Conversation analysis and the study of social institutions

Aceptado: 05/12/2013 The objective of this study is to show how conversation analysis, a sociological discipline, approaches the study of social institutions. Social institutions are conceived as the crystallization of members’ communicative, interactional practices. Two institutional domains—psychiatric interviews and broadcast news interviews —and a specific interactional practice—‘formulations’—are examined in this study. The results show that (1) in psychiatric interviews the psychiatrist uses formulations to transform the patients’ avowals and establish a psychiatric problem. (2) In broadcast news interviews, formulations might help the interviewer to clarify or transform the statements of the interviewee, or challenge his assertions. The comparison of formulations in two different institutional settings serves the purpose of (1) demonstrating how communicative conduct is adapted in particular settings in ways that invoke and configure distinct social institutions and (2) inspect the knowledge, practices, logic, etc., mobilized by members of the epistemic communities of psychiatry and journalism.


Introduction
As conversation analysis (henceforth CA) has shown (e.g., Atkinson and Drew, 1979;Clayman and Heritage, 2002;Heath, 1981;Heritage and Clayman, 2010;Wilson, 2012) there is a clear connection between sequences of interaction and the social institutions in which they take place.In this paper I take pains in demonstrating how the interactional and institutional orders reflexively shape each other.Traditional sociological perspectives conceive communicative conduct as the result or product of socio-cultural elements like age, gender, organizations, institutions, etc.The usual research strategy of such approaches is to show how socio-structural arrangements condition language use in a causal fashion.Other sociological disciplines that derive from an interpretive tradition (e.g., symbolic interactionism or cognitive sociology) relate social structure and interactional organization in a dialectical or reciprocal manner: social structure determines and is determined by interactional organization.The perspective adopted by CA-clearly stemming from the ethnomethodological concern with the reflexivity of social action-is somehow different from the two mentioned above: our sense of order is the result of interactional practices and thus created in and through talk. 1 This perspective urges the analyst to approach the study of the interactional and the institutional domains from a non-causal viewpoint.Instead, the emphasis is on the procedural 'work' that members2 mobilize in making social institutions 'visible' when they encounter one another (Schegloff, 1991).In other words, A reflexive approach to language, action, and social structure, then, means understanding how sequential organization and other aspects of the interaction order can be deployed in ways that are sensitive to the contingencies and relevances of a society's organizational and institutional settings (Maynard and Peräkylä, 2003, p. 252).
It is possible, then, to study social institutions through the talk that is designed by participants to locally invoke them and hence, examine how these participants produce and reproduce the institutional order.
When we study social institutions through the interaction that occurs in them we typically come across two types of participants: professionals and lay persons.Professionals, generally speaking, may be said to belong to certain epistemic communities by virtue of their ascription to the body of knowledge, practices, logic, etc., they put to work in their activities. 3The inspection of their interactional routines has proved to be a very fruitful line of enquiry: it has revealed, for instance, not only the characteristics of their professional stocks of knowledge (Peräkylä and Vehviläinen, 2003) but also that their encounters with lay participants involve a variety of interactional asymmetries.John C. Heritage (1997, p. 175) has indicated four different types of asymmetries in institutional interaction: (1) participation; (2) 'knowhow' of interactional and institutional arrangements; (3) knowledge; (4) rights to knowledge.The first is related to the fact that the participants in professional-lay interactions may have a differential participation regarding the activities they perform.The second source of asymmetrical interaction derives from the distinctive perspectives that participants adopt concerning the nature of their encounter: professionals tend to treat their counterparts as 'routine cases' and clients/customers conceive their case as unique and personal.The third point has to do with the proclivity of professionals to act cautiously or, in other words, to avoid taking a resolved position on the issues that concern the lay participants.The final point is intended to propose that while professionals are largely entitled to the 'right to know', lay participants seem to often have this right denied.So, as Heritage points out, "in institutional interaction, then, knowledge may not be enough; one must also be entitled to the knowledge, and have come to it in an appropriate way" (Heritage, 1997, p. 179).The analysis of interactional asymmetries is, then, particularly pertinent since it might help to exhibit how specific bodies of practical knowledge emerge and are brought into play in institutional encounters.
In this article I analyze the communicative conduct that takes place in two different institutional settings: the psychiatric (intake) interview and the news interview.A psychiatric intake interview is a variant of the psychiatric interview in which the psychiatrist's official task is to determine whether a person should be-voluntarily or involuntarily-hospitalized as a patient on the basis of the person's observable behavior during the interview.On the other hand, the news interview is a relatively common feature of television news shows in which a prominent public figure-usually a politician-is interviewed by the news anchor.The news interview has become a major vehicle for presenting broadcast news and political commentary, and a crucial intersection between the institutions of journalism and politics.Hence, it has become a prominent object of study for scholars in the social sciences in disciplines like communication, sociology, politics, etc.
In the following sections I first describe the methodological approach of CA, then explain how CA can study institutional interaction-and thus social institutions-and finally examine how an interactional device-'formulations'-is distinctively used in two diverse institutional settings: psychiatric interviews and news interviews.The purpose of this comparison is to show how language use is adapted in particular settings in ways that invoke and configure distinct social institutions.

The Methodology of Conversation Analysis
First developed by Harvey Sacks, CA is ethnomethodology's cognate discipline that takes conversations in real-life settings as the object of study.It starts from the premise that conversations are one of the central activities of social life and that through them a great deal of our social interaction is organized.Although CA remains the name for this discipline, the more comprehensive term of talk-in-interaction is at present preferred, since it seems to be inclusive of other interactional activities as well (e.g., non-vocal activities).Furthermore, the term conversation is heavily associated with the notion of informal talk, to which CA is not exclusively restricted.Given its interest in the study of all forms of social interaction, CA has developed into a discipline that aims to investigate all types of discursive orders.
CA sets out to record patterns within conversations in order to detect the underlying rules that enable communication to proceed in a largely orderly fashion.Special attention is paid to the sequential organization of interactional activities.It preferably employs audio or video recording techniques in order to preserve the naturally occurring attributes of interaction.
The first step to produce a study using CA is recording some interactional event/s.Once the recording has been made the conversation analyst characteristically produces a transcription of the interactional phenomena s/he has recorded.One of the obvious advantages of transcribing as a methodological principle is that any findings can be checked upon and, eventually, challenged.A further advantage of this process is that it helps in the practice of identifying interactional details that would go unnoticed.
One of the first discoveries of CA was that interactions proceed on a turn-by-turn basis, that is, they are sequentially organized.Hence, a turn at talk is understood as an action shaped by a previous utterance or utterances and what a speaker does in his/her turn.The early focus on 'adjacency pairs' proved that interactions were made possible through the availability for members of some kind of turn-taking system.Consider the following example: A: John?B: What?This is an example of an adjacency pair-in particular, a summons-answer adjacency pair.Although these utterances are produced by two different speakers, they form a unit; that is, they seem to have some kind of structure in that: (1) they are two turns in size; (2) there is a change of speaker after the production of each turn; (3) what is said in the first turn is relevant for the producer of the second turn; (4) what is said in the second turn is related to what is said in the first one.Their constant presence in conversation seems also to indicate that the production of the first part of the pair by a speaker normatively requires the production of a juxtaposed second action by another speaker.This fact has several consequences.It first implies a methodical basis on which the participants to a conversation could check whether specific next actions were recognizably absent.Second, it can also provide members with a tool to discriminate subsequent actions as misaligned when heard as unrelated to preceding utterances.Third, it sets the grounds on which the producers of unrelated or 'deficient' responses could be rendered normatively accountable.Finally, the discovery of juxtaposed actions in interaction proved to be important because: (a) it suggested the presence of a wider structure available for interactants; (b) it shows how members can display, in their next actions, a public understanding of the preceding talk; (c) it could demonstrate how shared meanings are created and sustained.In general, then, it furnishes members with a systematic apparatus to exhibit and bring up-to-date their public understandings of each other's talk (Heritage, 1988, p. 25).It also allows us to consider adjacency pairs as 'the building-blocks of intersubjectivity' (Heritage, 1984, p. 259).
The importance of the adjacency pair structure for the analysis of conversation does not stem from its massive presence in talk-in-interaction, but because the form of analysis addressed to these paired utterances is applied to more intricate and longer sequences of talk (Heritage, 1984, p. 253).In turn, this form of analysis is sociologically relevant because it shows that sequential organization is constituted of members' social actions and provides an environment for them to understand, appreciate and do things in and through talk.It can also show that: Social interaction is informed by institutionalized structural organizations of practices to which participants are normatively oriented.[…] these organizations of practices […] are fundamentally independent of the motivational, psychological or sociological characteristics of the participants.Rather than being dependant on these characteristics, conversational practices are the medium through which these sociological and psychological characteristics manifest themselves (Heritage 1995, p. 396).
If the above considerations are correct, we can appreciate that members have at their disposal a complex machinery that enables them to communicate with each other.The task of the conversation analyst is one of providing a formal description of that machinery while avoiding the production of abstract generalizations.Proceeding in this way, CA has turned into a discipline that can claim to have produced a cumulative body of findings.
In sum, the tradition of CA assumes a set of premises: (1) the primacy of ordinary conversation; (2) the use of naturally occurring recorded data; (3) an inductive, structural analysis of interactional practices.This form of research yields insight into the production of social actions by members in their daily, communicative routines, while retaining an interest in prominent sociological concerns like the normative character of, and the interpretive procedures involved in, social action.

CA and the study of Institutional Interaction
Almost since the inception of CA, many researchers within this discipline have dedicated a great number of their studies to the investigation of so-called 'institutional interaction/talk'.The interest in researching institutional interaction may have derived from three distinct concerns: first, it proved to be an adequate approach to study certain 'forms of talk'; second, it represented a way to accomplish a 'rapprochement' with the epistemological issues debated within orthodox sociologies (Schegloff, 1991, p. 45); finally, it would lead to a better understanding of social institutions (e.g.psychiatry, education, media, etc.) and a whole variety of epistemic communities (e.g.psychiatrists, journalists, teachers, etc.).
CA researchers dedicated to the study of some form of institutional interaction have discovered that some of the communicative phenomena analyzed may apparently be connected to the specific settings in which they take place.In particular, it has been noted that such settings-like, for instance, classrooms, the courts, medical settings, etc.-entail patterns of interaction that diverge somewhat from those found in ordinary conversation.Therefore, it might be suggested, context and communication are inextricably interrelated.However, that does not mean that the notion of context is exclusively related to the milieu where interactions take place.It includes other matters as well.In effect, different disciplines conceive of context in a variety of ways: for instance, as linguistic contexts, interactional contexts, social structural contexts, cultural contexts, etc.However, such disciplines may view the relationship between context and conduct as an opposition or dualism; in particular, context and conduct relate to each other in a one-directional way; namely, the contextual configuration determines conduct but not vice-versa.CA, instead, considers that context and conduct mutually constitute each other.CA also questions the view that conceives these two concepts as standing in opposition.Furthermore, the notion of context is regarded as being multidimensional: the setting where the interaction takes place, the linguistic resources mobilized, or social structural elements are all contextual dimensions susceptible to be studied.CA studies context and its dimensions as they are formulated and, therefore, produced and reproduced by members within talk-in-interaction.In other words, context is not external or extraneous to conduct; context and conduct reflexively configure each other.Therefore, the analysis of context should have as a central concern the analysis of face-to-face interaction; and this is so because First, face-to-face interaction provides the primordial locus for the production of talk.The features of face-to-face interaction thus constitute a primary exemplar of context.Second, in so far as face-to-face interaction is accomplished through the collaborative work of separate individuals it provides an elementary example of human social organization.[…] Third, within interaction participants are faced with the task of accomplishing understanding and, as part of this process, displaying to each other their understanding of the events in progress at a particular moment.[…] Finally, face-to-face interaction is inherently dynamic.Each subsequent utterance, and indeed events within a single utterance, change in subtle but profound ways the operative context of the moment.[…] Face-to-face interaction thus provides an opportunity to analyze language, culture, and social organization as integrated components of a single system of action (Goodwin and Duranti, 1992, pp. 22-23).
What I have tried to propose in the preceding paragraphs is a notion of context that is considered in terms of a members' local, endogenous and oriented-to accomplishment in any given setting.I have also tried to portray language use as a contextualizing and contextualized instrument.Consequently, if we are to show the institutional character of some interactional event, we should demonstrate in what ways the institutional context is constituted as such.That can be done by means of a comparative analysis.In effect, CA notes that in our everyday life, ordinary conversation is the paramount and generic vehicle for communicating.In this respect, ordinary conversation represents a kind of 'benchmark' against which other institutional forms of interaction can be compared (Wilkinson and Kitzinger, 2008, p. 60).As a result, it is supposed that these institutional forms of interaction, when contrasted with ordinary conversation, will show systematic differences and restrictions on activities and their design.
Harvey Sacks, Emanuel A. Schegloff and Gail Jefferson (1978) were the first to note the importance of a comparative analysis.In their seminal article they observe that turn-taking systems (among which ordinary conversation is one type) could be seen as forming a linear array in terms of their allocational arrangements: The linear array is one in which one polar type (which conversation instances) involves 'one turn at a time allocation'; that is, the use of local allocational means, and the other pole (which debates instance) involves 'preallocation of all turns,' and medial types (which meetings instance) involve various mixes of preallocational and allocations means (Sacks et al.,p. 46).
Having said that, Sacks, Schegloff and Jefferson, warn us about the dangers of treating speech-exchange systems as discrete communicative structures: We should not be understood thereby to be proposing the independent, or equal status of conversation and ceremony as polar types.For it appears likely that conversation should be considered the basic form of speech-exchange system, with other systems on the array representing a variety of transformations on conversation's turn-taking system to achieve other types of turn-taking systems.In this light, debate or ceremony would not be and independent polar type, but rather the most extreme transformation of conversation (Sacks,Schegloff and Jefferson,p. 47).
Similarly, we would not analytically profit from the production of hard and fast distinctions between the institutional and the non-institutional domains.Instead, a development of "family resemblances among cases of institutional talk" is preferred in order to determine the 'institutionality' of some interactional events (Drew and Heritage, 1992, p. 21).This should lead to a mode of analysis that would exhibit how the participants' conduct is institutionally oriented, illustrating the progressive (i.e., turn by turn), collaborative and negotiable, constitution of the institutional context in which the participants find themselves (Drew and Heritage, 1992, p. 21).
Institutional talk can be analyzed in different settings.Interactions in what has come to be termed 'formal settings' exhibit specific features.In particular, turn-taking systems in formal settings tend to substantially differ from ordinary conversation.By contrast, interactions in 'non-formal settings' tend to resemble those which take place in ordinary conversation.
It has been observed that the participants in formal settings such as courtrooms, classrooms, news interviews, etc., tend to methodically organize their conduct so as to exhibit and produce the 'institutional' quality of their encounter (Drew and Heritage, 1992, p. 26).Furthermore, their conduct is constrained by a diminution of the array of options and possibilities for action that usually take place in ordinary conversation.These variations from ordinary conversation may supply the analyst (and, in the first place, the participants) with the basis on which to identify each form of institutional talk.Or, as Paul Drew and John C. Heritage indicate, these variations may contribute to distinguish a unique 'fingerprint' for each speech-exchange system.(Drew and Heritage, 1992, p. 26).
Other settings-like medical or psychiatric environments, social services, etc.seem to comprise an interactional configuration less rigid than in more formal settings.Even though interactional asymmetries-as in formal settings-may be found, these do not appear to be the consequence of normatively instituted sanctions (Drew and Heritage, 1992, p. 28).Therefore, the non-formal character of the interactions which take place in these settings has to be appreciated in that: (a) they occur in private rather than public contexts; (b) there is room for negotiation as to the course the interaction will follow; (c) their turn-taking systems are similar to ordinary conversation.The question that remains is: if they are so similar to ordinary conversation, how can we distinguish their different turn-taking systems?Drew and Heritage suggest that in non-formal settings it is quite improbable to find specific interactional sequences, which can be indicative of its institutional character.Therefore, its institutionality will have to be Located in a complex of non-recursive interactional practices that may vary in their form and frequency.Systematic aspects of the organization of sequences (and of turn design within sequences) having to do with such matters as the opening and closing of encounters, with the ways in which the information is requested, delivered, and received, with the design of referring expressions, etc. (Drew and Heritage, 1992, p. 28).
In conclusion, CA approaches the study of social institutions in a similar vein as Georg Simmel: social institutions crystallize in the actions (and interactions) of members.CA stresses that social reality is an interactional accomplishment, and thus constructed by actors rather than being an established, antecedent structure.
When people talk, they are simultaneously and reflexively talking their relationships, organizations, and whole institutions into action or into 'being' (Boden 1994, p. 14).
Socio-cultural structures, then, should be understood as the product of people's actions rather than as simple 'containers' of those actions.When studying institutional interaction, conversation analysts bear in mind the reflexive character of interaction and social structure.By inspecting the interaction that takes place in institutional settings, CA has also been able to elucidate the practices, logic, knowledge, resources, etc., mobilized by members of certain epistemic communities (i.e., professionals) when dealing with lay and other participants.In the following section I describe one of such practices ('formulations') and subsequently analyze how it is distinctively used in two different settings: psychiatric interviews and news interviews.

Formulations
Let us consider the following (invented) example: A: Yesterday I quarreled with John.
B: Do you mean that you'll be separating?
What we can appreciate in this example is that: (1) B 'formulates' A's turn, i.e., B provides an interpretation of what A has just said; (2) B's turn should generate and adjacency pair whose relevant next action must be a decision; (3) That decision can be a confirmation or a disconfirmation; (4) B's formulation engenders an inspection of the topic/s of the conversation.
The above example shows that formulations (like for example, B's utterance) are a crucial feature for the achievement of intersubjective understanding in talk-in-interaction.As Harold Garfinkel and Harvey Sacks (1970) taught us, formulations are used by members as the ventures in conversation where they can display to one another their understanding of the sense of the talk-thus-far.Thus, formulations-which, in general, can be understood as an interactional device or practice-generate an examination of the conversational materials-topics, troubles, etc.-that have been a part of the conversation until 'now' (Heritage and Watson, 1979, p. 128).In this respect, they operate to exhibit apprehension.Other conversational devices, like repeat utterances (i.e., repair), are also designed to achieve understanding.However, as Heritage and Watson point out, formulations do not display the typical equivocality that is associated with repeat utterances.That is so because formulations involve a certain transformation or paraphrase of some prior utterance/s, and therefore "they preserve the relevant features of a prior utterance or utterances while also recasting them.They thus manifest three central properties: preservation, deletion and transformation" (Garfinkel and Sacks,p. 129).
Formulations have been characterized as operating at three different structural levels: utterance-by-utterance, topic organization and overall organization of the conversation.
At the utterance-by-utterance level, it is notable that formulations appear adjacently paired.In this sense, they generate a short sequence whose relevant next actions are 'decisions.'In particular, decisions typically take the form of 'confirmations' or 'disconfirmations' (Heritage and Watson, 1979, p. 141).As Heritage and Watson suggest, there is an intrinsic preference for confirmation over disconfirmation.That preference is founded upon the fact that plain disconfirmations could question the coconversationalist's capacity of (e.g.) attention to, or comprehension of, the talk so far-or, in other words, his/her competence as a member.Hence, it is not surprising to find disconfirmations of formulations conjoined with confirmatory elements (Heritage and Watson, 1979, pp. 144-146).
At the level of topic organization, since formulations constitute a gloss of what the conversation (or parts of it) has been about so far, it is clear that they must play an important role in the management of topic organization in the conversation.In effect, they may be used to align speakers to the same topic, control the course of the conversation, or collaboratively decide the importance or significance of a given topic or of the whole conversation (Heritage and Watson, 1979, pp. 149-151).
Finally, at the level of the overall organization of the conversation, formulations may serve the purpose of closing down the whole conversation.Since formulations may be used to close a topic, they may also be employed to close down a conversation as long as the topic so closed is the last topic of the conversation.In doing so, formulations also function to establish the preservable and reportable elements of a conversation (Heritage and Watson, 1979, pp. 155-6).
Overall, interactants may formulate the sense or gist (i.e., a summary) of what has been said (Heritage and Watson, 1979, p. 130), or produce an upshot (i.e., an implication or unexplicated version, as in the above example) of the conversation (Heritage and Watson, 1979, p. 134).
In comparing the use of formulations in institutional settings and ordinary conversation, Paul Drew notes that formulations are rarely found in ordinary conversation (Drew, 2003).According to Drew, this is so because "we do not need to arrive at compromises after long negotiations in mundane conversation" (Drew, 2003, p. 306).Further, in his study of formulations in a variety of institutional settings (psychotherapy, call-in radio programs, news interviews and industrial negotiations) these are taken to constitute the means for the management of these activities.
"We may engage in [doing formulations] in conversation from time to time, but they are not the kinds of routine, organizationally salient activities that they are for the settings discussed here" (Drew, 2003, p. 306).
From these comments, we may understand that the practice of formulating may be distinctively shaped according to particular communicative contexts like the psy-chiatric interview, the news interview or ordinary conversation.In the following analyses I describe how.

Data and Method
Data in this study form part of two different research projects.First, a study of political interviews which involved the recording of 20 interviews between prominent Spanish broadcasts journalists and politicians.The interviews studied were televised in Spain between 2006 and 2010 on private channels and public service broadcasters.Second, a study of psychiatric interviews consisting of eight sessions, with two different psychiatrists and eight various patients.The psychiatric interviews were videorecorded in a large Spanish hospital in 1999 for the purpose of my doctoral research.
The method of analysis is that developed in CA.CA's empirical tradition is particularly suited to study news and psychiatric interviews since it focuses on firstly, reallife instances of interactions and secondly, uncovering the processes and procedures which make the psychiatric interview or the news interview distinct forms of institutional interaction.
The two data extracts presented that support my analysis have been selected as examples that are not statistically representative of the interactional practices examined below.The patient, psychiatrist, broadcast journalist, politician or sequences, which constitute the object of this study, have not been selected as representative of the institutions of psychiatry or broadcast journalism.Rather, they have been chosen to reveal the accomplishment of particular institutional goals: the use of formulations to delimit a therapeutically-relevant problem in psychiatric interviews or broadcast journalists' employment of formulations to probe politicians' answers.

Formulations in Psychiatric Intake Interviews
A psychiatric intake interview is a variant of the psychiatric interview in which the psychiatrist's official task is to determine whether a person should be-voluntarily or involuntarily-hospitalized as a patient on the basis of the person's observable behavior during the interview.Customarily, this implies that the psychiatrist has to solicit the patient to talk about the problems that brought him/her to hospital.As we shall see below, this undertaking is accomplished through the psychiatrist's invitation of the patient's story and his/her inception of a variety of questions that demand particular topic elaborations.Once a problem has been disclosed by the patient, the psychiatrist's next assignment is to ascertain its psychopathological relevance.For this proce-dure to take place, obviously, the patient's lay interpretation of his/her problem has to be transformed.A useful psychiatrist's resource, as we shall see, is the use of formulations.(Patient 1, psychiatric interview, February15, 1999). 5he elicitation of a biographical account We may appreciate in the above data fragment how the 'reason of the conversation' is properly initiated by T1 in lines 1-2 with "Bueno pues (nos) explicas un poquito qué:: qué es lo que ha pasado por que te ingresaron ayer" ["Well so can you explain (to us) a bit wha::t what happened why you were admitted yesterday"] after some (omitted) preliminary talk.The kind of sequential structure that T1 is eliciting with her turn in lines 1 appears to be a 'story'.As Harvey Sacks (1992a;1992b) showed us, a story is a particular kind of collaborative narrative produced in conversation.Stories can be either volunteered or invited, and what they share in common is that: Firstly, each preface works to signal that a story is upcoming or "on offer", and to achieve coordinated entry by all participants into a bloc of talk embodying a narrative format (though there may exist variations in such a format), a major part of which involves the casting of interlocutors into the ca-pacities of teller and recipient(s) with appropriate distribution of participation rights.Furthermore, each story preface provides recipients with a built--in motive for listening to the story-recipients can and must listen for the materials prefigured in the preface.Another dimension of this built-in motive for listening is that the recipient(s) can adjudge whether the story is yet to be completed (Watson, 1990, pp. 275-276).
The following three-part is typical of volunteered stories (Sacks, 1992b): •Teller: story preface •Recipient: request to hear the story •Teller: story In contra-distinction to volunteered stories, invited stories are recipient-initiated stories (Watson, 1990).In other words, the recipient of the story provides the preface (or first utterance), whereby the materials produced by the putative recipient should be integrated into the putative teller's story.The putative teller, then, after s/he has been invited, might accept or decline to narrate the story s/he has been requested to produce.Thus, an invited story appears to be adjacently paired: the recipient invites the teller to produce a story through the provision of a story preface, and the teller either accepts or declines.The fact that this structure is adjacently juxtaposed also implies that some normative pressure is placed on the producer of the second action-i.e., the teller of the story.
Further differences between these two types of narratives can also be indicated.An obvious dissimilarity is that the materials of an invited story are furnished by the recipient.This element allocates the recipient of the invited story a wide margin of control over what can be said in the story (Watson, 1990, p. 276).Another difference is that the intercalations that the recipient is allowed to introduce into the narrated story are typically not restricted to 'response tokens'-the recipient also seems to be entitled to request story expansions by introducing (e.g.) questions.Having said that, the recipient's margin of control is also constrained: his/her intercalations should be perceived as 'on-story', that is to say, in consonance with the story preface.Further, the fact that an invited story is the teller's story, confers him/her-although only to a certain extent-the possibility of including materials s/he might consider as appropriate (Watson, 1990, p. 276).Finally, in a volunteered story the teller does not seem to face the interactional 'problem' of securing the floor to produce the story, since s/he is prompted to do so.
In the above data fragment, then, we may observe how T1 invites P1 to produce a story (lines 1-2).Of course, such an observation can be achieved by the fact that P1 shortly afterwards initiates the development of a story, which corroborates the contention that stories are collaboratively produced sequential structures.Thus, from line 7 we may appreciate that P1 has accepted to produce a story since she is furnishing one to T1.However, before that happens, P1 asks in line 4 "O sea lo que pasó realmente" ["So what really happened"].I indicated above that recipients (i.e., inviters) of invited stories have an increased margin of control over the story to be told by the teller, since they ask for a particular kind of a story that tellers have to estimate.Thus, the teller of the story might find it problematic to ascertain the kind of story that recipient might be requiring.In this respect, P1's utterance in line 4 (after a long pause of 2.5 seconds, which emphasizes her hesitation) might be heard as involving a check (or 'repair') on what an 'appropriate' story to T1's preface might be like.T1 then confirms that what she wants to hear is "what really happened".
Let us focus now on T1's invitation in lines 1-2.Such invitation takes the form of a question, which is a first pair part of an adjacency pair.Hence, perhaps, P1 might be seen as being doubly normatively constrained in the production of her subsequent action: in that her action is the second pair part of an adjacency pair (an 'answer'), and because T1's invitation somehow forces her next action to be of a specific type, namely, a particular kind of story.Notice as well that T1's question/invitation might be heard-and I am referring here, in the first place, to the recipient of the question/invitation-in several different ways.For instance, an appropriate answer could be 'because somebody brought me here', with more or less expansion, which would also fulfill the requirements of telling a story.However, as we may see in P1's incipient story, she hears T1's question (after some hesitation) as one that demands a biographical account of her life experiences.Having said that, we may appreciate that it is not just any biographical account, in the sense that such an account could be developed in the form of (e.g.) 'I was born in such place, in that year, and in the context of such a family'.It is a biographical account that expresses a contextual orientation 'to this place, on this occasion', and therefore, a set of 'problems'; in particular, as we can see, these can be glossed as 'marital problems'.

Using formulations to transform the patient's avowals
There has been a widespread interest in CA for the study of formulations in mentalhealth consultations.The pioneering study was that by Howard Schwartz (1976), which showed that formulations may be used by therapists to display the psychotherapeutic value of their own interpretations.Several recent studies are consistent with this finding (Antaki, 2008;Antaki, Barnes and Leudar, 2005;Peräkylä, 2004;Vehviläinen, 2003).On the other hand, formulations have been shown to be appropriate devices for the purposes of history-taking, selecting a candidate problem and establishing it as therapeutically relevant (Antaki et al., 2005;Davis, 1986;Hak and de Boer, 1996;Halonen, 2006;Vehviläinen, 2003).Formulations, also, have been described to be employed by patients to elaborate on the therapist's interpretations (Peräkylä, 2005).In general, these studies show that formulations might be used by professionals to exhibit the significance of certain elements of the patient's own accounts.
As we may observe, a chain of questions-which have been omitted for reasons of space-with the purpose of history-taking are initiated in line 19; after about two minutes P1 resumes her story from line 22.As it can be appreciated, the telling of her story is interrupted by T1 with an upshot formulation in lines 41-43 ("Entonces has trabajado de dependienta (básicamente) y tu madre e: y tu marido de diseñador" ["So you have worked as a shop assistant (basically) and your mother er and your husband as a designer"]), in which T1 self-corrects her initial miss-identification of 'mother' for 'husband'.This two-part formulation is accepted by P1 in lines 42 (her characterization as a shop assistant) and 44 (an inferred observation of her husband as being a designer).As we may understand, this formulation (and its subsequent acceptance) functions to fix a particular type of a problem, namely, P1's avowed marital problems.However, it also delimits the problem: it is now a case of subordination to her husband.This delimitation, though, has required certain transformations: the owner of a business (line 22) and an 'artist' (line 32) has been turned into (or better downgraded to) a 'shop assistant' '(basically)' (line 41).Likewise, P1's husband (a 'jeweler', line 28) has been turned into (or upgraded to) a 'designer' (line 43).As a result, through the transformation and establishment of specific professional hierarchies, T1 has been able to transfer her proposed suggestion of inequalities to the sphere of P1's marriage.

Formulations in the news interview
In the following extract I analyze a fragment of a news interview between the journalist Iñaki Gabilondo and the leader of the opposition party (in 2006) Mariano Rajoy.The interview was broadcast on the 14 th of September 2006 on the Spanish television channel Cuatro (Gabilondo and Izquierdo, 2006).

3
Pues que estaría en pleno vigor el pacto por las libertades y contra el Well that we'd have in full force the pact for liberty and against 4 terrorismo que firmamos en la anterior legislatura .hhhque era un terrorism that we signed in the previous legislature .hhhwhich was a 5 pacto que le decía a ETA simplemente dos cosas .hhoiga aquí estamos pact that simply told ETA two things .hhlisten here we are 6 el partido popular y el partido socialista que se supone son los únicos the people's party and the socialist party supposedly the only 7 que podemos gobernar España al menos en los próximos años (0.5) y likely parties to govern Spain at least in the coming years (0. Heritage notes that news interviewers (IRs) might use three types of formulations (Heritage, 1985, pp. 105-109): (1) Prompt (''to prompt interviewees to reconfirm and elaborate prior remarks''), (2) Cooperative recycle (''an interviewer may accurately and agreeably re-present the interviewee's stated position'') and ( 3) Inferentially elaborative probe (''a formulation is used to test or probe some aspect of an interviewee's actions, intentions, or attitude'').The inferential probe is often used by IRs when interviewing politicians.Many politicians provide unclear or evasive answers.Using an inferential probe allows the IR to (a): clarify to the audience (and IE) what the IE might actually mean, and (b): give a renewed topical focus while probing issues from the prior turn(s).As we shall see in the following analysis, the type of formulation that the IR is proffering in line 17 is the latter one.
Let's focus our attention on the question that IR asks in line 17, which produces an upshot or implication of the previous turn.As we can see, after having started a new topic (line 1) referring to a hypothetical situation-what the interviewee (IE) would do regarding ETA and the Basque Country if he was the prime minister-and after IE's elaborate answer, IR formulates the content of the previous turn.The implications of this formulation are threefold: (a) this formulation preserves the content of the previous turn as an object for further elaboration.Hence, it can be appreciated how the previous turn is topicalized and a new topic is not started, demonstrating the importance of formulations at the level of topic organization.It also allows us to see how this formulation works at the utterance-by-utterance level: generating an adjacency pair that projects a decision by the recipient.(b) It selects a particular element of the previous turn, which is 'proposed' for confirmation.The selection of a single element obviously involves the elimination of other aspects that do not appear in the formulation of the previous utterance.In effect, IE's long turn (lines 3-16) contains a variety of elements (the policy implemented to curtail ETA's activities, the cooperation between the two main parties, the absence of political negotiations with ETA, the effectiveness of the pact, etc..) that are not incorporated in IR's formulation.This formulation seems to select one of the elements mentioned in IE's turn: ETA's conduct in connection to governmental policy on terrorism.(c) It transforms or modifies, inasmuch as it paraphrases-in this case, an upshot or implication-the content of the previous turn.Thus, in the analyzed fragment, we may observe that "la única alternativa que tiene usted es dejar las armas y yo no voy a negociar políticamente nada con usted" ["the only choice you have is to give up your weapons and I'm not politically negotiating anything with you"] (lines 8-10)-i.e., ETA's abandonment of armed struggle accompanied by its own dissolution as a result of police pursuit and not political negotiation-is transformed in IR's formulation in line 17 as "Es decir que no habría ahora una tregua de ETA" ["So there would not be an ETA truce now"]-understanding by truce a temporary absence of terrorist activity, with no definite abandonment of armed struggle, but that could lead to political negotiation.In this sense, it might be interpreted that what IR is suggesting about IE's turn is this: if the IE was the current prime minister, as a result of his hypothetical policy on terrorism-exclusively based on police persecution-there would not be the right political and social climate for a truce which, in turn, could encourage the political negotiation of the conflict.
As can be seen in the transcript, IR's formulation is refused by IE (line 19), but occurs after a pause (one second) attributable to IE himself.The presence of this pause is quite relevant.Since in this type of sequence there is an intrinsic preference for confirmation over rejection, this pause mitigates the effects that a straight rejection could generate.Still, it is a rejection, and that prompts IR to transform his previous formulation in lines 20-22, which is again rejected-even more emphatically-in line 21, and after some considerable development, in the lines 28-30.
In general, the rejection of formulations generates what in AC is referred to as disagreement sequences (Pomerantz, 1984). Interlocutors, commonly (i.e., normatively), try to avoid their occurrence or at least limit their duration.This is so because they generate situations of conflict and, therefore, discomfort.And this is precisely what we can find in line 32: after two unsuccessful formulations, IR now tries with a direct question that reduces the scope of the previous formulations.He is now asking whether the truce-understood as the potential benefits of the absence of terrorist activity-is a worthwhile possibility, to which the respondent no longer offers resistance but vehement agreement (lines 37, 39 and 42), thus ending the sequence of disagreement.This agreement, however, has been achieved because both parties have given in.IR has reduced its original formulation to the obvious benefits of a situation of truce; IE has considered the hypothetical possibility of exploring a truce which, in the Spanish political context, could eventually lead to a political negotiation of the conflict.

Conclusions
The previous analyses show that formulations have distinct interactional functions in the two examined settings, i.e., participants seem to manage different institutional activities when they 'formulate'.As we have seen, a psychiatric intake interview is a type of institutionally-oriented-to encounter wherein patients' problems have to emerge.A useful resource that T1 has drawn on to elicit P1's talk of problems has been that of inviting her to narrate her story.Once she got P1 to talk, through the use of formulations-and other interactional devices not analyzed in this paper, such as repair or assessment sequences-she has been able to transform P1's own characterization and, consequently, select and establish a problem.In this respect, T1's candidate and progressive proposal of a therapy-relevant problem has been accepted and, hence, 'established'.This process suggests a somehow distinct use of formulating in this particular setting: it has helped the psychiatrist, with the patient's accord, to delimit a therapeutically-relevant problem.
Formulations seem to manage distinct interactional activities in the news interview.For IR, straight adherence to a journalistic style of impartiality and neutrality would imply, at the interactional level, a mere succession of questions.IR's role would be limited to simply providing subject headings that IEs could easily answer.To avoid the effects of such interviews-i.e., stillness during the interview and the resulting boredom in the audience-it is not surprising that journalists resort to devices such as formulations, which are somewhat like comments to IE's answers.Its use allows IR to attain certain objectives: clarify, transform, and propose alternatives to the statements of IE, or challenge his assertions.In short, they are useful to reach a more penetrating, flexible, dynamic and lively journalistic style which, in turn, can better engage the ultimate recipient of the interview: the audience.Furthermore, given the sequential properties of formulations, they seem to accomplish another practical purpose: they may allow IR-as we have seen in the above data extract-to pin down IE when the latter tries to avoid answering certain questions: IE must confirm or reject-but ultimately take a position on-what has been formulated.The analysis of formulations in two different settings has illustrated how CA approaches the study of social institutions.CA does not obviously aspire to explain every institutional aspect, but it shows a path to throw light on the intricacies of members' constitution-through their face-to-face encounters-of social institutions as such.If in ordinary conversation formulations seem to be designed to accomplish understanding, in psychiatric interviews and news interviews the practice of formulating is associated with different central institutional activities in each setting: delimiting a therapeutically-relevant problem and probing IE's answers, respectively.What this shows is that members enact, in their (seemingly irrelevant) discursive practices, the very social institutions that are an integral part of the social structures.The inspection of concerted social action can, therefore, exhibit the (seemingly invisible) features of social institutions.
When we study institutional interaction and the thereupon enacted social institutions we also have the opportunity to scrutinize the details of particular orders of knowledge and reasoning.For instance, in psychiatric interviews the knowledge of the two parties-psychiatrist and patient-may be said to lie at the ends of two opposite poles by virtue of their respective memberships.Thus, whereas the psychiatrist is assumed to be a knowledgeable party in dealing with matters of the mind, the patient is supposed to be a member of a category of people whose cognizance is taken to be defective.Given these considerations, such an asymmetry of knowledge may entitle the psychiatrist to determine those aspects of the patient's conduct susceptible of examination and modification.Furthermore, it appears to be in the patient's best interest to accept the psychiatrists' actions 'on his/her own behalf'.On the other hand, in broadcast news interviews, journalists' work is informed by a set of presuppositions about how to (e.g.) get IEs to talk or challenge their statements, while constituting the audience as the final recipient of information.Both institutional domains thus instantiate specific forms of knowledge that can be accessed through a detailed analysis of discursive activities.
CA, as I hope to have demonstrated, is a sociological discipline that can productively contribute to study social institutions.Through the analysis of situated language use, it can show how social institutions transpire and manifest themselves in the actions of professional and lay participants in institutional interaction.It can also help to delimit specific epistemic orders associated with specialized professional communities.Above all that, though, it can expose the structures of social action.
A: It's got four stories?Yeah B: It does!

Emphasis
Capitalization displays various forms of emphasis, including volume and/or pitch, and so on: A: She had a LOT of books

Aspiration
An 'h' marks audible breathing.The more 'h's', the longer the breath.A period preceding indicates 'inbreath'; no period denotes 'outbreath': A: .hhhhhh I guess so 8. Transcription conveniences Materials in double parentheses indicate difficult-to-transcribe vocal sounds, features of the setting, or characteristics of the talk: A: Well ((cough)) I don't know ((microphone noise)) B: ((whispered)) Neither do I 9. Inaudible utterances or words Materials in single parentheses indicate transcribers are not sure about words contained therein.If no words are within the parentheses, this indicates that talk was indecipherable to the transcriptionist: A: (Is that right?)B: (Perhaps) I am not sure