable and consequential one-the nonactivation of the practice of a confir- matory repeat. In excerpt 15 two women are discussing B's recent trip.

Excerpt 15 (SBL 2 :3 :2; Sacks [1992, vol. 2, p. 1401)

1 A: Didje have a nice time?
2 B: Oh, wonderful. -
3 A: Goo::d,[good.
4 B: [just -wonder [ful.

5 A: [Where'dju -go : : .

6 (0.6)

7 B : We were in northern California, up- (0.2)

8 weh (hhh)- (0.4) way up in the mountains too.

9 (0.4)

10 A: Oh well -we wen' up there oh : : about thr- . hh

11 I'd say about three weeks ago we was up at

12 Maripo:sa, . hh[hh

13 B: [Uh huh

14 A: an' up in the Mother Lode country en we

15 [wen' all through those -ghost tow : ns.

16 B: [( 1

17 B: Oh : : I see, Well we were up uh . hh intuh Red

18 (0.5) Red Blu:ff?

19 (0.4)

20 A: Oh::.

21 B: ( [ )

22 A: + [You wen' up further then.

23 B: 3Yes.

24 A: Uh [huh,

25 B: [Uh huh,

In response to A's inquiry about where she went, B responds by locating the general area ("northern California"), and then with a formulation that "does" referring to extended penetration into the mountains-"way up in the mountains" (lines 7-8). Keying on the reference to northern California, A then counters with her own recent trip to the area. Her references, however, display an orientation less to nature and the moun- tains than to conventional tourist destinations ("the Mother Lode coun- try," "those ghost towns," lines 14-15). As A has used B's account as the occasion for her own, the comparability of the account and their events has been potentially activated and A's trip is taken as offering a version of "way up in the mountains." To this, B then counterposes the extent of her own trip, offering as the point of comparison the name of the place defining the area she reached (lines 17-18). Note about it that it is offered as possibly less readily recognizable (and hence, perhaps, remote)-there is a pause and a bit of a search before it is articulated, the name itself is not delivered fluently, and when delivered it is packaged with an upward intonation, as if to ask whether A recognizes it or knows where it is (in an operation not unlike the "try-marking" discussed in Sacks and Schegloff [1979]). At the same time, it is offered for its tacit display that she (B) had traveled higher up into the mountains than had A (and that "up into the mountains" was the relevant orientation of her trip).

In many respects, then, this episode is similar to excerpt 2, in which one of the participants also characterizes a place with a practice that "does formulating it as far." In excerpt 2, it may be recalled, the recipi- ent responds by extracting the point as "That's far away"; here in ex- cerpt 15 the recipient does something similar-she extracts the upshot (if I may put it that way), "You went up further then." In excerpt 2, this was confirmed with a repeat, employing the practice already explicated. In 15, in an arguably comparable sequential environment, and with con- firming agreement also being done, the action is implemented with agree- ment tokens-"Yes, uh huh." B does not confirm, then, (Dare we say that she declines to confirm?) that her earlier formulation had been de- signed to convey indirectly the comparative reach of their respective trips, and she avoids thereby the "one-upsmanship" that this might have been heard to embody.

But there can be other grounds for not employing a confirming repeat in an environment apparently primed for it. In excerpt 16 a request is turned away by recounting a tale that conveys the difficulties the request entails without saying no in so many words. When the requester makes the upshot explicit, the possibility for a confirmatory repeat presents itself.

In excerpt 16 Madeline has called a second time asking to talk to her friend Gina and is told by Gina's mother, Marcia, that Gina will not be back home until later.

Excerpt 16 (Madeline 2, 1 :9-38; simplified)

1 Madeline: okay. =I was just wundering y'know .hhh (0.3) could
2   (.) d'you think you might (.) wannt rent (.) you know
3   like the bottom part a yer :(.) g'ra: :ge like to
4   -me : fer a whi :le, "a sump'm like that."
5   (0.3)
6 Marsha: wul[l-
7 Madeline: [(I think [ 11
8 Marsha:   [oh- you mean for] living in :madeline?
9 Madeline: ye :ah    
10   (0.3)    
11 Marsha: .hh It's just
-
   
12   (1.0)    
13 Madeline: no :t possible. =h[uh?

Marsha: [ye:na tri :ed that at one

:ah we- gi ti : [me . =b't-] Madeline : [ -I: re] member. =she was doin that o [nce. Marsha: [.hhh we could not get it uh :: (0.4)-clo-: :sed.

(0.3)
Madeline: oh : : (y') couldn't -close it?
Marsha: ye :ah.
Madeline: -oh : =

Marsha: and it jus' (wz) -free :zing. ='n then we bought a -heater : . ='n it was -still freezing. = we bought

-['n -

another heater it was] still freezing. . hh [h Madeline: + [(o (h) h @))I [no: : +it didn't work ou :t -did it.

(0.7) Marsha: =f no : :. an-=an :d uh hh (.) in the mean time I'm afraid

that the insect lif :e has moved ba: ck. =

If we continue to exploit paired comparisons between instances, as we did in juxtaposing excerpt 15 with excerpt 2, then excerpt 16 invites juxtaposition with excerpts 11 and 14. In excerpt 11 Vera could be under- stood to have built into her telling of a tale a stance, a point of view, that her interlocutor could extract and offer for confirmation; when offered by Jenny, that stance was confirmed by Vera with a form that confirmed that it had indeed been conveyed by the telling. In excerpt 14 Shane recounted the tale of getting something to drink in a fashion that con- veyed the unyielding stinginess of the service person and barely escaped confirming that he had conveyed this in the telling. Similarly in excerpt 16, Marsha's telling (lines 14-18, 23-25) is designed to convey the nonvi- ability of Madeline's proposal by reporting its nonviability in the past. (Note that the telling is launched in response to Madeline's anticipation of a rejection, "It's just not possible, huh?" at line 13.) Indeed, it is so overbuilt to convey this that a confirmation by repeating Madeline's proposed gist, "It didn't work out, did it" (line 27) might well have been hearable as a dry, cutting sarcasm. Instead, Marsha confirms with an agreement token and goes on to add to the account additional difficulties as grounds for not accommodating the request.

Excerpt 16 provides an occasion for taking up briefly another ingredi- ent of this "confirmation venue" and this confirmation practice. Confirming repeats can convey a range of affective stances toward the other and toward the matters being addressed in the talk. They can convey satisfaction, congratulation, mockery, and the like-even perhaps dis- covery, as when the upshot proposed by the interlocutor reveals to the teller a tack they had taken of which they were unaware. In the excerpts examined earlier in this article, for example, a note of resignation is conveyed in excerpts 2 ("far away") and 12 ("not that well"). Just as Shane in extract 14 veered away from doing a confirmatory repeat and thereby avoided the appearance of having himself deflected blame for the incident being discussed, so can we entertain the possibility (though no comparable analytical basis has been developed for it here) that Mar- sha in excerpt 16 eschews a confirming repeat by reference to the cutting sarcasm it could represent after as broadly conveyed a stance as she had portrayed.

There is a suggestion here, then, of yet another level of action, of practice, and of analysis that can be relevant to the phenomenon being examined. Confirming by repeating is a practice, I have argued, for confirming that the candidate understanding distilled by the other had indeed been previously conveyed inexplicitly. Now we are noting that the doing of this action can itself become the vehicle for another, in a fashion even more tightly woven into the fabric of the local context. If the inexplicit conveying had been subtle, sophisticated, challenging for this recipient, then a confirmatory repeat can appreciate the other's per- spicacity. If in the face of a broad and heavy-handed "indirection" a recipient still undertakes to check it out, a confirmatory repeat can come off as sardonic and patronizing. But aside from its relationship to the preceding inexplicit conveying, other resources can be involved in achiev- ing these effects, notably those of prosody and intonation, whose prac- tices at the present time still elude compelling analytic specification. For now this additional layer can itself only be broadly indicated, but cannot be explored with any empirical or analytic detail. On the horizon, how- ever, appears yet another lamination that an empirical account of action must be able to deal with.

Finally, it seems cogent to remark that in excerpts 15 and 16 it is virtually certain that nothing of interest would have been seen at all were we not already familiar with the practice of confirming allusions and its environments of possible occurrence. Here then we may have some of the most distinctive fruits of inquiry in rendering what would otherwise be invisible visible in its very absence.

THE FIT BETWEEN THE PRACTICE AND THE ACTION: A SKETCH

How are we to understand the fit between this practice-repeating more or less identically the candidate understanding of one's prior talk or other circumstances proposed by another-and the action which it appears to accomplish-confirming both the candidate understanding and that its gist had been previously conveyed inexplicitly by the one who now con- firms it? What suits this practice for doing this action? How does it work? There is space here for only a sketch.

Such an account begins with the observation that parties to talk-in- interaction may also apply the practice of identical repeats to their own prior talk.14 For example, they may do this when the start of their turn becomes implicated in overlap with other talk-whether a simultaneous start by another incipient speaker or an extension of the preceding talk by the prior speaker. As that overlap is resolved, the speaker of the "surviving" turn may "recycle" the beginning of their turn, now in the clear, as does Kathy at the arrowed turn in excerpt 17.

Excerpt 17 (KC-4 :07; taken from Schegloff [(1973) 19871) 1 Rubin: Well thee uhm (.) (a paz) they must have grown a culture. 2 (0.5) 3 Rubin: You know, (.) they must've I mean how lo- he's 4 been in the hospital for a few day: s, right? 5 {(1.0)/. hhh} 6 Rubin: Takes a[bout a week to grow a culture,] 7 Kathy: -[ I don think they grow a ] I don think 8 +they grow a culture to do a biopsy. 9 Rubin: No : : . (.) They did the biopsy while he was on the table.

One import of this practice is a display or claim by the speaker that what they are saying now is what they were saying before (in the overlap), rather than (for example) addressing responsively what the other party was saying during the overlap. Note that the resources of language make it possible to say the same thing (or do the same thing) in diffeerent words (see, e.g., Schegloff [I9761 1984, pp. 40-41); saying the same words ap- pears designed specifically to underscore that what is being said now is the very thing that was being said before.

Resaying in the same words to show/claim that what is being said now is the same thing that was being said before is not limited to use by the "victor" in the environment of emergence from overlapping talk. It can be used by the apparent "loser" after the victor has completed the turn they were producing. In excerpt 18, Kathy and her husband, Dave, are commenting on a weaving of Kathy's which their guest, Rubin, has admired.

l4 See the first specification of the domain being explored in this article, at p. 177 above.

Excerpt 18 (KC-4 :36-7)

1 Kathy: So once I'd set up the warp, i' w's very simple to

2 jus' keep- jus' to weave it.

3 ((pause))

4 Kathy: Youknow [( )

5 Dave: + [(But listen tuh how long 11

6 Rubin: [In other words, you gotta string up thee]

7 you gotta string up thee colors, is that it =

8 Kathy: =[Right ]

9 Rubin: =[in thee] in thee [warp. ]

10 Kathy: [right] right.

11 Dave: 3(But listen) tuh [how long it took to put in the] =

12 Kathy: + [A n d t h e n each weft-]

13 Dave: =the the warps [(though)

14 Kathy: 3 [And then each weft y'know then I did

15 I s- my warp was. . .

Note Dave's reuse of the same words at line 11 after "losing" to Rubin in the overlap at lines 5-6. More striking still is Kathy's reentry at line 14 after losing to Dave at lines 11-12. She begins her reentry by reusing the same words ("And then each weft") to establish that what is being said now is the same thing that was being said before . . . and then, having accomplished that, she drops that way of talking in favor of another.

Nor is this practice of repeating one's own prior words limited to proximate environments of overlap and its resolution. Across various spans of talk and various routes by which a speaker's talk comes (in their view) not to have been addressed, they may use a repeat of their own earlier words to reintroduce their earlier contribution into the talk, and to do so in a manner that claims that they are doing just that-doing/ saying again what they didlsaid before, and that they (and it) are not being responsive to what may have transpired in the interaction since the previous saying.

If repeating one's own prior words is a way of doing "what I'm saying now is what I was saying before," why is that practice not employed in the contexts being examined in this article? That is, why does the speaker whose prior talk has been formulated by its recipient in a candidate understanding not repeat that prior talk as a way of showing that they are saying now what they said before? In the turn after a proffered candidate understanding, a repeat of the previously articulated saying (the inexplicit saying) would be declining to confirm when confirmation had been solicited. It would, in effect, reject the candidate understanding proffered by the recipient and reassert the adequacy, indeed insist on the necessity, of the way in which it had been said before.

In excerpt 19, for example, Colonel Lehroff, the director of civil de- fense, is calling the home of the manager of the municipal truck yard- one "Phil."

Excerpt 19 (CDHQ, 15; Openings, 299; from Schegloff, 1992b, p. 1322) 1 Phil: Hello? 2 Lehroff: Phil! 3 Phil: Yeh. 4 Lehroff: Josh Lehroff. 5 Phil: Yeh. 6 Lehroff: Ah: :what've you gotten so far. Any requests to 7 dispatch any trucks in any areas, 8 Phil: +Oh you want my daddy.

-

9 Lehroff: 3Yeah, Phi [1, 10 Phil: [Well he's outta town at a convention.

As it happens, the phone has been answered by the truck manager's son, whose name is also Phil and who, upon hearing the caller's identification (line 2), understands himself to have been identified and ratifies the "rec- ognition" (line 3). The mutual identifications being completed, the caller proceeds to the first topic, a topic recipient-designed for the answerer whom the caller thinks he has identified and recognized and whom he had called to raise this very topic. But raising that topic reveals to Phil that his understanding of Lehroff's first turn was wrong, and he offers (line 8) his now-transformed understanding of it. When Lehroff responds (line 9), he first confirms the candidate understanding with an agreement token and then repeats his own earlier turn whose understanding he has just confirmed. But in this sequential context, that repetition has the effect of confirming the prior saying with a protestation of sorts. Its effect is to underscore that he had said just the right thing in thefirst instance, one which should not have been a source of trouble. So, although the repeat of his own prior talk does the work of saying here what was said there, when done in the face of a differently formulated and articulated account, it insists on the form of the prior saying and does not simply confirm the proffered understanding of it. The latter is done by a simple agreement token.

In an environment in which there has been an allusion or inexplicit conveying in the prior talk of which the recipient has proposed an explicit upshot, an identical repeat of the prior saying by its speaker amounts to an insistence on that mode of saying in the face of a proposed alternative, and thus a possible rejection of the alternative-the explication. But the earlier point remains: a way of identifying a later saying with an earlier one is to use the same words. Doing an identical repeat, but of the explica- tion, confirms the understanding that it proposes and does so with a practice otherwise employed for marking the thing being now said as the thing previously said.

PROVENANCE

Virtually all of the instances so far presented of the phenomenon and practice of which we have been developing an account come from occa- sions of interaction that could be vernacularly characterized as between intimates-between members of a family, long-standing friends, and so on. And the occurrences could be plausibly understood as contextually specific along these lines; for example, it is to the intimately related (it might be proposed) that speakers undertake to convey something inex- plicitly and confirm such undertakings when it becomes relevant to do so; it might even be taken as a mark or indicator of intimacy, or as a move to invoke or establish such a relationship.

This turns out not to be the case. It is striking that the practice de- scribed here and the action that it implements are found across a range of settings and social relationships from quite intimate ones to rather formal and public ones, and across variation in other variables (such as class, gender, culture, power, etc.) conventionally understood to embody facets of the decisive relevance of context for interaction. In addition to the earlier exemplars (e.g., excerpt 8, which is drawn from the mass media setting of the news interview with the novelist), only a few are displayed here to provide a sense of the ways in which the practice appears in this diverse range of social settings-from a teaching case conference in a major university medical school hospital (excerpt 20 be- low), to an editorial conference at a major metropolitan newspaper (ex- cerpt 2 1 below), to a criminal courtroom in session (excerpt 22 below), and a commercial transaction at an outdoor farmers' market (excerpt 23 below). If the analyses of the several segments discussed here (and others not displayed in order to conserve space) are on target, then this practice of confirming allusions has been successfully employed in each instance by the participants in that occasion, in that setting, in that relationship, and were so understood by the other participants, across variations in these parameters of context and participation. Consider then:

Excerpt 20 is taken from a case conference in a major university medi- cal school and hospital. About 45-50 staff members are in attendance, including senior faculty, residents, interns, and so on. A resident in pedi- atrics ("Pr" in the transcript) has been called upon to launch the presen- tation of a report on a recent case of great interest. She has been reporting the presenting symptoms and the previous treatment history. Just before the beginning of the excerpt, she has inadvertently juxtaposed two incon- gruous observations that have elicited an outburst of laughter, and she is searching the room for her supervising hemotologist (LLM2" in the transcript). He has just arrived (late) at the conference, and he responds to being spotted by Pr with "Don't blame me" (line 5).

Excerpt 20 (Boyd, leukemia case conference, 1 :29-2 : 14)

1 Pr: -Hey I'm doing this all by myself. Where's my- -Oh.

2 There he is. (.) Ha hah .

3 Grp: Ha ha ha ha huh huh huh.

4 Pr: Ha hah hah hah hah hah:

5 M2 : (Don't blame me for it. )

6 Pr: So.(.)U:m(.)

7 M2 : J [ a-] (Jackie) [Stop there =

8 Pr: [No.] [Oh.

9 M2 : =There's almost an -infinite number of things tuh

10 dis- discuss on- on this case between thuh new

11 technique of dialysis we use an' a helluva lot of

-

12 other things. You presented thee :I'm sorry tuh be

13 late I wz uh (.) ( 1.

14 Pr: N-Ij's presented thuh presenting histry 'n I

15 haven't gone into his subsequent course. =

16 =I wi [ll quickly go into 'that ='

17 M2: [That's fine.

18 M2 : =No. I'd like tuh talk about where we are.

19 *You talked abou'what happened at thee other hospital?

20 Pr: 3I talked about what happened at thee other hospital.
2 1 He did git one (0.2)
22 ??: (right)

23 Pr: do- he got one course of- they started him on :
24 uh : protocol, (.) an' he got (.) h wz started on
25 unchemo .hh thuh day before 'n -that's whin he

2 6 subsequently had an -acute deterioration. . .

When Pr is about to resume her account, M2 intervenes in the case presentation. In the utterance at lines 9-13, which he self-interrupts with his apology for coming late, he had begun to formulate what Pr had already presented, and at lines 14-16 she delivers her account of her presentation to that point, "I just presented the presenting history," and how she proposes to continue. M2 then undertakes to redirect the case presentation, but asks Pr whether she had discussed what treatment the patient had received at the other hospital, something one might have thought would have been included under "the presenting history" and which therefore Pr has already suggested has been conveyed. To this inquiry she responds (line 20) with a confirming repeat, "I talked about what happened at the other hospital," an allusion confirmation here delivered in a manner (intonationally) that carries overtones of testiness (conveying "Of course I talked . . . ").

Excerpt 2 1 is taken from the daily editorial news conference of a major metropolitan newspaper, at which the stories to be carried, their place- ment in the paper, and so on, are determined." Some 15 participants contribute vocally to the proceedings, including-in the excerpt to fol- low-the business editor (BE), the managing editor (ME), and the city editor (CE), the last of whom puts forward a candidate story for the consideration of the others.

Excerpt 2 1 (Clayman : editorial conference, 7/24/89 : 12-13)

1 CE: This story is cute.
2 ?F: hhh! (0.8) hhh
3 CE: --Kid yesterday w- uh Sunday goes out to the park with
4   his parents, (0.4) [really liked -the park.
5 ?F: [((whispering))
6   (0.5)
7 8 CE: ?F: So what happens?
*(This is ) *
9 CE: wakes up in the middle of the night, (.) Sunday-
He -
10   --(.) Sunday night, (0.3) decides that he is gonna get
11   outta the house, this is in Logan Squa :re, and go back -
12   to the park. (1.0) This is a three -year old kid.
13 ??: hhh
14 CE: Gets lost -on the way, (0.2) police- somebody calls the
15   cops-at about three a'clock in the morning and says
16   -hey there's a kid on the street. (0.2) Cops come -out,
17   pick the kid -up, (0.3) take him to the station, .hh he
18   goes through hamburgers, he goes through all kinds of
19   --uh junk food, (1.0) finally the cops: give up. =I mean
20   the- uh little- the little kid can't remember his
2 1   name or isn't telling it, (0.5) can't remember his -
22   address,-(0.4) uh : cops take him out on the street
23   --just for a- a walk along tour of the neighborhood

l5 I am indebted to Steve Clayman for sharing this material.

24 hoping that you know maybe if the kid will remember

-25 something, .hh kid spots his cat -(0.8) follow the cat 26 into this apartment building he's follow the cat- 2 7 follow the cat -into an apartment building (.) knock on 28 the door (0.7) Mommy comes- comes out 'nd says WHAT 2 9 THE HECK -ARE YOU DOING OUTTA HERE. (0.4) As 3 0 it turns out she's always hated the cat -until this morning. 31 (1.0)

32 ME: (She) loves the [cat now ((laughter)) 33 BE: + [She didn't realize her son was missing?

-

34 CE: ZDidn't realize her son was missing. 35 (0.6) 36 GE: Sleeping at three -in the morning I'd think I'd

37 BE: >Oh what time -was h- you know what time he was 38 brought home?

-

39 CE: Uh: :early today. 40 BE: Oh 41 (3.0)

The story is recounted in a manner that requires its hearers to "solve it" for its interest as (presumably) a "human interest story." In particular, the final components of the telling are left for "working up" by its recipients, and the telling is directly followed by such interpretive upshots. First the final component of the story has its implied contrast made explicit; "hated the cat until this morning" (line 30) being contrasted with "loves the cat now" (line 32), while leaving the basis for this turnaround still unexplicated. Then the penultimate component of the telling, which had been delivered as a stripped-down direct quotation of the mother's utterance upon opening the door and seeing the child outside ("What the heck are you doing out here?" lines 28-29), has its import formulated by a recipient for confirmation by the teller ("She didn't realize her son was missing"). The teller both confirms the understanding that it makes explicit and confirms that the telling had been designed to convey it without saying it in so many words-perhaps achieving thereby a demonstration of its potential solvability by readers as well and, accordingly, its worthy candidacy as a publishable human interest story.

In excerpt 22 we are afforded not only a direct representation of an episode of confirming an allusion but a journalistic account of it as well, in which this understanding is embodied at several levels. The exchange occurred in a court of law, during the early legal maneuverings in the 0.J. Simpson case, in which a popular celebrity was accused of murder- ing his ex-wife and a friend of hers, an event whose aftermath was given extensive and prolonged exposure in the news media.

Excerpt 22 (New York Times,national edition, June 23, 1994, p. A12)

1 As a grand jury looking into the murder of Mrs. Simpson and
2 Mr. Goldman continued to hear evidence today, Mr. Simpson's
3 lawyer, Robert L. Shapiro, appeared with the defendant in
4 Municipal Court and demanded that prosecutors speed their release
5 to him of the evidence they are presenting to the grand jurors. . .
6 Mr. Shapiro went out of his way to point to the
7 discrepancies between many details that have been reported by
8 local and national news organizations and the small number of
9 details publicly confirmed.
10 "I've been hearing reports of all this evidence that's been
11 found, and the material we have doesn't show any of it," the
12 lawyer told Judge Patty Jo McKay, referring to documents turned
13 over to him by the District Attorney's office. He specifically
14 mentioned a bloody ski mask that, according to some press reports,
15 was found in Mr. Simpson's home after the killings.
16 When Ms. Clark, the prosecutor, replied that the papers
17 given Mr. Shapiro listed all the evidence in the District
18 Attorney's possession, Judge McKay interrupted and asked

19 +directly, "So you are saying there is no ski mask?" The

20 Zprosecutor replied, "There is no ski mask." [end of story]

Note several points about the account in excerpt 22: First, the journal- ist takes it that the judge has heard the prosecutor to be conveying some- thing indirectly, which the judge then addresses by llask[ing] directly" (lines 18-19). Second, the prosecutor's reply confirms the judge's suppo- sition of what she was saying and confirms that she was conveying it indirectly; the reply thereby suggests that the prosecutor also (like the journalist) understood the judge to have heard that she (the prosecutor) was conveying something indirectly, answered in a way that addresses that possible hearing, and confirms it. Third, this was for the journalist, and the editorial oversight agent, a pointed-and apparently satisfac- tory-way of ending the story. (On the journalistic practices involved here, especially concerning the journalist's incorporation of the question in reporting the information delivered in the answer, see Clayman [1990].)

Finally, in excerpt 23, Liz, a purveyor of "air plants" at an outdoor farmers' market, is approached by a customer who had previously pur- chased plants from her.

1 Liz: Hi::
2 Cust: I want to ask you something.
3 Liz: Sure
4 Cust: I bought three of those uh you ((points to a
5   know like ( ). tray of plants))
6   One of them died out.  
7   ((gap))  
8 Cust: Uhh  
9 Liz: It did,  
10 Cust: Yeah  

11 Liz: +The other ones are doing well,

12 Cust: ZThe other ones are doing well.

13 Liz: +They were all in the same area,

14 Cust: ZSame thing, yeah.

15 Liz: Ohh.

16 Cust: Yeah

17 Liz: They were 'n the little shells,

18 Cust: Yeah, in the little shells, yeah.

This episode allows us again to refocus on the central action with which we have been concerned-confirming allusions-by juxtaposing in con- trast the practices for confirmation of other suppositions with the confir- mation of a supposition grounded in a prior inexplicit conveyance. Note, then, that the specificity of an assertion ("One of them died out"; line 6) may be taken to convey that other than the specified referent is not implicated in the utterance. When Liz undertakes to make that explicit (line ll), her inference is met with a confirming repeat (line 12). By contrast, other inquiries by Liz about the plants in question, which are also confirmed by the questioner, were not previously conveyed indi- rectly, and these confirmations (lines 13-14, 17-18) are not done by repeats of that which is being confirmed. (But note the equivocality of the exchange at lines 17-18, which may reflect the equivocality of the pointing at lines 4-5.)

These episodes from work settings, ranging from the informality of an open market to the formality of a court of law in session, underscore the robustness of the practice of confirming allusions across social and interactional contexts and across compositional features of the partici- pants. Here, at least, the interactional phenomenon does not appear to vary across "contexts" as conventionally understood. It appears then that not everything must.

CONCLUDING CONSIDERATIONS

So that is the little phenomenon I have meant to introduce: a practice of talk-in-interaction used to implement a distinctive action-an agreeing repeat used to do "confirming an allusion." Not one of the apparently major actions that compose the fabric of the everyday life of the society, whether dramatic or mundane . . . any more than E. coli or Drosophila were apparently important organisms in the larger biological scheme of things. Not, then, like requests and offers and their acceptances and declines, by which goods and services get moved around a society. Not, then, like askings and tellings and the like, by which information gets circulated around a society. But it is one of the actions that compose the culture's repertoire, and from it there may be something to be learned about actions as an object of inquiry (as was the case with other, appar- ently "limited" actions [Schegloff 19861). As compared to the grosser types of action involved in the sequences we have examined-which we could formulate as "confirmation" and "disconfirmation," one or the other of which is virtually mandated by an interlocutor's offering of a candidate understanding to a speaker of what the speaker has just said- the form that the confirmation takes appears to be what might be called an "optional action." Confirming by repeating the proposed upshot as a way of ratifying its prior inexplicit conveyance is, then, a relatively minor hue in the cultural, discursive, and behavioral palette of common- place interaction. The breadth of its provenance (in spite of its relative infrequency) should indicate that the fabric of ordinary sociality may be more finely woven and more subtly shaded than has been generally entertained, and that this is so across the conventional differentiations of social categories that professional social science and vernacular opinion are both inclined to believe fundamental to social organization. Yet it is socially organized and socially deployed, and accessible to systematic, empirical inquiry.

Perhaps the "minor" status of our object of inquiry should be appreci- ated from another perspective. The preoccupation of the present report is but a corner of a much larger domain omnipresent in talk (and arguably in any use of language) in which talk is laced through and through with inexplicitness and indexicality and, in the nature of the case, requires "solving" by hearers. In the episodes examined in this inquiry we peek at a corner of this domain, in which hearers make their understandings- their solutions-explicit, seek confirmation of them, and get that confir- mation in a particular form. But for talk (or other forms of conduct, for that matter) to go on and-through it-interaction (and all "practical action," as was argued by Garfinkel [1967]), such inexplicitness is being "solved" all the time. Its results are displayed (even when not formu- lated) in the ensuing talk and action and are subject to repair there if found problematic (Schegloff 1992b). The arguable omnipresence of inexplicitness and indexicality is not incompatible with its differential importance and treatment across the ranges of social settings. The text of this report suggests types of interactional "moments" for which inex- plicitness is specially relevant. On a different-sociopolitical-canvas, the study of political discourse has long been familiar with the ways in which repressive social and political organization can put a premium on the exploitation of inexplicitness in discourse-as, for example, in Leo Strauss's (1952) account of the exoteric and esoteric traditions of writing in the discourse of political theorizing and the interpretive possibilities thereby opened up. So the "downplaying" of the import of our topic needs to be properly understood. It is not that the domain involved here is minor; it is rather that this particular action is of relatively modulated import when compared to some other types of action and when assessed relative to the larger domain of the working of language in interaction of which it taps one type of development.

One last observation. Until I grappled with a collection of actual, naturally occurring repeats, and with the assembling of this core collec- tion in particular, I had not the slightest idea that there was such a function, such an action, such a practice in talk as "confirming that something had been conveyed inexplicitly,"-confirming both the allusion and that it had been an allusion. This is not the sort of action that is part of the articulable vernacular culture (what Giddens [I9841 terms "discursive culture"); there is no speech act term for it; it is not readily accessible to intuition, although without a native's cultural knowledge it might not be detectable in an examination of interactional materials. (The bearing of this point on the possibility of a universal, rational, aprioristic pragmatics of the sort projected by Habermas cannot be ex- plored here, but should be weighed by those committed to such a project. The Habermasian view presumes that we know, or can know, preanalyt- ically and preempirically what persons actually do and how, and this may well not be so.)

At a time when the possibility of genuine discovery grounded in careful examination of empirical material is often taken to be compromised by the particular cultural commitments of inquirers or of their culture, or to be arguably nothing more than a function of the presuppositions of (the) inquiry per se (and all this especially relevant where the cultural is involved), it may be salutary to weigh the evidence of the present inquiry. Surely the presuppositions of the enterprise are germane to the formula- tion of the problem-concerned with what action some form of talk is doing-as well as to the conception of what an answer might look like. But arguably these presuppositions serve not to subvert the claims of the inquiry but to undemrite them; for it seems clear that a pervasive, if not omnirelevant, orientation and concern of parties to interaction about any utterance of an interlocutor is "Why that now?" (Schegloff and Sacks 1973, p. 299), and a prime candidate solution generically is, "What is the speaker doing thereby?"-that is, what action is that utterance doing or being used to do? The presuppositions of this academic inquiry, then, are themselves meant to be grounded in the generic relevancies of ordinary actors in the quotidian activities out of which their lives in interaction are composed.

If the proposals of this article are on target, then, it reports the finding of something that was not presupposed, was not even imagined. It is possible to find new things, things that could not have been conceived before the inquiry, though if inquirers presume otherwise and act accord- ingly, they will guarantee the reaffirmation of their presumption. This has direct consequences for how one does one's work. With materials of the sort being examined here, one cannot simply go through the stream of conduct seeking to identify which previously known action, phenome- non, practice, and so on, a given "bit" exemplifies. ("Confirming an allusion" would not have been an available analysis of an utterance in conversation before this inquiry, and confirming repeats would have been left unanalyzed, or worse, given an ad hoc-and most probably a vernac- ular-interpretation.) One needs to ask about each new object of exami- nation what it could be possibly doing, and ask, as well, what could constitute an object of examination to which this question could be put, that had not been so considered before.

With such a study policy applied to the raw material of the society's life, how many other such unknown jobs, functions, actions, practices- such as confirming allusions-might there turn out to be? Hundreds? Thousands? Yet these make up the cultural inventory of the society, elements of which the parties assemble and parse in the moment-to- moment constituting of its quotidian embodiment, and that serve as the prima facie rationale for the deployment of language. It is about these that a theory of action is presumably theorizing. Without knowing them, there is the prospect that we lack articulate, theoretically informed access to a substantial bulk of our culture, and rely on our vernacular under- standing of typicalized courses of action-designed for the working of the society and not for the working of its disciplined study-to process much of the basic material of our inquiry. We lack the social equivalent of the presumed 90% of the physical matter of the universe now unaccounted for and termed "dark matter." Without bringing this inexplicit part of our culture and of the basic repertoire of practical conduct under empiri- cally grounded analytic control, we are unlikely to grasp at the deepest level how either language or action or sociality work.

APPENDIX

Transcription Conventions

A brief guide to a few of the conventions employed in the transcripts may help the reader in what may appear to be a formidable undertaking. It is apparent from the excerpts printed in this article that some effort is made to have the spelling of the words roughly indicate the manner of their production, and there is often, therefore, a departure from standard orthography. Otherwise:

+ Arrows in the margin point to the lines of transcript rele- vant to the point being made in the text.
( 1 Empty parentheses indicate talk too obscure to transcribe. Letters inside such parentheses indicate the transcriber's best estimate of what is being said.
hhh The letter "h" is used to indicate hearable aspiration, its length roughly proportional to the number of "h"s. If pre- ceded by a dot, the aspiration is in-breath. Aspiration in- ternal to a word is enclosed in parentheses. Otherwise "h"s may indicate anything from ordinary breathing to sighing to laughing, etc.
[ Left-side brackets indicate where overlapping talk begins.
  Right-side brackets indicate where overlapping talk ends, or marks alignments within a continuing stream of over- lapping talk.
((points))   Words in double parentheses indicate transcriber's ments, not transcriptions. com
(0.8) Numbers in parentheses indicate periods of silence, in tenths of a second.
(.) A dot in parentheses indicates silence of less than two tenths of a second.
  Colons indicate a lengthening of the sound just preceding them, proportional to the number of colons.
becau- A hyphen indicates an abrupt cut-off or self-interruption

2 12

Confirming Allusions
of the sound in progress indicated by the preceding letteris)
(the example here represents a self-interrupted "be
cause").
He says - Underlining indicates stress or emphasis
drAink A "hat" or circumflex accent symbol indicates a marked
  pitch rise.
-- Equal signs (ordinarily at the end of one line and the start
  of an ensuing one) indicates a "latched" relationship-no
  silence at all between them.
"Then0 Material between degree marks is markedly quieter than
  surrounding talk.

Fuller glossaries may be found in Sacks et al. (1974, pp. 731-34) and in Atkinson and Heritage (1984, pp. ix-xvi).

REFERENCES

Austin, John L. 1962. How to Do Things with Words. Oxford: Clarendon Press. , 1970. Philosophical Papers. Oxford: Oxford University Press. Bales, Robert F. 1950. Interaction Process Analysis: A Method for the Study of Small Groups. Reading, Mass.: Addison-Wesley.

Brun-Cottan, Francoise. 1988. "Between Us: Competitive-Cooperation, Networking, and Other Hollywood Development Job-Finding Practices in Conversational Inter- action." Ph.D. dissertation. university of California, Los Angeles, Department of Anthropology.

Clayman, Steven. 1990. "From Talk to Text: Newspaper Accounts of Reporter- Source Interactions." Media, Culture and Society 12 (1): 79-104. Drew, Paul, and John C. Heritage. 1992. "Analyzing Talk at Work: An Introduc- tion." Pp. 3-66 in Talk at Work: Interaction in Institutional Settings, edited by

P. Drew and J. Heritage. Cambridge: Cambridge University Press. Duranti, Alessandro. 1988. "Intentions, Language and Social Action in a Samoan Context." Journal of Pragmatics 12: 13-33. Frake, Charles 0. 1964. "How to Ask for a Drink in Subanun." American Anthropol- ogist 66, no. 6, pt. 2, pp. 127-32. Garfinkel, Harold. 1967. Studies in Ethnomethodology. Englewood Cliffs, N.J.:

Prentice-Hall.

Geertz, Clifford. 1973. The Interpretation of Cultures. New York: Basic Books.

Giddens, A. 1984. The Constitution of Society: Outline of the Theory of Structuration.

Berkeley and Los Angeles: University of California Press. Goffman, Erving. 1955. "On Face-Work: An Analysis of Ritual Elements in Social Interaction." Psychiatry 18 (3): 2 13-31. . 1959. The Presentation of Self in Everyday Lve. Garden City, N.Y.: Dou- bleday. . 1963a. Behavior in Public Places: Notes on the Social Organization of Gather- ing. New York: Free Press. . 1963b. Stigma: Notes on the Management of Spoiled Identity. New York: Simon & Schuster.

. 1967. Interaction Ritual: Essays in Face to Face Behavior. Garden City,

N.Y.: Doubleday.
. 1976. "Replies and Responses." Language in Society 5:257-313.
. 1981. Forms of Talk. Philadelphia: University of Pennsylvania Press.

Gonzales, Patrick. 1995. "Re-Presentation in Science: Reported Speech and Reformu- lation in Talk among Physicists." Manuscript. University of California, Los Angeles, Department of TESL and Applied Linguistics.

Goodenough, Ward H. 1957. "Cultural Anthropology and Linguistics." Pp. 16773 in Report of the Seventh Annual Roundtable Meeting on Linguistics and Lan- guage Study, edited by P. L. Garvin. Washington, D.C.: Georgetown University Press.

--. 1964. Introduction to Explorations in Cultural Anthropology, edited by

W. H. Goodenough. New York: McGraw-Hill. Gumperz, John J., and Dell Hymes, eds. 1964. Special issue, "The Ethnography of Communication." American Anthropologist 66, no. 6, pt. 2. Habermas, Jiirgen. 1970. "Toward a Theory of Communicative Competence." Pp. 114-48 in Recent Sociology No. 2, edited by H. P. Dreitzel. New York: Macmillian. . 1984-87. The Theory of Communicative Action. Boston: Beacon Press.

Heritage, John C. 1984a. "A Change-of-State Token and Aspects of Its Sequential Placement." Pp. 299-345 in Structures of Social Action, edited by J. M. Atkinson and J. C. Heritage. Cambridge: Cambridge University Press.

. 19846. Gayfinkel and Ethnomethodology. Cambridge: Polity Press. . 1985. "Analyzing News Interviews: Aspects of the Production of Talk for an Overhearing Audience." 3:95-119 in Handbook of Discourse Analysis, edited

T. A. Dijk. New York: Academic Press.

Heritage, John C., and Marja-Leena Sorjonen. 1994. "Constituting and Maintaining Activities across Sequences: And-Prefacing as a Feature of Question Design." Language in Society 1:l-29.

Homans, George C. 1961. Social Behavior: Its Elementary Forms. New York: Har- court, Brace & World. Jefferson, Gail. 1972. "Side Sequences." Pp. 294-338 in Studies in Social Interac- tion, edited by D. N. Sudnow. New York: Free Press. . 1980. "On 'Trouble-Premonitory' Response to Inquiry." Sociological Inquiry 50 (34): 153-85. . 1984. "Notes on a Systematic Deployment of the Acknowledgement Tokens 'Yeah' and 'Mm hm.' " Papers in Linguistics 17:197-216.

. 1986. "Colligation as a Device for Minimizing Repair or Disagreement." Paper presented at the Conference on Talk and Social Structure. University of California, Santa Barbara.

. (1983) 1993. "Caveat Speaker: Preliminary Notes on Recipient Topic-Shift Implicature." Research on Language and Social Interaction 26 (1): 1-30.

Jefferson, Gail, Harvey Sacks, and Emanuel A. Schegloff. 1987. "Notes on Laughter in the Pursuit of Intimacy." Pp. 152-205 in Talk and Social Organisation, edited by G. Button and J. R. E. Lee. Clevedon: Multilingual Matters.

Johnstone, Barbara, ed. 1987. Special issue, "Perspectives on Repetition." Text 7 (3).

Keenan, Elinor Ochs. 1977. "Making It Last: Repetition in Children's Discourse." Pp. 125-38 in Child Discourse, edited by S. Ervin Tripp and C. Mitchell Kernan. New York: Academic Press.

Labov, William, and David Fanshel. 1977. Therapeutic Discourse: Psychotherapy as Conversation. New York: Academic Press.

Lerner, Gene H. 1987. "Collaborative Turn Sequences: Sentence Construction and Social Action." Ph.D. dissertation. University of California at Irvine, Department of Psychology.

Levinson, Stephen C. 1983. Pragmatics. Cambridge: Cambridge University Press.

Mandelbaum, Jennifer. 1993. "Assigning Responsibility in Conversational Storytell- ing: The Interactional Construction of Reality." Text 13 (2): 247-66.

Matza, David. 1969. Becoming Deviant. Englewood Cliffs, N.J.: Prentice Hall.

Miner, Horace. 1956. "Body Ritual among the Nacirema." American Anthropologist 58:503-7. Norrick, Neal R. 1987. "Functions of Repetition in Conversation." Text 7 (3): 245-64.

Ochs, Elinor. 1988. Culture and Language Development: Language Acquisition and Language Socialization in a Samoan Village. Cambridge: Cambridge University Press.

Parsons, Talcott. 1937. The Structure of Social Action. New York: McGrawHill. ~omerantz,Anita M. 1978. "Attributions of Responsibility: Blamings." Sociology 12 (1): 115-21.

. 19786. "Compliment Responses: Notes on the Cooperation of Multiple Con- straints." Pp. 79-112 in Studies in the Organization of Conversational Interaction, edited by J. Schenkein. New York: Academic Press.

. 1980. "Telling My Side: 'Limited Access' as a 'Fishing' Device." Sociological Inquiry 50:186-98.

. 1984. "Agreeing and Disagreeing with Assessments: Some Features of Pre- ferred1Dispreferred Turn Shapes." Pp. 57-101 in Structures of Social Action: Stud- ies in Conversation Analysis, edited by J. M. Atkinson and J. Heritage. Cambridge: Cambridge University Press.

. 1987. "Descriptions in Legal Settings." Pp. 226-43 in Talk and Social Orga- nisation, edited by G. Button and J. R. E. Lee. Clevedon: Multilingual Matters.

Sacks, Harvey. 1972. "On the Analyzability of Stories by Children." Pp. 325-45 in Directions in Sociolinguistics: The Ethnography of Communication, edited by J. J. Gumperz and D. Hymes. New York: Holt, Rinehart & Winston.

. 1975. "Everyone Has to Lie." Pp. 57-80 in Sociocultural Dimensions of Language Use, edited by M. Sanches and B. G. Blount. New York: Academic Press.

. (1973) 1987. "On the Preferences for Agreement and Contiguity in Sequences in Conversation." Pp. 54-69 in Talk and Social Organisation, edited by G. Button and J. R. E. Lee. Clevedon: Multilingual Matters.

. 1992.Lectures on Conversation, 2 vols. Edited by G. Jefferson with introduc- tions by E. A. Schegloff. Oxford: Blackwell.

Sacks, Harvey, and Emanuel A. Schegloff. 1979. "Two Preferences in the Organiza- tion of Reference to Persons and Their Interaction." Pp. 15-21 in Everyday Lan- guage: Studies in Ethnomethodology, edited by G. Psathas. New York: Irvington.

Sacks, Harvey, Emanuel A. Schegloff, and Gail Jefferson. 1974. "A Simplest System- atics for the Organization of Turn-Taking for Conversation." Language 50: 696-735.

Safire, William. 1991. "Language: Frank, Forthright and Forthcoming." International Herald Tribune, May 13.

Schegloff, Emanuel A. 1967. "The First Five Seconds: The Order of Conversational Openings." Ph.D, dissertation. University of California, Berkeley, Department of Sociology. . 1968. "Sequencing in Conversational Openings." American Anthropologist 70:1075-95. . 1972. "Notes on a Conversational Practice: Formulating Place." Pp. 75-119 in Studies in Social Interaction, edited by D. N. Sudnow. New York: Free Press. . (1973) 1987. "Recycled Turn Beginnings: A Precise Repair Mechanism in Conversation's Turn-Taking Organisation." Pp. 70-85 in Talk and Social Organi- sation, edited by G. Button and J. R. E. Lee. Clevedon: Multilingual Matters. . (1976) 1984. "On Some Questions and Ambiguities in Conversation." Pp.

28-52 in Structures of Social Action, edited by J. M. Atkinson and J. C. Heritage. Cambridge: Cambridge University Press. . 1980. "Preliminaries to Preliminaries: 'Can I Ask You a Question?' " Socio

logical Inquiry 50: 104-52. . 1986. "The Routine as Achievement." Human Studies 9: 111-5 1. . 1988a. "Goffman and the Analysis of Conversation." Pp. 89-135 in Erving

Goffman: Exploring the Interaction Order, edited by P. Drew and A. Wootton. Cambridge: Polity. . 19886. "On an Actual Virtual Servo-Mechanism for Guessing Bad News: A Single Case Conjecture." Social Problems 35 (4): 442-57. . 1988c. "Presequences and Indirection: Applying Speech Act Theory to Ordi- nary Conversation." Journal of Pragmatics 12:55-62. . 1992a. Introduction to Harvey Sacks: Lectures on Conversation, vol. 1. Ed- ited by G. Jefferson. Oxford: Blackwell.

. 19926. "Repair after Next Turn: The Last Structurally Provided Place for the Defense of Intersubjectivity in Conversation." American Journal of Sociology 95 (5): 1295-1345.

. 1993. "Reflections on Quantification in the Study of Conversation." Research on Language and Social Interaction 26 (1): 99-128.

Schegloff, Emanuel A., Gail Jefferson, and Harvey Sacks. 1977. "The Preference for Self-correction in the Organization of Repair in Conversation." Language 53: 361-82.

Schegloff, Emanuel A., and Harvey Sacks. 1973. "Opening Up Closings." Semiotics 7:289-327. Searle, John R. 1965. "What Is a Speech Act?" Pp. 22-139 in Philosophy in America,

edited by M. Black. Ithaca, N.Y.: Cornell University Press. . 1969. Speech Acts. Cambridge: Cambridge University Press. , 1976. "The Classification of Illocutionary Acts." Language in Society 5:l-24.

Strauss, Leo 1952. Persecution and the Art of Writing. Glencoe, Ill.: Free Press. Tannen, Deborah. 1987a. "Repetition in Conversation: Toward a Poetics of Talk." Language 63 (3): 574-605. --. 19876. "Repetition in Conversation as Spontaneous Formulaicity." Text 7 (3): 215-43. . 1989. Talking Voices: Repetition, Dialogue, and Imagery in Conversational Discourse. Cambridge: Cambridge University Press.

Comments
  • Recommend Us