El potencial del análisis conversacional al estudio de la psicoterapia

23
http://tcp.sagepub.com/ The Counseling Psychologist http://tcp.sagepub.com/content/29/3/413 The online version of this article can be found at: DOI: 10.1177/0011000001293006 2001 29: 413 The Counseling Psychologist Anna Madill, Sue Widdicombe and Michael Barkham The Potential of Conversation Analysis for Psychotherapy Research Published by: http://www.sagepublications.com On behalf of: Division of Counseling Psychology of the American Psychological Association can be found at: The Counseling Psychologist Additional services and information for http://tcp.sagepub.com/cgi/alerts Email Alerts: http://tcp.sagepub.com/subscriptions Subscriptions: http://www.sagepub.com/journalsReprints.nav Reprints: http://www.sagepub.com/journalsPermissions.nav Permissions: http://tcp.sagepub.com/content/29/3/413.refs.html Citations: at University of Leeds on August 1, 2011 tcp.sagepub.com Downloaded from

Transcript of El potencial del análisis conversacional al estudio de la psicoterapia

http://tcp.sagepub.com/The Counseling Psychologist

http://tcp.sagepub.com/content/29/3/413The online version of this article can be found at:

 DOI: 10.1177/0011000001293006

2001 29: 413The Counseling PsychologistAnna Madill, Sue Widdicombe and Michael Barkham

The Potential of Conversation Analysis for Psychotherapy Research  

Published by:

http://www.sagepublications.com

On behalf of: 

  Division of Counseling Psychology of the American Psychological Association

can be found at:The Counseling PsychologistAdditional services and information for     

  http://tcp.sagepub.com/cgi/alertsEmail Alerts:

 

http://tcp.sagepub.com/subscriptionsSubscriptions:  

http://www.sagepub.com/journalsReprints.navReprints:  

http://www.sagepub.com/journalsPermissions.navPermissions:  

http://tcp.sagepub.com/content/29/3/413.refs.htmlCitations:  

at University of Leeds on August 1, 2011tcp.sagepub.comDownloaded from

THE COUNSELING PSYCHOLOGIST / May 2001Madill et al. / CONVERSATION ANALYSIS� INTERNATIONAL FORUM

The Potential of Conversation Analysisfor Psychotherapy Research

Anna MadillUniversity of Leeds

Sue WiddicombeUniversity of Edinburgh

Michael BarkhamUniversity of Leeds

An illustrative analysis is offered demonstrating the potential of conversation analysisfor psychotherapy research. Extracts are presented from an unresolved problematictheme selected from an unsuccessful eight-session psychodynamic-interpersonal ther-apy of a female client presenting with a major depressive episode. The authors identifydescriptive devices and communicative skills on which the client and therapist draw inthe process of problem formulation. They demonstrate the way in which differing ver-sions of the client’s problems were presented and how the client and therapist pursueddiffering attributional projects. Specifically, the authors examine how internal and exter-nal accounts of the client’s underlying problems were accomplished and warranted andhow the therapist attempted the transformation of problems. The quality criteria pertain-ing to conversation analysis are indicated, the implications for therapeutic practiceexplored, and the relationship between the present findings and previous researchdiscussed.

Our aim in this article is to introduce a relatively novel methodologicalapproach, conversation analysis, which we shall argue has the potential toprovide useful insights into the nature of the psychotherapy process (Gale,1991; Nofsinger, 1991). We begin by outlining conversation analysis. Wethen make several observations about the kinds of interactional issues thatneed to be examined in relation to the process of problem formulation,because this has been described as a crucial part of psychotherapy (Peyrot,

413

The first two authors are in alphabetical order. Anna Madill was supported by a MedicalResearch Council Training Award. We would like to thank David Newman and Mark Cheesmanfor their time and careful consideration of the therapy tapes, Janet Needham for her help tran-scribing sections of these tapes, Sarah Edmunds for her secretarial support, and our reviewersand editor for their tremendous help shaping this article. Finally, we are most indebted to theanonymous client and therapist for allowing us to conduct research on this psychotherapy. Corre-spondence concerning this article should be addressed to Anna Madill, School of Psychology,University of Leeds, Leeds, LS2 9JT, UK; e-mail: [email protected].

THE COUNSELING PSYCHOLOGIST, Vol. 29 No. 3, May 2001 413-434© 2001 by the Division of Counseling Psychology.

at University of Leeds on August 1, 2011tcp.sagepub.comDownloaded from

1987). Finally, we specify the relationship of conversation analysis to otherqualitative approaches currently employed in psychotherapy research.

Conversation analysis is the study of talk in interaction and of how,through talk, people accomplish actions and make sense of the world aroundthem. There are three distinguishing features of conversation analysis’ treat-ment of interaction and language. The first is its activity focus. This is mani-fested in the analytic aim of identifying how social and institutional activi-ties, such as making a diagnosis, are accomplished in a stretch of interaction.It is argued that actions are accomplished through the communicative compe-tencies and tacit skills that underpin everyday interaction and institutionaltalk. Conversation analysis aims to identify such skills in the fine-graineddetail of recordings of talk.

The second feature is that conversation analysis entails examining turn-by-turn sequences of utterances, which are characterized in terms of the actionsthey perform. Adjacent turns in conversation constitute paired actions, whichshould display alignment such that given a first action (e.g., an invitation), alimited set of next actions can be expected (an acceptance or rejection).Moreover, by producing the relevant next action, participants display theirunderstanding of what the prior turn was doing. The turn-by-turn organiza-tion of conversations therefore provides the context within which “publiclydisplayed and continuously updated intersubjective understandings is sys-tematically sustained” (Heritage, 1984, p. 259). This is also an important ana-lytic tool because these public displays are available for inspection and pro-vide the opportunity to see how joint understanding is produced. Of course,examining next turns may not provide direct access to participants’ under-standings of prior talk; this is often exhibited indirectly (so, e.g., “accepting”is a way of displaying an understanding that the prior turn was an invitation).In addition, next utterances may interpret the prior talk in a way that is rele-vant to the project of the next speaker, and this may not coincide with that ofthe first speaker (Heritage & Watson, 1979). Therefore, we should see how aspeaker may try to influence the trajectory of talk by not pursuing a course ofaction made relevant by a prior turn. Speakers may thereby transform mean-ings and oppose and modify prior actions, and in this sense, the products ofinteraction are jointly produced.

A third feature of conversation analysis is its emphasis on participants’orientation to the business at hand. The goal is not to impose an interpretationof what is going on in the talk but to see how participants themselves analyzeand classify the kind of business that a turn in dialogue is attending to. Inaddition, rather than assuming the relevance of particular roles or identities,the institutional context or, indeed, power and status relations, it is incumbentupon analysts to show that these are oriented to by participants and made con-sequential for the interaction. In fact, studies have shown that basic features

414 THE COUNSELING PSYCHOLOGIST / May 2001

at University of Leeds on August 1, 2011tcp.sagepub.comDownloaded from

of conversational interaction are used by participants in institutional settingsto manage particular role-specific activities and that these are the meansthrough which asymmetries in status are realized. For example, professionalscan strategically direct talk through the capacity to define topics as relevant orirrelevant to the matter at hand (e.g., Tannen & Wallat, 1987). However, thereare also studies that delineate the means through which laypersons mayaccomplish opposition to professionals’ activities. For example, Heritageand Sefi (1992) pointed to the unilateral ways in which health visitors initiateadvice giving to new mothers and how the latter’s opposition to this generallytakes a passive form. Furthermore, it is often through such devices thatprofessional-layperson identities are made relevant to and consequential forthe talk and therefore, in an important sense, how institutions are grounded intalk in interaction (Maynard, 1991).

Its action focus, turn-by-turn analysis, and emphasis on participants’ ori-entation make conversation analysis ideally suited to examining empiricallythe nature of the psychotherapeutic process. Here, we focus on the process ofproblem formulation. Previous work suggests that problem formulationtakes place in three distinct stages: definition of the problem, gathering evi-dence, and organization of the client’s consent to work on the problem(Davis, 1986). Problem identification is not, however, a matter of straightfor-ward and objective diagnosis. Instead, it is a product of interaction betweenclient and therapist in which the meaning and significance of the client’sexperiences are negotiated (Buttny, 1990; Hill, 1982).

This process of negotiation is subject to particular constraints related tothe therapeutic protocol. For example, psychodynamic-interpersonal therapy(as in the following illustrative analysis) is based on the assumption that theclient’s problems arise from disturbances in significant personal relation-ships, and it focuses on the client’s affect or reaction. Thus, the therapistoffers, as the basis of negotiation, hypotheses regarding the client’s feelings(Shapiro & Firth, 1985), which in turn may have the effect of transforming aclient’s account of externally located problems into internal, affective ones.For example, Davis (1986) has shown how a client’s presenting problemswith respect to her pregnancy, position as housewife and mother, and inequal-ities in her relationship with her husband were reformulated by the therapistinto an issue regarding her inability to talk about her feelings.

The implications of this process of transformation have been criticized.For example, drawing on insights from the fields of feminism, social con-struction, and poststructuralism, Hare-Mustin and Marecek (1997) sug-gested, first, that focusing on the individual as the locus of problems deflectsattention away from the role of societal conditions in generating some prob-lems and in exacerbating others. Second, because of the dictates of therapyprotocol, the problems identified may not be fully negotiated but rather a

Madill et al. / CONVERSATION ANALYSIS 415

at University of Leeds on August 1, 2011tcp.sagepub.comDownloaded from

product of the therapist’s reformulation of the client’s prior version and theclient’s subsequent confirmation or consent. Third, the therapist is guided byparticular forms of knowledge through which he or she constructs a view ofreality. In doing so, the therapist may selectively disregard diverse accounts,disconfirming evidence and alternative views of human experience. Thus, itis argued that the therapeutic relationship is inevitably characterized by dif-ferences in expertise, authority, and power. Nevertheless, because the thera-pist’s problem formulation requires the client’s approval or at least compli-ance, this means that there is always the possibility of opposition to orrejection of the therapist’s version. This has not been studied in detail,although Davis (1986) made some reference to the possibility of such resis-tance in her study.

Finally, in promoting conversation analysis, it is useful to specify its rela-tion to other forms of qualitative approach currently developed in the field.From the early 1980s, qualitative methods have enjoyed a renaissance in psy-chotherapy research. Arguably, the most developed of these are task analysis(Greenberg, 1984) and comprehensive process analysis (Elliott, 1989). Thefeature that most distinguishes conversation analysis from these is that it doesnot base research on therapist insight or on assumptions inherent in the thera-peutic protocol but on a growing literature about how conversations work.For example, task analysis requires a prima facie description of the task thatthe client must perform in the segment selected for analysis and possiblestrategies that the client and therapist may use to facilitate this work. Com-prehensive process analysis provides a structure outlining features of thetherapeutic process considered relevant for the researchers’ attention. Whatconversation analysis offers is a data-driven perspective from outside theinstitution of psychotherapy. At the same time, the approach has potential tofurther our understanding of psychotherapeutically important constructs,such as, for instance, the alliance (e.g., Bordin, 1979), and to be used in com-bination with certain preestablished frameworks, such as the significantevents paradigm (Rice & Greenberg, 1984). However, it would do so throughexamining how significant events (e.g., a rupture in the alliance) unfoldedsequentially within the interaction itself. Most important, it would demon-strate how participants themselves made certain tasks (e.g., repair of the alli-ance) relevant, or indeed irrelevant, to the matter at hand during the ongoingconversation.

In this article, we demonstrate the potential of conversation analysis forpsychotherapy research. To do so, we examine the processes of problem for-mulation and attributional accounting by a therapist and client within thecourse of psychodynamic-interpersonal psychotherapy. The case we exam-ine, which is defined as unsuccessful, was selected because such cases havebeen relatively neglected in process research but may shed light on elements

416 THE COUNSELING PSYCHOLOGIST / May 2001

at University of Leeds on August 1, 2011tcp.sagepub.comDownloaded from

necessary for successful therapy. In the Analysis section, we show how par-ticipants fail to collaborate in the production of the client’s problem, and wealso consider when and how different roles are mobilized. This example illus-trates the three key features of conversation analysis: (a) identification of theactions performed within the dialogue, for example, formulations; through(b) a turn-by-turn analysis of the unfolding dialogue; which (c) focuses onparticipants’ own understanding of conversational meaning as demonstratedthrough their talk. The implications of this analysis, and hence potential util-ity of conversation analysis for psychotherapy practice, are explored in theDiscussion.

METHOD

Participants and Case Selection

The case was a therapy of a female client who completed all sessions of atime-limited, psychodynamic-interpersonal psychotherapy consisting of eight1-hour weekly sessions. This case was selected from a British study, the Sec-ond Sheffield Psychotherapy Project (Shapiro et al., 1994), which provided117 audiotaped therapy cases of clients obtaining a diagnosis of majordepressive episode (Diagnostic and Statistical Manual of Mental Disorders,3rd Ed. [DSM-III], American Psychiatric Association, 1980). We selectedthis case because the client’s Beck Depression Inventory (BDI) scores wereparticularly unsuccessful in comparison with the other 29 clients receivingsimilar therapy. This client’s scores were as follows: 28 at initial screening,24 at intake, 22 immediately prior to first session, 23 at 2-weeks posttherapy,19 at the 3-month follow-up, and 10 at the 1-year follow-up. Although the cli-ent’s BDI score fell to almost normal levels 1 year after therapy, theyremained high during and immediately after therapy. Hence, this case wouldbe defined as unsuccessful in the short term. Moreover, data at the 12-monthfollow-up showed that the client had used community education courses,including transactional analysis, relaxation techniques, and assertiveness forwomen. Hence, improvement scores at 12-month follow-up are likely to becontaminated by the effects of these interventions. The client was not on anymedication.

The client was White, female, in her 40s, and in full-time white-collaremployment. She lived with her partner and their two young children. Theclient’s partner was separated from his wife (referred to as “ex-wife” in thefollowing analysis) and their child. However, he remained in regular contactwith them. The therapist was White, male, of similar age to the client, and had18 years of experience with psychodynamic-interpersonal therapy. The ther-

Madill et al. / CONVERSATION ANALYSIS 417

at University of Leeds on August 1, 2011tcp.sagepub.comDownloaded from

apy was conducted within a research context, and written informed consentto use audiotapes of this therapy for research purposes was obtained from theclient after therapy completion.

Selection of Extracts

Three weeks after the initial assessment, the client was asked to select par-ticular problems to work on during therapy (Personal Questionnaire) (Phil-lips, 1986). The client rated her selected problems each week according tohow much the problems were bothering her from 1 (not at all) to 7 (extremely).The problem chosen for analysis was “difficulty shouldering all the responsi-bility at home.” This theme was selected because it was specific enough toallow identification of relevant material. Moreover, because the scoresranged from 4 to 6, it seems the client was considerably bothered by this issuethroughout therapy and that it remained unresolved at therapy completion.

To obtain transcripts of conversation pertaining to this issue, three psy-chology graduates in clinical-related research listened to audiotapes of thecase for relevant passages. Each of the eight sessions had extracts chosenindependently by at least two of the graduates. They were requested to be, ifanything, overinclusive in their selection, because the aim of this process wasto identify as much thematically pertinent material as possible. All selectedextracts were transcribed. This provided 16 passages varying from a few linesto five pages of transcript. The transcription conventions adopted in this studyare standard in conversation analytic research (Atkinson & Heritage, 1984)(see Table 1).

Analytic Procedure

First, the selected extracts were read and reread carefully, and severalsequences were chosen for further analysis. The basis of this choice was theobservation that there were a number of sequences in which the client andtherapist appeared to be debating the definition and source of the client’sproblems. Second, the actions within each of these sequences and the rela-tions between actions were identified. This was helped by continually asking,“What is the speaker doing in this turn?” The more detailed analysis revealedthat, in these sequences, the therapist failed to take up the client’s problemformulation and vice versa. Then, three of these sequences were selected forfine-grained analysis because they offered good examples of the identifiedpattern (a form of theoretical sampling; see Strauss & Corbin, 1998). Third,we considered the design of utterances, including selection of descriptiveterms and characterization of states of affairs, and how they provided for cer-tain understandings of the actions being performed and the matters talked

418 THE COUNSELING PSYCHOLOGIST / May 2001

at University of Leeds on August 1, 2011tcp.sagepub.comDownloaded from

about. This involved identifying the use and functional significance in situ ofconversational devices observed in other studies, such as contrasts, three-partlists, gists, and hypothetical examples (each of which are seen in the extractsexamined below). This process was aided by asking “What are the conse-quences of putting it this way rather than that way?” “What inferences areavailable from this use of this term rather than another?” “How is the actionaccomplished through such devices?” We also noted how the turn was takenand terminated, silences, pauses, and overlapping utterances and how theycontributed to the action sequences. Finally, we examined the ways that theaccomplishment of actions also implicated certain identities, roles, or rela-tionships between participants. Further details of these methodological pro-cedures are to be found in Pomerantz and Fehr (1997) and ten Have (1998).The analysis of the three extracts selected for detailed inspection is presentedin full in the following section.

ANALYSIS

In Extracts 1 and 2, the client constructs several accounts in which sheattributes the source of her distress to her domestic situation and in particularher partner’s behavior. We will examine how she constructs and warrants hercomplaints about him prior to considering the therapist’s responses.

Madill et al. / CONVERSATION ANALYSIS 419

TABLE 1: Transcription Conventions

Simultaneous utterances Client: just to go and visit you kn[ow]Therapist: [Yes]

Overlapping utterances Client: Sc[enesTherapist: [Epi[sodesClient: [Yes]Therapist: [Sessions you called them

Contiguous utterances Therapist: or it’s not=Client: =I think I’m more frightened of

Intervals between utterances Client: things he should do to make me feel(timed in seconds) (0.5)

Therapist: So he is spending time with herIntervals within utterances Client: Um (0.5) although you know he saysExtension of sound Therapist: the colla::pse the e::mptynessHalting or abrupt cut-off Client: and - it’s almost as well as if - you know -Rise and fall in intonation Therapist: once ag�ai�n its the g�uiltEmphasis Client: I feel very nervous actually askingLoudly spoken Client: NOT for being a bad motherQuick pace Client: a meeting > this that and that < wellAudible intake of breath Therapist: about you somehow .hh having y’know

at University of Leeds on August 1, 2011tcp.sagepub.comDownloaded from

Extract 1, Session 2 [C is the client, T is the therapist]

1 C: . . .he see::ms to spe::nd an awful lot of ti:::me - th�e::re2 and �I find it very very difficult to co::pe with -3 �um (0.5) although y�ou kn�ow he says there’s4 nothing in it and a::ll the re::st of it I st�ill5 feel that there - are things that he sh�ould d�o to6 make m�e:: feel7 (0.5)8 T: So he is spe::nding t�i::::e with h�e::::r=9 C: =.Hh just to go - and v�is�it you kn[�ow]=10 T: [Y�e::s]11 C: =But to m�e::: it’s a lot of t�i::me because we have a12 very busy - l�i::fe - u:::m - and �I:: feel as if �I::’m13 l�eft - with you kn�ow collecting the children14 taking them h�o:::me - tea y�ou:: kn�o:::w >we’ve go’

a15 solid fuel cooker th�a’ needs stoking and all16 the the rest of it< - fire to light .hhh - and17 e�veryth�ing seems to be down to m�e:: and after18 w�o::rk - >h�e’s got l�ots of meetings and things19 but quite< - quite o::ften if he’s got a meeting20 he’ll go there - for an hour f�i::rst and then go21 �o::ff - and �I feel I’m just left with �e::veryth�ing22 - um tt- ss- so �I can supp�ort the way he wants to �do::::23 things um

Extract 2, Session 2

1 C: . . . um (0.5) �a::nd - it’s �almost as well as if - you2 kn�ow - his car�eer his:: what he wants to d�o:: is3 param�ount it it’s taken for gr�a::n�ted you know4 if he’s got a m�eeting >this tha’ and tha’< - well -5 got a m�ee::�ting and that’s �it - if �I have to do6 anyth�i::ng - y’ know I’ve got to sort of book7 in three weeks in adv�a::nce y’ know say is it8 alr�i::::ght (0.5) for me to - you know even a st�a::ff9 meeting - or anything like tha::’ -10 T: Yea:::h - this is - mayb�e:: something which h�its you::11 very h�a:::rd - part�i::cular�ly because of this idea12 that .hh we’ve looked at already about y�ou somehow .hh

420 THE COUNSELING PSYCHOLOGIST / May 2001

at University of Leeds on August 1, 2011tcp.sagepub.comDownloaded from

13 having y’ know N::�Othing inside the coll�A::pse the14 �e::mptiness or what[ever=15 C: [Mm hm]16 T: =that so::mehow you ca:::n’t - you can’t imp�i:::nge on him17 you you don’t really c�ou::nt for �anything=18 C: =Mm (0.5) I f�eel very n�e::rv�ous actually �a::sking for19 any[�thing -20 T: [Mmm]21 C: [- it takes me a long time to work �u::p to it -22 T: Mm23 C: - even if it’s:: s�omething y’ know fairly24 insign�i::fica[nt=25 T: [Mm]26 C: =in the li::ves of ordinary folk as it were=27 T: =Mm=28 C: =� Er - y�ou kn�o::w - if u:::m - �I kn�ow if I want to29 go in- into S�ai:::nsb�urys or something one e::ven�ing -30 there’s a::lw�ays:: u::m I always have to take the31 ch�i::ld�ren because he’s �always got something �o::n32 you k�now and [and=33 T: [Mm]34 C: =and jus’s- sa:::ying well - are you going to help me35 with the sh�o::pp�ing tonight it takes an36 aw::ful an aw::ful lot of �e::ff�ort to actually - s�a::y it37 because I kn�o::w I’m fr�ight�ened I suppose of the38 resp�o::�nse because I think I kn�o::w what it’s g�oing39 to b�e:: - um40 (5)

The basis of the client’s first complaint (Extract 1, lines 1-6) is that herpartner spends a lot of time at his ex-wife’s house, and there are things that heshould do but does not. She specifies the nature of the problem as the amountof time he spends there by characterizing it as an “awful lot,” thus implyingthat it is excessive. In addition, she addresses and discounts other reasons forher distress, which are made potentially relevant through her description. Forexample, she uses the term there to describe the place where he spends time(rather than, say, “at his ex-wife’s place”) and thereby plays down the signifi-cance of the location. She also uses the idiomatic phrase, “He says there’snothing in it and all the rest of it.” This is the kind of phrase used to character-ize a relationship as innocent or platonic in circumstances where its status assuch may be open to doubt or to discount possible ulterior motives. Thesecharacterizations therefore work to discount the attribution of the source of

Madill et al. / CONVERSATION ANALYSIS 421

at University of Leeds on August 1, 2011tcp.sagepub.comDownloaded from

the problem to her partner’s continuing relationship with his ex-wife and herown subsequent feelings or reactions.

In her second account (Extract 1, lines 11-23), she provides a moredetailed description of her circumstances, which is oriented to clarifying andjustifying her complaint against him. This is done in part by producing athree-part list of the things that she and her partner should be doing: “collect-ing the children (and) taking them home,” making tea and stoking the cooker(and “all the rest of it”), and lighting the fire. Potter (1996b), following Jeffer-son (1990), observed that three-part lists are frequently used to indicate thatthe individual instances stated stand for something more general; here, herformulation indicates the general class of domestic chores. Moreover, this listis constructed as a list of things that have to be done, and where pronouns areused, these indicate that such things are a feature of their shared lives (e.g.,“We’ve got a solid fuel cooker”). In this way, the client constructs the objec-tive fact of the household tasks and, at the same time, indicates that they aretasks for which she and her partner are mutually responsible. These detailsare significant because justifying her complaint concerning the inadequacyof her partner’s contribution depends on her first establishing the extent of theshared domestic chores that she suggests she is left with. In addition, shesays, “after work . . . if he’s got a meeting he’ll go there for an hour first andthen go off.” She thereby clarifies that he does have some time available butdoes not prioritize domestic responsibilities.

In the first part of Extract 2, the client builds a complaint about theinequality in their relationship by drawing a series of contrasts between herpartner’s unquestioned pursuit of his activities and constraints imposed onher own. First, she says that if she wants to do anything, then she has “got tosort of book in three weeks in advance”; if he’s got a meeting, “that’s it.” Sec-ond, she says that she has to ask, so he can “say is it alright”; his attendance atmeetings is taken for granted. Third, she makes relevant the general class ofhis activities through the generalizing phrase “this that and that” and thenstates that such constraints are applied to her even for a staff meeting or “any-thing like that.”

Two further descriptive strategies are used in Extracts 1 and 2. The firstconcerns the way that she also constructs her partner’s identity through herdescription of her domestic circumstances, which in turn reinforces the basisof her complaint. In Extract 1, he is described as someone who has a lot ofmeetings but who also “quite often goes there [to his ex-wife’s place] for anhour first.” She thus implies that he is someone who does not take his respon-sibilities seriously, either toward domestic duties or toward her. In addition,she says that being left with everything is “so that I can support the way hewants to do things.” Compared with, say, “the way he does things,” whichwould perhaps suggest that his actions are just thoughtless, “the way he wants

422 THE COUNSELING PSYCHOLOGIST / May 2001

at University of Leeds on August 1, 2011tcp.sagepub.comDownloaded from

to do things” implies that his actions are also selfishly motivated. Similarly,in Extract 2, she suggests that he does what he wants to do without consultingher or taking her into account: “What he wants to do is paramount,” and “ifhe’s got a meeting. . . . Well got a meeting and that’s it.” In these ways, then,she constructs his character as selfish and inconsiderate, and this functions tostrengthen her complaint that his conduct is the source of her distress.

A second strategy is seen in the way that she portrays the frequent occurrenceand hence achieves the routine nature of this state of affairs. In Extract 1, thisis accomplished through the insertion of the statement that “he’s got lots ofmeetings and things,” which implies that being left to do everything occursfrequently. In Extract 2, a similar effect is achieved in several ways, for exam-ple, by describing what he wants as taken for granted. She also produces ahypothetical case—“if he’s got a meeting.” It has been observed elsewherethat hypothetical cases allow speakers to distill recurrent features from anumber of events and bring them together in a form that indicates their gener-ality or their routineness. Moreover, with such examples, there is no recourseto a direct interrogation of the details of the events (Widdicombe & Wooffitt,1995). Finally, she makes relevant meetings and “this that and that” and thusimplies the generality of such incidents. This is significant because the justi-fiable nature of her complaints depends on the state of affairs being regularrather than exceptional. Furthermore, it has also been observed that formulat-ing behavior as routine is used to imply that it is related to the character of aperson (Edwards, 1997). Here, such formulations serve to reinforce the por-trayal of her partner’s character as selfish and egocentric.

To summarize, the client employs several descriptive devices to constructand warrant her complaint about her partner and to attribute the source of herdistress to his conduct. These include the descriptive terms used, listing, con-trast structures, as well as the way she portrays his character and the routinenature of the circumstances.

In both extracts, however, the therapist offers a different formulation ofher problems. In particular, he makes relevant underlying feelings as thesource of her distress and employs several devices to bring about this shift inthe focus of the interaction. For example, in Extract 1 the therapist produces agist of what the client has said (line 8). Gists formulate the sense achieved upto a point in the conversation. Simultaneously, they preserve certain featuresof the preceding talk while others may be glossed or recast (Heritage & Wat-son, 1979). Here, the therapist produces the gist “so he is spending time withher.” This preserves certain features of the preceding account (spendingtime) but it overlooks the location (there) and the amount of time. Instead, itmakes relevant who he spends time with. Thus, his gist ignores the work doneby the client to counter such an inference (e.g., through the use of the idiom-atic statement “nothing in it”), and the problem is now recast in terms of her

Madill et al. / CONVERSATION ANALYSIS 423

at University of Leeds on August 1, 2011tcp.sagepub.comDownloaded from

partner being with his ex-wife. This in turn suggests that the underlying prob-lem is perhaps the client’s jealousy or insecurity. In this way, he attempts toinitiate a topic change by inviting discussion of their relationship or her feel-ings about it rather than, say, seeking clarification of what her partner shoulddo and how he should make her feel.

In Extract 2, the therapist produces an assessment of her account, “yeahthis is maybe something which hits you very hard” (lines 10-11), therebyindirectly acknowledging that she has made a complaint. However, he makesno reference to the complainable aspects of her partner’s behavior nor to herrelationship with him. Instead, the therapist’s version makes her affectivereaction the significant feature of the difficulties she has described. He thusmanages a topic shift to her feelings, and he is then able to offer two accountsof the source of her distress. He says first that she has “nothing inside” andsecond that she “can’t impinge on her partner” because she doesn’t “reallycount for anything.” He thus invokes her sense of worthlessness as the under-lying basis for her problems. Moreover, like the client, he constructs therecurrent nature of the problem by describing it as “this idea that we’velooked at already.” He thereby implies that her complaint is a further mani-festation of an internal, psychological problem rather than one related to hercircumstances.

Thus, in both extracts we have seen how, through his use of gists andassessments, the therapist attempts to bring about topic shifts to focus on theclient’s feelings. Nevertheless, the successful accomplishment of a topicchange or reformulation requires uptake and ratification by the client. In theextracts above, the client does not comply with this requirement. For exam-ple, in Extract 1, the client first orients to and rejects the romantic implica-tions of the therapist’s gist, “spending time with her,” by saying “just to goand visit” (line 9). It is moreover noteworthy that this rejection is immediate;there is no noticeable gap between the therapist’s and the client’s utterances.This is followed by a further account of her troubles (lines 11-23), which, incontrast to the first version, is rich in descriptive detail. That is, her first for-mulation (lines 1-6) is minimally produced in the sense that it is marked byvagueness (e.g., spending time there is characterized as something “heseems” to do; the things that he should do are not specified). Her second for-mulation, by contrast, is comparatively rich. Therefore, it can be suggestedthat the character and detail of her second version are oriented to reassertingor reinforcing her prior problem formulation.

Support for this conclusion is derived from her use of similar strategies inExtract 2. Specifically, the client rejects the therapist’s suggestion that herproblems stem from a feeling that she “can’t impinge on him.” So that,although she complies in part with the topical redirection by referring to herfeelings, “I feel very nervous actually asking for anything” (lines 18-19), at

424 THE COUNSELING PSYCHOLOGIST / May 2001

at University of Leeds on August 1, 2011tcp.sagepub.comDownloaded from

the same time she transforms the underlying reason for this state of affairs. Byattributing it to her partner’s likely reaction, she reasserts her claim that he isthe source of her problems. As in Extract 1, her rejection of the therapist-identified problem in favor of her prior claims is worked up through the pro-vision of a detailed account of her circumstances and by reasserting the rea-sonableness of her complaint. This is done by setting up a contrast betweenthe ordinary and trivial nature of her requests and the degree of nervousnessand fear she experiences in relation to them. That is, she says that she feels“very nervous actually asking for anything,” that “it takes me a long time towork up to it,” and it “takes an awful lot of effort” to ask for help. She thendescribes the kind of requests that are associated with such nervousness:“Even if it’s something fairly insignificant in the lives of ordinary folk.” Inaddition, she provides an example of the kind of circumstances to which sheis referring: going shopping and having to take the children with her.

The mismatch between her requests and her fear is then attributed to hislikely response (lines 37-39). The nature of his response is not stated, but itcan be inferred from the earlier contrast between her requests and fear that itwill be extreme, negative, not warranted, and predictable (“I think I knowwhat it’s going to be”). In this way, she suggests that her unwillingness to askis not a result of feeling that she cannot impinge on him, but his likely reactionif she did so. Moreover, through making relevant his unreasonable responseto her mundane requests, she adds to the character building accomplished inthe first part of this extract. She thereby reinforces her claim that her partner isthe cause of her problems.

In summary, the client does not take up the therapist’s invitation to focuson her underlying feelings, and the therapist does not engage with the client’saccount of her domestic situation and partner’s behavior. This is accom-plished through the use of several devices, especially those related to manag-ing and rejecting topic shifts. Moreover, these devices serve the useful func-tion of enabling the participants to maintain the appearance of collaborativeaccounting and thus to avoid overt conflict in the interaction while pursuingtheir own rather different projects.

There is, however, a further set of observations that we want to make aboutthese extracts that concern the way that the participants’ descriptive strate-gies are simultaneously oriented to and mobilize particular institutional iden-tities. In Extract 1, we noted that the client’s first account is characterized byvagueness; moreover, her final statement “to make me feel” (lines 5-6) is notcompleted. It is possible that these vague formulations are oriented to theasymmetries in their roles and invite the therapist, as expert, to seek relevantclarification and to take the lead in specifying the problem. Indeed, someindication that this is the case is given by the way that he takes up this invita-tion by producing a gist of what she has said.

Madill et al. / CONVERSATION ANALYSIS 425

at University of Leeds on August 1, 2011tcp.sagepub.comDownloaded from

In Extract 2, the therapist employs similarly vague formulations in hisstatements about the source of her distress. For example, instead of specify-ing the nature of the underlying feeling, he orients to a general class of feel-ings through producing a three-part list. That is, he refers to having “nothinginside,” “the collapse, the emptiness, or whatever” (lines 13-14). He also usesqualifying terms “somehow you can’t impinge on him” and “you don’t reallycount for anything” (lines 16-17). The vagueness of his characterization ofher psychological state can be seen to invite her to confirm or upgrade hisinterpretation. These formulations therefore orient to and mobilize his role ascollaborator in identifying the problem. Moreover, by referring to “this ideathat we’ve looked at already” (lines 11-12), he invokes their identities as jointcollaborators in negotiating the problem and implies that some kind of com-mon agreement has been reached previously.

In the following extract, which occurred later in the course of the eighttherapy sessions, we observe the use of similar and further strategies of prob-lem formulation and show how, through them, the therapist orients to a differ-ent identity, as therapist rather than collaborator.

Extract 3, Session 6

1 T: I’m I’m loo::king at - what you were s�a:::�ying2 about feeling al�o:::ne with the problem -3 C: Mm4 T: - which is p�a::rtly about not being able to sha::re5 it - as you’ve decided to do h�e::re >but it’s also6 about feeling that< .hh you w�ou::ldn’t get the support7 if you d�i:::d (0.3) or it’s n�o:t=8 C: =I think I’m m�o::’ frightened of being cond�e::mn�ed=9 T: =Yeah -10 C: - Um (1.5) NOT fo:::’ - being a bad mother >in inverted

commas< bu’ for re�a:::cting toostr�o::ngly to [it=

12 T: [Y�e:::s::]13 C: =[and=14 T: [Mm �hm]15 C: =cr�ea::ting the[se=16 T: [Mm �hm]17 C: =um=18 T: =Sc�e:::nes=19 C: =Sce:[nes=20 T: [�Epi[s�odes=

426 THE COUNSELING PSYCHOLOGIST / May 2001

at University of Leeds on August 1, 2011tcp.sagepub.comDownloaded from

21 C: [=Yes=]22 T: [=S�e:::ssions you called

[them=23 C: [=Yes]24 T: =Yes (0.5) yes (0.5) yes so this is the - this is o::nce25 ag�ai::: �n it’s the - it’s the g�ui::::lt and the sense of26 being open to accus�a::: �tion - which I sti:ll feel27 is arou:::sed for you:: when you rush to your28 m�o::ther’s de�fence because you see:: - your29 perception of me::: as blaming your m�o::: �ther is30 rea::lly s- stemming from a sense that y�ou:: were31 blameworthy - as a mo::�ther32 (5)

As before, the therapist makes the client’s reaction “feeling alone with theproblem” the significant feature of her complaint. Whereas earlier, he used agist (Extract 1) and an assessment (Extract 2) to accomplish this, here heintroduces this topic by referring to the client’s own report, “what you weresaying about” (lines 1-2). As in Extract 2, this feeling is in turn attributed totwo sources: “not being able to share the problem” and “feeling that youwouldn’t get the support if you did.” These characterizations locate the prob-lem in her reticence rather than in her circumstances or her partner’s behav-ior. Specifically, he refers to “not being able to share it as you’ve decided todo here.” He thus suggests that sharing the problem is incumbent on a deci-sion to do so and thereby makes available the inference that she has decidednot to talk to her partner. Similarly, he describes her feeling that she would notget the support “if she did.” The implication is that she has not tried to discussthe problem with him.

Nevertheless, the client does not ratify the therapist’s version, and we canobserve the way she does this through her use of strategies similar to thoseidentified earlier. For example, she rejects the therapist’s account through therevision and thereby rejects his version. It is noteworthy that her rejection isproduced as an interruption. She starts speaking before the therapist com-pletes his turn on line 7, and there is no gap indicating that the therapist isinviting her to take the floor. She attributes her feelings to her fear of her part-ner’s response: his condemnation of her “for reacting too strongly.” Her useof the term condemned portrays an extreme response, which, in the context ofher (reasonable) concerns, implies that his reaction is unreasonable. How-ever, she explicitly discounts another possible reason for his condemnation:“not for being a bad mother.” She thus makes clear that it is not her abilities asa mother that are the focus of his rebuke (although this is the focus of the ther-apist’s subsequent attributional work).

Madill et al. / CONVERSATION ANALYSIS 427

at University of Leeds on August 1, 2011tcp.sagepub.comDownloaded from

Whereas the client identifies the problem as being her partner’s criticismof her and by implication his lack of understanding, the therapist subse-quently reformulates that as a feeling of “being open to accusation” and “theguilt” that arises from a feeling of blameworthiness as a mother. Thus, beingcondemned is transformed into a feature of the client’s own feelings and anunderlying personal problem. Moreover, he constructs the routine nature ofthe feeling of guilt through the use of the definite article (“the guilt”) and thephrase “once again the guilt.” His version is further warranted in severalways. First, he implies that his formulation is derived from his continuingobservations (“I still feel is aroused for you when you rush to your mother’sdefense”). Second, he produces evidence for his attribution by characterizingher actions in therapy as defending her mother, which he says is related to her“perception of me as blaming your mother.” This perception is then attributedto her “sense that you were blameworthy as a mother” (lines 30-31), and thisis described as what her guilt is “really stemming” from. Moreover, by char-acterizing the client’s perspective as a perception, he indicates that it is sub-jective and, by implication, perhaps mistaken rather than real. The therapist’scharacterization of the client’s behavior in therapy and attribution to the realsource therefore functions to reinforce the status of the client’s guilt as theunderlying and pervasive problem.

In both versions, the therapist uses authoritative accounting in doing thisattributional work. In this way, he simultaneously orients to his institutionalidentity and role as therapist rather than collaborator in coproducing the cli-ent’s problems. In line 1, for example, he says “I’m looking at what you weresaying” (cf. his earlier use of the pronoun “we”). Moreover, whereas inExtract 1 his statements displayed some vagueness and as such invited theclient’s confirmation or elaboration, here his statements are produced as fac-tual descriptions of the underlying problem. That is, he states that “feelingalone with the problem” is “partly about . . . but it’s also about.” In addition,there is a more explicit reference to the therapeutic encounter in his statementthat she has decided to share the problem “here.” Finally, in lines 24 through31, the therapist uses authoritative formulations such as “it’s the guilt and thesense of being open to accusation” as well as an unqualified characterizationof her action as “rushing to her mother’s defense” and what this is “reallystemming from.”

Subsequent to this second formulation, there is a 5-second silence, whichindicates the client’s refusal to take her turn. This seems to represent a furtherstrategy of nonuptake, specifically by withdrawing cooperation in the inter-action. This is also displayed in her refusal to accept the therapist’s invitationto take the floor in the exchange that occurs at lines 15 through 24, in whichshe passively confers with the therapist’s suggestions that she creates“scenes.” Her withdrawal of cooperation in the interaction means that he is

428 THE COUNSELING PSYCHOLOGIST / May 2001

at University of Leeds on August 1, 2011tcp.sagepub.comDownloaded from

perhaps forced to adopt an institutional role as therapist and expert to main-tain some form of interaction.

DISCUSSION

The advantage of the kind of analytic approach adopted is that we canappreciate, at the level of the interactional detail through which it is done,how the participants’ projects do not cohere and therefore why no resolutionof the client’s problem is reached, even though the interaction is sustained.The client’s project is one of blaming her partner, while the therapist tries toengage her in talking about her feelings and the mutual construction of theproblem around this issue. However, the client resisted talk of her feelings,and the therapist resisted attending to the details of her complaint. In examin-ing the details of how this is accomplished interactionally, we have identifiedseveral strategies of resistance. The term resistance is used here at aninteractional level as a way of describing the effects of a set of actions in theextracts we have analyzed. These strategies include reasserting or revisingthe prior case, nonuptake of aspects of a prior turn, managing and rejectingtopic shifts, and the withdrawal of cooperation. Moreover, the failure to nego-tiate the problem was compounded by the way that the therapist finally appearedto deal, interactionally, with the failure to construct a mutual account bymobilizing an identity as expert. As in previous studies, we saw how the clientcomplied with this role by withholding a response (cf. Heritage & Sefi,1992).

Clearly, this notion of resistance and the approach to issues of power rela-tions and the construction of knowledge in formulations of client problemsdiffer from the ways that similar issues have been addressed in the feministand poststructuralist literature. Conversation analysis is not primarily ori-ented to critique or to producing inferences about psychotherapy. It is,instead, a powerful tool for describing and understanding in detail howaspects of our social lives get done by working with actual recorded interac-tions. By contrast, feminist and other interpretations of psychotherapy aimspecifically to mount a critique of the process and its implications and do solargely by drawing on a body of theoretical and conceptual tools rather thanfine-grained empirical analyses (see, e.g., Hare-Mustin & Marecek, 1997,outlined in our Introduction). Nevertheless, it is likely that such critiques willbenefit from conversation analysis’ observations and identification of strate-gies through which psychotherapy gets done.

Conversation analysis constitutes a rigorous method of analysis becauseof its requirement to base conclusions in the details of what is actually said.However, further quality criteria pertaining to this form of analysis have been

Madill et al. / CONVERSATION ANALYSIS 429

at University of Leeds on August 1, 2011tcp.sagepub.comDownloaded from

developed because concepts such as reliability and validity, as they aredefined in relation to quantitative research within a naive or scientific realistepistemology, are not directly transferable to conversation analysis, which isethnomethodological in origin (see Madill, Jordan, & Shirley, 2000). In par-ticular, Potter (1996a) identified four criteria by which to judge the quality ofdiscourse or conversation analytic research, although he noted that not allwill be evident in any one analysis. These are readers’ evaluation, partici-pants’ understanding, deviant case analysis, and coherence.

First, readers’ evaluation concerns the way that the extracts and the pro-cess of analysis are laid out and available for inspection and not just the con-clusions. Second, there is the extent to which the researcher fulfills therequirement to attend to how the participants themselves appear to be makingsense of the matter at hand. This understanding is available through thesequential unfolding of conversation and the paired actions therein embod-ied. These criteria are fully used in the present study.

The third and fourth criteria, deviant case analysis and coherence, point toways in which the present work can be developed. Conversation analysisoften commences with the identification of interactional patterns and thenproceeds to find counterexamples. The reasoning here is that if the pattern isgenerally valid, participants should orient to the anomaly as unusual. Forexample, past research suggests that agreements are by far the preferredresponse within conversations, whereas disagreements are marked by severalfeatures such as delay in responding, indirectness, and the provision of anexplanation (Pomerantz, 1984). It may be, then, that the interactional patternfound in this study is itself the deviant case and that the extended justifica-tions the client provides demonstrate her orientation to the way that disagree-ment is an accountable matter for which she has to provide good reasons. Ouranalysis suggests that focusing on strategies of uptake and nonuptake ofproblem formulations is a potentially fruitful focus for further analyses,which may consolidate and extend our preliminary findings.

Finally, coherence refers to the ability of an analysis to inform practice,build on past research, and contribute to the development of new research.Our three extracts were selected as good examples of a pattern (problemformulation/disconfirmation), and it is likely this process was detrimental totherapy. However, we must be tentative in considering this the only or maincause of failure. Furthermore, we do not know if the client’s strategies aretypical of unsuccessful therapy, although as we noted previously, the use ofminimal response as passive opposition has been documented elsewhere.Hence, at this point it is appropriate to explore the implications of this studyfor therapeutic practice and to discuss some of the past research on which the

430 THE COUNSELING PSYCHOLOGIST / May 2001

at University of Leeds on August 1, 2011tcp.sagepub.comDownloaded from

present findings build in establishing the coherence of this study with a cumu-lative body of work (e.g., Friedlander & Heatherington, 1998; Reandeau &Wampold, 1991; Tracey & Ray, 1984).

The principle of many psychotherapeutic approaches is that progressrequires the location of a core dynamic within the client and that one task oftherapy is the identification of problems that belong or are internal to the cli-ent because these are considered to be directly amenable to intervention.However, our analysis suggests that when the client refuses to conform toprotocol and comply with its demand to focus on his or her feelings, then noresolution is reached for that particular problem domain. Two points followfrom this: Successful therapy may rest on the client complying with the thera-pist’s topic shifts or else on the therapist abandoning protocol in the face ofthe client’s continued opposition. This, in turn, requires a sensitivity to thestrategies through which the client and therapist may try to resist the other’sprojects. Conversation analysis provides a way of identifying such strategiesand therefore of sensitizing the therapist to their use and role in therapy. Thesecond, broader point is that the analysis raises questions about the nature,functions, and legitimization of such alternative client-identified problemaccounts.

From a clinical perspective, a client’s noncompliance with attempts toidentify an internal dynamic may be understood as a defensive externaliza-tion (Freud, 1946). For example, the therapist has a number of psychologicalmechanisms (projection, denial, dysfunctional cognitions, etc.) to draw on asrhetorical devices to account for such disjunctures, and these may function tolegitimate the therapeutic perspective by undermining the client’s preferredaccount (Schafer, 1980). However, our analysis suggests that overlooking thenature of the client’s externalized complaints or treating them as defensivemay not be fruitful insofar as no mutually constructed problem was achieved,and therefore, there was no basis established for resolving the client’s dis-tress. There is then some indication that conversation analytic research maybe used to inform therapeutic practice.

Our study is also compatible with a growing literature on the therapeuticprocess. In attending to the turn-by-turn development of verbal communica-tion, conversation analysis has something in common with sequential analy-sis. This is a statistical technique used in psychotherapy research for identify-ing patterns in the flow of categorized client and therapist statements. Forexample, Tracey and Ray (1984) demonstrated that successful therapy wascharacterized by a pattern of complementary interaction defined as mutualcooperation with topic initiation in which there was a high-low-high patternof complementarity across therapy. On the other hand, unsuccessful therapy

Madill et al. / CONVERSATION ANALYSIS 431

at University of Leeds on August 1, 2011tcp.sagepub.comDownloaded from

showed a consistent level of low or high complementarity over time. Accord-ingly, the unsuccessful case presented here suggests low complementarity inwhich each participant is unwilling to follow the topic initiated by the other.

Sequential analysis has also demonstrated that high-alliance cases arecharacterized by therapist high-power messages followed by client low-power/high-involvement messages (Reandeau & Wampold, 1991). In contrast, asseen in the present study, low alliance cases contained elements of clientavoidance, such as refusal to take his or her turn in the conversation. Whatconversation analysis adds to sequential analysis is a detailed, functionalexamination of interaction exploring, for example, how clients resist therapist-initiated topic shifts, including avoidance strategies, and the implications thiscan have for, say, control over problem definition. Moreover, rather than usepredefined categories, conversation analysis attends to the ways in whichparticipants themselves orient to the matter at hand.

Validating the importance of problem formulation and the potential forcompeting formulations, a number of scales have been developed to tap rele-vant constructs within therapy dialogue, for example, the Construction ofProblems Scale (Heatherington et al., 1998) and Cognitive ConstructionsCoding System (CCCS) (Friedlander & Heatherington, 1998). The latter dis-tinguishes a number of dimensions, including internal-external problem con-struction, found to be of particular relevance in the present study. Moreover,the CCCS has been used to trace therapist-facilitated problem transforma-tion, which substantiates the potential ubiquitousness of this process withinpsychotherapy (Coulehan, Friedlander, & Heatherington, 1998). Other researchhas investigated the related topic of differences in the etiological attributionsof clients and their therapists. For example, Worthington and Atkinson(1996) showed that clients rated their therapist as more credible when theirattributions regarding the causal antecedents of the problem were similar tothe client’s own. So when, as in the current study, the therapist’s etiologicalattributions differ from those of the client, the therapist may lose the client’sconfidence, and the therapy is likely to suffer as a consequence.

In this article, we have introduced conversation analysis and argued that ithas potential to further understanding of the psychotherapy process. It pro-vides a perspective from outside the institution of psychotherapy yet is com-patible with the new paradigm of qualitative research that is being developedwithin the field and a growing body of literature on how therapy gets done.We identified descriptive devices and communicative skills on which the cli-ent and therapist drew in the process of problem formulation, how the thera-pist attempted topic shifts and mobilized collaborator and expert roles inmanaging his reformulations, and how this related to the client’s strategies ofresistance. Strategies of uptake and nonuptake of problem formulationswould be a useful focus for further analyses. This would offer practitioners

432 THE COUNSELING PSYCHOLOGIST / May 2001

at University of Leeds on August 1, 2011tcp.sagepub.comDownloaded from

more information about the ways in which client and therapist may try toresist the other’s projects and provide an opportunity for the therapist to steerthe conversation in a more productive direction.

REFERENCES

American Psychiatric Association. (1980). Diagnostic and statistical manual of mental disor-ders (3rd ed.). Washington, DC: Author.

Atkinson, J. M., & Heritage, J. (Eds.). (1984). Structures of social action: Studies in conversa-tion analysis. Cambridge, UK: Cambridge University Press.

Bordin, E. S. (1979). The generalizability of the psychoanalytic concept of the working alliance.Psychotherapy: Theory, research and practice, 16, 252-260.

Buttny, R. (1990). Blame-account sequences in therapy: The negotiation of relational meanings.Semiotica, 78, 219-247.

Coulehan, R., Friedlander, M. L., & Heatherington, L. (1998). Transforming narratives: Achange event in constructivist family therapy. Family Process, 37, 17-33.

Davis, K. (1986). The process of problem (re)formulation in psychotherapy. Sociology of Healthand Illness, 8, 44-74.

Edwards, D. (1997). Discourse and cognition. London: Sage Ltd.Elliott, R. (1989). Comprehensive process analysis: Understanding the change process in signif-

icant events. In M. J. Packer & R. B. Addison (Eds.), Entering the circle: Hermeneutic inves-tigations in psychology. Albany, NY: SUNY Press.

Freud, A. (1946). The ego and the mechanisms of defense. London: Hogarth Press.Friedlander, M. L., & Heatherington, L. (1998). Assessing clients’ constructions of their prob-

lems in family therapy discourse. Journal of Marriage and Family Counseling, 24, 289-303.Gale, J. E. (1991). Conversation analysis of therapeutic discourse: The pursuit of a therapeutic

agenda. Norwood, NJ: Ablex.Greenberg, L. S. (1984). Task analysis: The general approach. In L. N. Rice & L. S. Greenberg

(Eds.), Patterns of change: Intensive analysis of psychotherapy process (pp. 124-148). NewYork: Guilford.

Hare-Mustin, R. T., & Marecek, J. (1997). Abnormal and clinical psychology: The politics ofmadness. In D. Fox & I. Prilleltensky (Eds.), Critical psychology: An introduction (pp. 104-120). London: Sage Ltd.

Heatherington, L., Johnson, B., Burke, L. E., Friedlander, M. L., Buchanan, R. M., & Shaw, D. M.(1998). Assessing individual family members’ constructions of family problems. FamilyProcess, 37, 167-184.

Heritage, J. C. (1984). Garfinkel and ethnomethodology. Cambridge, UK: Polity Press.Heritage, J. C., & Sefi, S. (1992). Dilemmas of advice: Aspects of the delivery and reception of

advice in interaction between health visitors and first-time mothers. In P. Drew & J. Heritage(Eds.), Talk at work: Interaction in institutional settings (pp. 359-417). Cambridge, UK:Cambridge University Press.

Heritage, J. C., & Watson, D. R. (1979). Formulations as conversational objects. In G. Psathas(Ed.), Everyday language: Studies in ethnomethodology (pp. 123-162). New York: Irvington.

Hill, C. E. (1982). Counseling process research: Philosophical and methodological dilemmas.The Counseling Psychologist, 10, 7-19.

Jefferson, G. (1990). List construction as a task and resource. In G. Psathas (Ed.), Interactioncompetence. Lanham, MD: University Press of America.

Madill et al. / CONVERSATION ANALYSIS 433

at University of Leeds on August 1, 2011tcp.sagepub.comDownloaded from

Madill, A., Jordan, A., & Shirley, C. (2000). Objectivity and reliability in qualitative analysis:Realist, contextualist, and radical constructionist epistemologies. British Journal of Psy-chology, 91, 1-20.

Maynard, D. (1991). On the interactional and institutional bases of asymmetry in clinical dis-course. American Journal of Sociology, 92, 448-495.

Nofsinger, R. E. (1991). Everyday conversation. London: Sage Ltd.Peyrot, M. (1987). Circumspection in psychotherapy: Structures and strategies of counselor-

client interaction. Semiotica, 65, 249-268.Phillips, J.P.N. (1986). Shapiro Personal Questionnaire and Generalized Personal Questionnaire

techniques: A repeated measures individualized outcome measurement. In L. S. Greenberg &W. M. Pinsof (Eds.), The psychotherapeutic process: A research handbook. New York:Guilford.

Pomerantz, A. (1984). Agreeing and disagreeing with assessments: Some features of preferredand dispreferred turn shapes. In J. M. Atkinson & J. Heritage (Eds.), Structures of socialaction: Studies in conversation analysis. Cambridge, UK: Cambridge University Press.

Pomerantz, A., & Fehr, B. J. (1997). Conversation analysis: An approach to the study of socialaction as sense making practices. In T. A. van Dijk (Ed.), Discourse as social interaction.London: Sage Ltd.

Potter, J. (1996a). Discourse analysis and constructionist approaches: Theoretical background.In J.T.E. Richardson (Ed.), Handbook of qualitative research methods for psychology andthe social sciences. Leicester, UK: BPS Books.

Potter, J. (1996b). Representing reality: Discourse, rhetoric and social constructionism. Lon-don: Sage Ltd.

Reandeau, S. G., & Wampold, B. E. (1991). Relationship of power and involvement to workingalliance: A multiple-case sequential analysis of brief therapy. Journal of Counseling Psy-chology, 38, 107-114.

Rice, L. N., & Greenberg, L. S. (Eds.) (1984). Patterns of change: Intensive analysis of psycho-therapy process. New York: Guilford.

Schafer, R. (1980). Narration in the psychoanalytic dialogue. Critical Inquiry, 7, 29-53.Shapiro, D. A., Barkham, M., Rees, A., Hardy, G. E., Reynolds, S., & Startup, M. (1994). Effects

of treatment duration and severity of depression on the effectiveness of cognitive-behavioraland psychodynamic-interpersonal psychotherapy. Journal of Consulting and Clinical Psy-chology, 62, 522-534.

Shapiro, D. A., & Firth, J. (1985). Exploratory therapy manual for the Sheffield psychotherapyproject (Psychological Therapies Research Centre Memo). Leeds, UK: University of Leeds,School of Psychology.

Strauss, A., & Corbin, J. (1998). Basics of qualitative research (2nd ed.). London: Sage Ltd.Tannen, D., & Wallat, C. (1987). Interactive frames and knowledge schemas in interaction:

Examples from a medical examination/interview. Social Psychology Quarterly, 50, 205-216.ten Have, P. (1998). Doing conversation analysis: A practical guide. London: Sage Ltd.Tracey, T. J., & Ray, P. B. (1984). Stages of successful time-limited counseling: An interactional

examination. Journal of Counseling Psychology, 31, 13-27.Widdicombe, S., & Wooffitt, R. (1995). The language of youth subcultures: Social identity in

action. Hemel Hempstead, UK: Harvester Wheatsheaf.Worthington, R. L., & Atkinson, D. R. (1996). Effects of perceived etiology attribution similar-

ity on client ratings of counselor credibility. Journal of Counseling Psychology, 43, 423-429.

434 THE COUNSELING PSYCHOLOGIST / May 2001

at University of Leeds on August 1, 2011tcp.sagepub.comDownloaded from