Paper about Sociolinguistics analysis

CHAPTER 9 What Drives Social Action?

Jenny Mandelbaum Rutgets University

Anita Pomerantz Temple University

Conversation analysts aim to make sense of sense-making-to explicate the met hods through which everyday interactants produce discourse. This involves the close examination of tape-recorded, transcribed everyday conversations. Some of the topics studied by conversation analysts include how interactants take and give up turns (Sacks, Schegloff, & Jefferson, 1974), orient to sequences (Schegloff &Sacks, 1973), repair their own and others' talk (Schegloff, Jefferson, & Sacks, 1977), issue invitations (Drew, 1984). seek information (Pomerantz, 1988), and tell about their troubles (Jefferson, 1980). '

Throughout these and other studies, we (as conversation analysts) are funda- mentally interested in participants' concerns, orientations, and enterprises. At the same time, however, we have resisted addressing the question of intention. We often have written about participants' orientations and concerns in almost behavioristic ways, giving discourse features as evidence of proposed orienta- tions and concerns. We have often proposed actors to be concerned about discourse features, a s opposed to being concerned about pursuing actions. Traditionally conversation analysts have described participants' concerns in terms of their orientations to features of conversation. For instance, participants a r e portrayed as being concerned about whether or not a particular turn is to be understood to be the reason for the call, or whether and how the conversation is to be closed. and so forth (Schegloff, 1986; Schegloff & Sacks, 1973). These orientations are formulated as participants' answers to the question, "Why that now?"-why is he or'she saying that thing a t this particular point?

Whereas conversation-analytic studies have tended to downplay or be inex-

152 MANDELBAUM AND POMERANTZ

plicit about the role of intentions, studies in speech communication have fre- quently unquestioningly treated interaction as wholly intenlional and goal- driven. As Craig (1 986, p. 257) has noted, tlie assumption that lluman behavior is goal-oriented is pervasive in many fields. Much of the work on goals in the speech communication field seems to be influenced by cognitive science models of action (e.g., Schank & Abelson, 1977). This work implicitly. and sometimes explicitly, takes the position that discourse is constructed in such a way as to enable interactants to achieve goals. I f it is a routine goal, actors will have a familiar script to play out in order to achieve the goal. I f not, a plan will be formulated for engaging in behavior that will lead to the satisfaction of the goal. According to this view. interaction is seen to be designed by a pre-existing goal. (For instance, cf. Rule & Bisanz, 1987; Canary, Cunningham, & Cody, 1988; Dillard, 1990, re: compliance-gaining goals. See Manusov, 1989. for a useful review.) Actions that a r e not in concert with that goal a r e sometimes described as "digressions" (Jacobs. Jackson, Hall, & Stearns, this volume).

Conversation analysis's tendency to disattend or be inexplicit about partici- pants' concerns, and speech communication's tendency to regard all of social action a s goal-driven, suggest the need for a closer look at what drives social action, if indeed "driving" is the appropriate metaphor. Conversation analysis's particular strength is its ability to identify details of social action by looking closely a t interaction. In this chapter we address the question of participants' intentions in social action from a (our) conversation-analytic perspective. Drawing inferences from analysis of a fragment of conversation, we propose that. rather than generally being planfully driven by goals, .social action some- times is shaped by various types of concerns. We also note, however, that there a r e taken-for-granted features of interaction, and some more formal properties of it. that influence how social action is organized. In this chapter, we show how a close look suggests that social action is worked out between participants i n a n ongoing way, rather than driven by goals.

In order to answer the question, "What drives social action?", we discuss a telephone conversation in which it is rather clear what the participants were trying to achieve. One participant was asking for help, and the other was declining t o help. This conversation provides us with materials for illustrating the distinctions we wish to make. Whereas some discourse choices were respon- sive to primary, multiple and contingent, or prerequisite concerns, other dis- course choices were influenced by taken-for-granted considerations or by other properties of social interaction. Below we clarify the proposed three concerns and two other influences by showing how they are manifested in conversation. Before turning to the telephone call, we briefly outline some of the issues raised by these distinctions.

First, we emphasize that we are dealing with inferredconcerns. Our data are the details of interaction. This approach does not permit us to determine what a n a r t i c i ~ a n t "actually" is concerned about. Indeed, we take it that such informa-

tion is currently very hard to obtain. Even self-reports of "actual" concerns a r e mediated by the situation in which they a r e solicited and may simply provide an account of what a participant remembers himself or herself to have been concerned about when prompted to think about it by a researcher. We inferred the mental states of the interlocutors from patterns in their interaction.

Second, level of awareness or consciousness of concerns is another cognitive matter to which we have no direct access. As noted above, when we propose that someone had a concern, we are involved in a process of inference. Sirni- larly, we inferred probable consciousness or absence of it from various features of the participants' behavior. Although these parlicipants made their concerns apparent for one another and acted in ways that indicated that they wereat least sbhewhat aware of what they were achieving, we mark as problematic the answers to (a) how "aware" or "conscious" participants are, and@) the affects of their awareness (or lack of it) on their actions.

The following telephone call' provides us with data for exploring an answer to, "What drives social action?"

O Ring

1 Sheila: Hello?

2 Rorlny: 'lo Sheila,

I Ronny: ((7's) Rjonny.

5 Sheila: Hi Ro~iny.

G Ronny: Guess what.hh

7 Sheila: What.

8 Ronny: .hh My ca:r is sta::lled.

9 (0.2)

10 Ronny: ('n) I'm u p here in the Gle117

1 1 Sheila: 011::.

12 Ronny: h h h

13 Ronny: A:nd.lrh (0.2) 1 don' know if it's: po:ssible, but

14 .Ill111 see I heveta open up the ba:nk.hh

'We wish lo thank Manny Scheglolf lor providing the data and for his contributions to theanalysis. We are also ~ r a t e l u l for his feedback on a late draft of this chapter. See appendix for transcription symbols.

154 MANDELBAUM AND POMERANTZ

16 Ronny: a:t uh: (.) in Brentwood?hh=

17 Sheila: =Yeah:- en I know you want- (.) en I whoa-

18 (.) en I would, but- except I've gotta leave

19 in about five min(h)utes. =

20 Ronny: [=Okay then I gotta call somebody else.right away.

21. Sheila: [(hheh)

23 Ronny: Okay?=

24 Sheila: =Okay [Ron 1

25 Ronny: [Thanksla lot.= Bye-.

26 Sheila: Bye:.

We offer five distinctions to help us better understand what drives the social action in this fragment of talk. These include primary concerns, rnultlple and contingent concerns, prerequisite concerns, taken-for-granted sustained achievements, and unmindful actions.

PRIHARY CONCERNS

We infer that Ronny's primary concern here was to get help (most likely a ride) and that, in order to satisfy this concern, he called Sheila. The inference that this was Ronny's primary concern is based on the following features of his talk. The call was organized around getting help. It was his first and only official business after identification was achieved. Ronny's talk was limited to what was neces- sary for determining whether or not Sheila could help him: He identified the answerer as Sheila and himself to Sheila (lines 2-5), he indicated he had news to tell (line 6): he reported a problem that Sheila might have helped him to solve (lines 8, 10, 13-14, 16), and he closed the call immediately upon learning that Sheila could or would not help him out (line 20).

These details suggest that this was a single-purpose call. One could hypothe- size that there was an additional motive to the request. In the way that friend- ships may be tested with requests (Baxter & Wilmot, 1984), Ronny might have been testing whether Sheila was prepared to put herself out for him in time of need.2 w e do not know enough about the doing of secret tests to say whether or

'We are differentiating Ronny's making the request to test Sheila's friendship from an effect that compliance with or relusal lo a request may have on a relationship. More specifically, although

9. WHAT DRIVES SOCIAL ACTION? 155

not this is likely. We see rlotlling in the discourse, however, to indicate that Ronny was concerned with testing the relationship.

MULTIPLE A N D CONTINGENT CONCERNS

In writing of "dialectical goals," Craig (1986) pointed out that "people often face the need to do more than one thing a t the same time" @. 264). For example, In asking for someone's help, participants often a r e concerned both with getting the help they seek and with not imposing on the other more than is necessary. Likewise, in declining to help friends, participants often are concerned both with not committing themselves to perform the help-tasks and with not engendering negative reactions.

We propose that both Ronny and Sheila were dealing with multiple concerns. More specifically, we propose that, when Ronny and Sheila enacted their primary concerns (i.e., getting help and declining to give help, respectively), there were additional concerns attached to those enactments. Next, we describe in detail how Ronny asked for help from Sheila in a way that minimized the degree to which he was imposing on her. This concern not to impose can be seen a s an additional concern that came about contingently, that is, upon finding that he needed to turn to someone for help. Sheila declined to help Ronny while displaying a sympathetic and friendly attitude toward his asking. This concern to be sympathetic and friendly also can be seen a s a contingent concern that came into being when Sheila chose not to help Ronny.

A characteristic of Rontly's talk that makes available his contingent concern is its neutrality with respect to what he wanted. He began with "Guess what," which indicated news but did not suggest whether the news was good or bad. He told his problem, without stating the upshot of it: "My ca:r is sta::lled . . . ('n) I'm up here in the Glen? . . . see I haveta open up the ba:nk.hh . . . a:t uh: (.) in Rrentwood?hh." Thus. Sheila was put in a position of "inferring" why he was reporting the situation to her. By describing a problem without making a n explicit request, Ronny gave Sheila a chance either to volunteer to help or to treat what he said as news. Sheila treated it as news in line 11 by indicating, with her "Oh::.," that she had just been informed by what he said (see Heritage, 1984, re: change of state tokens). In reporting an event that can be heard a s a "trouble" or "problem," Ronny most likely was counting on Sheila's understanding of problems as needing solutions. In telling her the trouble or problem, he made available to her that she might help him, but he did not directly indicate anaction .

Sheila's not having given help (or giving help had she done so) may or may not have had consequences for how Ronny viewed their friendship, that is quite separate lrom whether he made the rrn~lmt a s a Imt of her Irie~~dship.

9. WI-1AT DRIVES SOCIAL AC'KION? 157 156 MANDELBAUM AND POMERANTZ

that he would have liked Sheila to take in response to his p r ~ b l e m . ~ Thus Ronny gave Sheila an opportunity to offer the help or not, rather than putting her in a position of having to respond yes or no to an explicit request.

In addition to not making an explicit request for help. Ronny gave Sheila an "out." In saying. "A:nd.hh (0.2) 1 don' know i f it's: po:ssible" (line 13), Ronny was displaying a n awareness that Sheila may have to turn him down. Ronny offered a basis that Sheila might have for not helping him that minimized the poter~tial offensiveness of her not helping him out. He proposed not that she would not want to but that she may not be able to, that is, that it would not be her choice. Hence he was making it easier for Sheila to not help, should that be the option she was selecting.

We a r e arguing, then, that through the way that Ronny constructed his talk, he showed Sheila that he had concerns that were attached to getting help. The way that he constructed his talk shows that he was also concerned with not imposing on Sheila. In reporting his problem without officially requesting Iiclp, Ronny enabled Sheila to volunteer to help him. In naming an inoffensive basis for her not helping him, he may have made it easier for her to decline to help. Thus, he attempted to actualize his primary concern in a way that was sensitive to his concern of not imposing on her.

Our account of what motivates Ronny here is based in details intrinsic to the talk, as is usual in conversation-analytic inquiry. ltis also possible to posit a more social psychological explanation for the phenomenon. As the following social psychological explanation indicates, though, it may be difficult to ground this explanation in the details of the talk.

Rawlins (1983a, 1983b) proposed a "dialectic" of friendship, in which friends a r e torn between the contradictory impulses of the freedom to be independent and the freedom to be dependent. Rawlins says. "While each person is free to pursue individual goals and interests separate from the other and without the friend's interference o r help, each retains the liberty to call on the other for assistance, should it be necessary" (1983a, pp. 259-260).

In friendships, tensions often arise between wanting to help a friend, on the one hand, and wanting to pursue one's own concerns, on the other. In asking a favor of a friend, interactants often a r e sensitive to this dilemma and attempt to minimize the expected inconvenience, avoid having the friend sacrifice per- sonal priorities, and/or insure that the friend perceives that h e o r she can say no. Ronny's concern for getting help (his primary concern) carries with it a contin- gent concern of not overly imposing on Sheila.

Ronny again displayed this contingent concern in dealing with Sheila's

'See Drew. 1984. for a similar method used wilh respect lo making and lurning down ir~vilations. Drew noted that. when a speaker offers n report and leaves Ihe recipient lo draw the upsl~ot of tlic report. this is a device a spenker may use lo avoid making "rejecting" the talk's ofliclal business.

declination to help. In lines 17-19, Sheila "anticipated" what Ronny was calling for and offered a reasoll for not helping him:

17 Sheila: =Yeah:- ell I know you want- (.)en 1 whoa-

18 (.) en I would, but- except I've golta leave

1 $1 in about five tnin(11)utes. =

In response to her giving a reason, lionny right away accepted her refusal. He did not question or challenge her reason. He did not plead, cajole, threaten, promise, or make any appeals designed to make her change her mind. Rather, he immediately accepted the reason and announced his intention to solve the problem another way:

2 0 ~ o n n y : [=Okay Ihen 1 gotta call somebody else.right away.

Ronny's abandoning his attempt to get help from Sheila may have been related to his not wanting to put pressure on her to do something that she did not want to do. This concern might have emerged in that he was attempting to solicit Sheila's help, where Sheila's giving help may involve some degree of sacrifice on her part.

To reiterate, Ronny's alertness to the possibility of imposing was suggested in the ways in which he made the request in the firs1 place and dealt with her response to it. He presented the matter in a way that allowed her to volunteer, and, once she indicated a problenl with giving the help that she inferred he wanted, he was very quick to accept her position.

Deriving our observations from this conversation we have argued that a favor asker may orient to the presumed imposition or cost of the recipient's compli- ance by providing a choice and giving a legitimate out for the recipient. In addition, a favor asker may orient to the presumed imposition in deciding whom to ask. If one imposes to some degree on the selected target when soliciting help, then one selecls a target where that imposition is more or lessappropriate to the relationship. Ronny selected Sheila as a target for his request. When Sheila indicated unwillingness to help, Ronny quickly moved to report that he would select another target. 'The multiple concerns of getting help while not "overly" imposing would seem to affect both the selection of the target and the manner in which help is sought.

Similarly, we infer that Sheila had contingent concerns in turning down Ronny's request for help. Although Sheila clearly was concerned with pursuing her own activities, she turned him down in a way that showed friendliness toward him and a sympathetic attitude toward his problem. In lines 18-19, she turned dowri liis inexplicit request i l l a way that claimed a willingness to have done tlle favor but for circumstances:

158 MANDELBAUM AND POMERANT? 9. WHAT DRIVES SOCIAL ACTION7 159

18 Sheila: . . . en I would, but- except I've gotta leave 19 In about five rnin(h)utes. =

Sheila cojoined a claim of willingness ("en 1 would") with a claim of uncontrol- lable circumstances ("I've golta leave in about live min(h)utes0 [emphasis a d d ed]). With this format, she treated his request as legitimate, that is, as one that was appropriate for him to askof her. In other words, in her refusal sl~e.in~plicitly claims that she was an appropriate target for the request, despite the refusal. Additionally, she may have displayed a sympathetic attitude toward his request by producing the refusal with disfluency, that is, behaving in a way that was interpretable as a show of discomfort a t turning him down.

Both Ronny's and Sheila's enactments show that contingent concerns accom- panied their primary concerns. Faced with the prospect of "needing to get help" or "turning down a request." interactants may experience additional concerns, such as not imposing or offending the cointeractant. These additional concerns shape the way the actions are carried out.

A qualification is necessary here: We have called "not imposing" and "main- taining a friendship" contingent concerns because they manifested themselves in the enactments of Ronny's getting help and Sheila's declining to help, respec- tively. Presumably, a contingent concern in one circumstance may be a primary concern in another.

PREREQUISITE CONCERNS

Assume that an actor has a primary concern and engages in some interaction in an effort to satisfy that concern. Some of the behaviors displayed a r e shaped by his or her attempting to satisfy the primary concern (e.g., getting help). There also a r e behaviors that a r e shaped by contingent concerns (e.g., not overly imposing). Still other behaviors are products of yet another kind of concern: concerns that emerge by virtue of getting into the position of attempting to satisfy the primary concern. We call these prerequisite concerns.

A prerequisite concern is one that is a prerequisite to achieving the business of the conversation. For instance, prerequisite to Ronny's getting help from Sheila is for Ronny to reach her. We assume that Ronny was oriented to whether Sheila was home or not, because his being able to reach her was a prerequisite for his seeking help from her. We d o not know whether Ronny presu~ned that Sheila would be home, thought it likely, or was indeed concerned about catching her. In the discourse that we have from the start of the phone call, we see no display of this concern.

We offer another example of a prerequisite concern. In order for Ronny to try to get Sheila's help, he had to know that she knew that he was the caller. Hadshe

shown that she recognized his voice, he would have known that she knew his identity. As it turned out, she did not show that she knew who was calling (line 3):

O Ring

I Sheila: Hello?

2 Ronny: 'lo Sheila,

In line 3, she confirmed her own identity without offering to identify the caller. In response tosheila's just confirming her own identity, Ronny identified himself (line' 4):

2 Ronny: 'lo Sheila.

3 Sheila: Yea[:h 1

4 Ronny: [('t's) Klol~ny.

5 Sheila: I l i Ronny.

Ronny's specific concern to be recognized emerged after Sheila's behavior in line 3 suggested that she may not have recognized him. Although in some phone calls the interactants unproblematically recognize each other's voice, in other phone calls, interactants nlay become concerned with being recognized and/or recognizing the other by voice (Schegloff, 1979, 1986). In this case, we suggest that Ronny's concern about whether Sheila recognized him may have emerged when her talk exhibited some doubt about it.

The two prerequisite concerns that we discussed are Ronny's possible con- cern with whether Sheila was at home and Ronny's emergent concern with whether Sheila recognized him as the caller. Both a r e offered as concerns that Ronny may have dealt with before being able to seek help from Sheila.

TAKEN-FOR-GRANTED SUSTAINED ACHIEVEMENTS

'Throughout this chapler, we have been fighting with, and hedging on, whether using phrases like, "The inleractant's apparent concern was. . ." or "A concern emerged. . ." implies that the actors were aware of the proposed concerns. We now want to mark out one area in which interactants generally are not aware. That area involves achieving the taken-for-granted aspects 9f our lives.

E r v i n ~ Goffman (1 967) dislinguished between circumstances in which obliga

160 MANDELUAUM AND POMERANTL 9. WHAT DRIVES SOCIAL ACTION? 161

tions and expectations are met without thinking or feeling them and circum- stances in which obligations and expectations are felt:

In fact, most actions which are guided by rules of conduct are performed u~llllink- ingly, the questioned actor saying he performs "for no reason" or because he "fell like doing so." Only when his routines are blocked may he discover . . . that liis failure to perform them can become a matter of shame and humiliation. Similarly, he may so take for granted his expectations regarding others that only when things

g o unexpectedly wrong will he suddenly discover that he has grounds for indigna- tion. @p. 49-50)

In line with Coffman's argument, we suggest that when actions or relationships are routine, established, and/or unproblematic, participants accomplish these actions or relationships without specifically having a concern for accomplishing them.

As an example. Ronny and Sheila interacted as friends with some common knowledge of each other's circumstances. Some of the behaviors that they enacted help sustain a sense of themselves as friends with common knowledge of each other's circumstances.

Ronny and Sheila used address terms for each other. Although address terms are available as "tie-signs" (Coffman, 1971, chapter 5) for outsiders to infer the relationship between the parties, Ronny and Sheila most likely would take for granted their ordinary address terms for each other. Doing so is part of achieving a sustained friendship.

One way in which we "do" having a stable relationship is to be able to produce appropriate address terms and references without their becoming objects of our attention. This should be regarded as an achievement-something to which participants pay some attention-however, it is usually unreflexively achieved in established relationships and is not usually a matter for concern. I t is generally assumed rather than being something to which participants pay active attention; however, there a r e circumstances where such taken-for-granted ac- tions become concerns. For instance, in a developing romantic relationship between individuals of unequal status, the transition to first-name address terms may be the object of considerable attention and concern. (cf. Carfinkel, 1967, for a detailed treatnient of the achieved character of social life).

In the conversation between Sheila and Ronny, first names were used in identifying the parties to the call. Ronny's self-identification as "Ronny" indi- cated his expectation that Sheila should have been able to recognize him from his first name. He addressed her as Sheila, claiming a n entitlement to call her by her first name.

The way in which Ronny made place references, for example, "the glen" and "the bank," also indicated his expectations regarding Sheila's knowledge. In telling the problem he was facing, he made reference to "the bank," that is, to a

place that Sheila could recognize as the one he was talking about. Selecting an appropriate reference, then, is a n achievement.

Sheila clid not respond for three tenths of a second, at which point he added, "a:t uh: (.) in BrentwoodYhh."

13 Ronny: A:nd. h h (0.2) 1 doll' know i f it's: po:ssible, but

I 4 . I ~ h h see I llaveta open u p the ba:nk.hh

16 Ronny: a:t uh: (.) in Brentwood?hh=

By expanding the reference, he treated her nonimmediate uptake as indicating her not having been able to understand the previous reference (Pornerantz, 1984). Whereas we are often unaware of the process of selecting reference terms tliat are appropriate to the knowledge presumed of the recipient, when trouble emerges we tend to become more aware (cf. Sacks & Schegloff, 1979, re: procedures for references to persons). That is, there a r e circumstances in which recognition of a taken-for-granted reference ("the bank") turns out to be p r o b lematic, and then gettiug recognition of the referent may become a concern.

The reference forms that Ronny used suggest the access Ronny and Sheila had (and were expected to have) to one another's worlds. These references provided Ronny and Sheila (and us, as analysts) with a loose sort of guide to the level and nature of their relationship. That is, by choosing these references, Ronny instantiated the nature of their relationship. However we would not say that this was something that they necessarily were consciously o r planfully doing or were actively concerned with. Rather, the selection of these place references would have been taken for granted and would only arise as an active concern in situations of uncertainty or trouble (cf. Schegloff, 1972 re: the organization of place reference).

UNMINDFUL ACTIONS

Questions about the exlent to which people a r e aware or conscious of their actions have been raised by several researchers (see, e.g., Langer, 1978, re: "mindlessness"; Berger & Roloff, 1980, re: "thoughtful" and "thoughtless" speech). Langer and others have proposed that we are able to perform actions mindlessly because, in some situations, we follow scripts for actions (cf. Benoit & Benoit, 1986, for a review and assessment of this position).

Langer (1989) described several different kinds of mindlessness. For our purpose here, we concentrate on the phenomena of individuals being unaware of what they are doing and/or how they a r e doing it. She wrote, "In the routine

162 MANDELB'UM AND POMERANTZ

of daily life we do not notice what we are doing urlless there is a ~)roblenl" (198'3, p. 43).

One circumstance in which an actor may be mindless is when he or she performs a task expertly. Langer argued that when we become experts, we assume that we can perform the given task, and we do perform it, yet we no longer know how we perform it: ". . . if we know a task so well that we can perform it 'expertly' (mindlessly), these steps may no longer be corlsciously available. . . ." (1989, p. 20). She argued that becoming conscious or mindful of each step may incapacitate the expert.

Another circumstance in which a person may be mindless is when he or she performs a sequence of routine activities. Whereas the person may well be aware of his or her actions while initiating the sequence, once started, he or she may go on "automatic pilot." Langer illustrated this with the following anecdote: "William James tells a story of starting to get ready for a dinner party, uiidress- ing. washing, and then climbing into bed" (1989. p. 43).

One kind of communicative behavior that frequently is repeated and cer- tainly may become routine is greetings. By closely examining greetings and openings, Schegloff (1986) has pointed out that even something as routine and often repeated as the opening of a telephone call is an achievement throughout. Although "mindlessness" suggests a n actor's lack of awareness on some level of an action that he or she performs, analysis of the details of those actions reveals their achieved character.

Unmindful actions a r e those we "find ourselves having done." Often, they are the unintended consequences of an action. An example may be when a speaker finds that he or she has insulted another without realizing it. Goffman (1959) referred to such actions as "faux pas." where a speaker, for instance, calls someone with a history of psychological problems a "nut" and does not realize until afterwards that it may be understood as being more than just playful. Similarly, people may come to hear unintended meanings in their utterances. The following example is from a field note. Upon warming meatballs a person said, "These meatballs aren't too hot." Apparently, upon hearing the two possible meanings of "hot," they added, "I should warm them up some more" so as to clarify which of the two available meanings they wish to direct their interlocutor to. The added unit of talk clarifies the possible double entendre of "hot," showing it to refer to temperature rather than quality. In this way, a semantic property of the word "hot" (it has at least two available meanings in this context) may result in our doing a n action (criticizing the taste) without realizing until afterwards that we have done it.

In the meatball illustration, the unmindful action became an object of atten- tion when its perpetrator showed herself to have noticed it. Often, though, unmindful actions may pass completely unnoticed. Some unmindful actions appear to be produced by characteristics of talk and the way these characteris- tics operate on s ~ e a k e r s . For instance, Sacks (1972) noted structural features of

9. WFIA'I' DRIVES SOCIAL ACTION? 163

talk that provide for the routine (Jccurrence of unintentional and often unnoticed puns in narrative terminations. Jefferson (1977) arlalyzed what she and Sacks called the "poetics" of talk, where subsequent words appear to be touched off by the sounds and meanings of prior talk. Although puns and "poetic" features of talk may sometimes be intentional many are unintentional and apparently not attended to. This suggests that some leatures of talk exert an influence on how interaction is done without participants being aware of them. Features of this kind can be subject to analysis by examining interaction.

CONCLUDING DISCUSSION

As the preceding discussion indicates, much can be inferred from the details of talk. From looking at the conversation between Ronny and Sheila, we were able to infer Ronny's primary concern and suggest how he shaped his behavior in order to achieve it. From features of the discourse, we inferred Ronny's and Sheila's contingent concerns that accompanied their primary concerns. We suggested the possibility of prerequisite concerns that are dealt with to enable participants to be in a position to act on their primary concerns. We showed how taken-for-granted sustained achievements may explain how some action comes about a r ~ d may become concerns at certain junctures. Finally, we saw how some actions may be unmindful and may become concerns only in their aftermath, if at all.

Of course, we did not have access to inside the participants' heads, particu- larly to their moment-to-moment states of awareness. Our concern is that, in our lay usage, the terms concerns, goals, and orientations conjure up a picture of the actors' awareness. We offer the preceding distinctions as a reminder that we should not overemphasize the role of awareness in the production of social action. As we have argued, some behaviors are explainable as directed to satisfying concerns, whereas others may be completely unintentional.

Our discussion of primary concerns, multiple contingent concerns, prerequi- site concerns, taken-for-granted sustained achievements, and unmindful actions suggests that, in order to understand what drives social action, we need to be alert to the details of interaction. Using these details, we (like participants) may infer various ways in which actioil moves forward. We attempted to infer concerns and other forces that influence how social action is shaped by at- tending closely to the details of talk. We had fairly high confidence regarding some of our inferences; with others, we offered mere possibilities.

The data-based distinctions we have been able to make between different types of actors' concerns, on the one hand, and different ways in which actors perform social action, on the other, suggest that i t is sometimes possible for conversation analysts to clarify their local uses of terms such as participants'