Keywords

Joint Decision-Making in Cognitive Behavioral Therapy

The theory underlying Cognitive Behavioral Therapy (CBT) places great importance on joint decision-making between therapist and client (Blackburn & Davidson, 1990; Wright, Basco, & Thase, 2006). Unlike other therapies, the therapeutic relationship in CBT is guided by a specific working alliance referred to as “collaborative empiricism” (Wright et al., 2006) that involves therapists and clients working together to gather data that disconfirm core depressive beliefs or thoughts (Beck, Rush, Shaw, & Emery, 1979). Therapists are also encouraged to engage clients in a highly collaborative process in which there is shared responsibility for aspects such as setting goals and agendas, giving and receiving feedback, and putting CBT methods into action, both inside and outside the therapy session (Beck et al., 1979; Neenan & Dryden, 2000; Wright et al., 2006). CBT theory thus encourages therapists and clients to work together to question clients’ cognitive distortions and unproductive behavioral patterns, with the aim of revealing opportunities for increased rationality, reduced symptoms of depression, and improved personal effectiveness.

One aspect of CBT in which the working alliance of “collaborative empiricism” is considered particularly important is the technique of “Behavioral Activation.” According to CBT theory, Behavioral Activation involves engaging clients in a process of change that is designed to stimulate positive thought and hope, or help them solve a problem (Blackburn & Davidson, 1990). Therapists aim to assist clients in choosing one or two actions that make a positive difference to how they feel, and develop a plan to carry out these actions. Therapists are encouraged to engage clients through the use of Socratic questioning, and while they can help guide clients toward actions that might be helpful, whenever possible therapists are expected to allow clients to make the choice (Wright et al., 2006). In particular, CBT theory suggests asking a series of inductive or open questions—in a form that does not provide proposals/suggestions that the client can accept/reject, but which steer the client toward making the proposal for future action themselves. For example, a therapist might ask: “What action could you take in the next couple of days that would begin to make a difference?” A key reason for encouraging clients to make their own choices about behavioral changes within CBT is to avoid client resistance. Resistance to change on the part of clients is considered a major limiting factor to the success of CBT treatment for depression (Leahy, 2001).

Little is currently known, however, about how therapists and clients come to make their decisions about Behavioral Activation tasks within interaction during CBT sessions. This chapter explores how these decisions are made across the sequence of interaction, with a particular focus on who (therapist or client) makes the initial proposal for a behavioral change, and how this proposal for future action is responded to by the recipient.

Joint Decision-Making in Social Interaction: The Negotiation of Epistemic and Deontic Status

For a joint decision to occur in any social interaction, there must be a proposal from one recipient and an approval and commitment to the proposed future action by another participant (Stevanovic, 2012). Each participant thus has some level of involvement in the decision-making process. In order to be involved in decision-making, each person needs to have sufficient knowledge, as well as rights to make decisions (Stevanovic & Peräkylä, 2012). The extent to which a joint decision is made is thus underpinned by each person’s epistemic rights (i.e., rights to knowledge) and deontic rights (i.e., rights to determine future action) over the situation (Stevanovic & Peräkylä, 2012). Within CBT sessions, therapists carry the authority of a professional perspective (the epistemics of expertise), however they only have secondary access to knowledge about a client’s life and situation (the epistemics of experience) (Heritage, 2013b). Clients will always have ultimate epistemic access to knowledge of how situations have played out in their lives, and they also have the ultimate right to decide on their behavior in the future.

A growing body of conversation-analytic studies has examined the ways in which participants in interaction attend to and manage their rights and responsibilities in relation to knowledge and information in various settings (e.g., Butler, Potter, Danby, Emmison, & Hepburn, 2010; Heritage, 2010, 2012; Heritage & Raymond, 2005; Land & Kitzinger, 2007; Lerner & Kitzinger, 2007; Raymond, 2010; Raymond & Heritage, 2006; Stivers, 2005). Rights and responsibilities concerning what participants know, and have rights to describe, are explicitly oriented to by participants in conversation (Heritage & Raymond, 2005). The majority of work on the management of epistemic status has been conducted in two local environments of interaction: where participants are engaging in assessments (e.g., Clift, 2006; Heritage & Raymond, 2005; Lerner & Kitzinger, 2007; Raymond & Heritage, 2006; Stivers, 2005) and in question–answer sequences (e.g., Heritage, 2010, 2012; Heritage & Raymond, 2010; Raymond, 2010).

Other recent conversation-analytic research has examined how participants in conversation also attend to their deontic rights and obligations (e.g., Kent, 2012; Landmark, Gulbrandsen, & Svennevig, 2015; Lindström & Weatherall, 2015; Stevanovic, 2013; Stevanovic & Peräkylä, 2012, 2014; Zinken & Ogiermann, 2011). Deontic status is concerned with participants’ rights to determine future courses of action, that is, to determine “how the world ought to be” (Stevanovic & Peräkylä, 2012). As with epistemic status, people have varying deontic rights in different contexts, and this status is typically managed by parties in particular ways. For example, announcing a decision suggests the speaker has higher deontic status than the recipient, whereas proposing a decision claims more equal distribution of deontic status. Recipients’ responses may accept or resist the previous speakers’ claims about deontic status. Establishing a joint decision involves an interplay of the epistemic and deontic status of the participants. In order to have the right to decide on a future course of action, the participants must also have sufficient knowledge of the context that informs that course of action.

Within medical primary care consultations, researchers in conversation analysis have explored how clinicians and patients manage their epistemic and deontic rights to make decisions about treatments. The ways in which clinicians present treatment options have been described in terms of positioning across a deontic gradient. Some forms of treatment presentation set up a steep deontic gradient in that they involve expressions that limit or constrain patients’ deontic rights (e.g., providing a recommendation for a particular treatment). Other forms set up a shallower deontic gradient in that they mark the decision as the patients’ to be made (e.g., listing options for possible treatments). As Toerien, Shaw, and Reuber (2013) demonstrate, option-listing, in comparison to recommending, creates a more open space for the type of response that patients can provide, thus allowing for more patient involvement. However, research has also shown that clinicians sometimes deliver options in ways that are biased toward a particular decision, or that exclude potential options (Landmark et al., 2015; Toerien et al., 2013). This body of research on deontic rights in treatment decision-making indicates that even when professionals use communication practices that appear to involve patients in the decision-making process (e.g., recommending a treatment, rather than making a stronger claim), the extent to which patients are involved can vary and is influenced by many interactional factors. Patients have been observed to resist their deontic right to make a decision by claiming their inferior, or lack of, knowledge relative to the clinician (Landmark et al., 2015). However, in other cases patients have been found to assert their deontic rights to make the ultimate decision to accept or reject a treatment proposal (Lindström & Weatherall, 2015). Joint decision-making is thus a complex interactional task that requires ongoing negotiation by the clinician and patient/client across the consultation/session.

Research Questions

This chapter builds on this prior conversation-analytic work on epistemic and deontic rights in interaction by examining how therapists and clients managed decision-making about clients’ future courses of action in real-life, video-recorded CBT sessions. In particular, the analysis aimed to explore (1) how therapists initiated discussions regarding the client’s future course of action; (2) to what extent clients were given the opportunity to propose ideas for their future action(s); (3) how clients responded to therapists within these decision-making sequences; and (4) how clients and therapists’ epistemic and deontic rights over the clients’ experience and future actions were managed within the interaction.

Data and Method

The data for this study is a corpus of 20 audio-recorded CBT sessions involving 9 therapists (1 male and 8 female) and 19 clients (1 male and 18 female, all over 18 years old) who were being treated for depression. One client had two of her sessions recorded (these sessions were with two different therapists). The recordings were collected in a free, university-affiliated clinic in Australia that specializes in CBT treatment. Clients at the clinic have weekly appointments, typically for 8–12 weeks. Each recorded session fell in the middle of the client’s course of treatment, with therapist and client having established a therapeutic relationship. Sessions involved one client and one therapist in each case, with an average duration of approximately 56 minutes. The total time for all recorded sessions combined was 16 hours, 46 minutes. The study was approved by the University of Adelaide ethics committee. Recordings were transcribed using the Jeffersonian transcription system (Jefferson, 2004), and analyzed using conversation analysis. The analysis is based on a corpus of 34 extended extracts that involved discussion of behavior change, in which therapists proposed a future course of action for the client.

The Accomplishment of Joint Decision-Making in CBT Interactions

Across the corpus of recorded CBT sessions there was evidence of variation in whether it was the therapist or client who made the initial proposal for the client’s future course of action. In some instances, therapists invited clients to propose their own idea (see also Chapter 11). In doing so, the therapists claimed a lower epistemic and deontic status relative to the client. More commonly, however, it was the therapist who made a proposal for a future course of action for the client. These proposals set up acceptance/rejection as the relevant response from the client rather than the more open-ended slot where the client could make their own suggestion. In proposing a course of action, the therapists claimed epistemic and deontic status in relation to the decision about what the client should do. The typical response by clients to these proposals for change from therapists was to resist—typically, again, by asserting their deontic right to be more involved in decisions about their own future behaviors.

Therapists Inviting Clients to Make a Proposal for Future Action

In some instances in the corpus, therapists used information-soliciting questions (e.g., Is there anything you could do…) to invite clients to propose a future action. These questions were designed to prompt clients to provide the first suggestion for a possible change they could implement in their lives. For example, the first extract, below, follows an extended troubles-telling by the client concerning her feelings of being overwhelmed. She is dealing with problems involving her daughter; trying to prepare for Christmas; and not getting enough sleep. The extract begins with the therapist’s formulation of the client’s trouble (lines 1–5). The therapist’s information-soliciting question comes at line 12. In each of the extracts, T = therapist, and C = client.

figure a

At the beginning of the extract the therapist is offering a gist formulation (Antaki, 2008) of the client’s trouble (lines 1–5). The formulation validates the client’s prior troubles-telling, while also focusing the discussion on the topic of the client prioritizing other things over herself. The client provides multiple acknowledgments (mm line 2, umhm line 4, mm line 6, and yeah line 8) throughout the therapist’s turn, and at line 10, looks as if she is about to expand on the formulation in some way (I guess: (0.5) it’s:), but then redoes her turn to form another confirmation (yeah that does). The therapist then asks the client a question (line 12), which invites her to make a proposal for a course of action that would address her trouble (Is there anything that you could do ta (0.3) h↑elp with that?). With this information-soliciting question, the therapist claims lesser epistemic and deontic rights than the client in making a decision for future action. She provides an opportunity for the client to draw upon her own knowledge of the situation in order to generate an idea. Thus, she allows the client the right to make the initial proposal and to be actively involved in the decision-making process (see also Chapter 11 of this volume).

A gap ensues in the interaction here (2.8 seconds, line 14), following which the client begins her next turn with I dunno. After some brief hesitation, she then moves to provide an option for behavior change (wri:ting in my list a bit of time out time), thus orienting to the preference for provision of information that was set up by the therapist’s interrogative. The therapist responds with an okay (line 17) that accepts the client’s suggestion. In providing this acceptance, the therapist reasserts her involvement in the decision-making process. Not only has the therapist provided the client with an opportunity to suggest a behavior change, but the therapist can then accept the change and thus “have the last word” in the third position. This third-position turn shows the therapist managing her deontic right to accept/reject the client’s proposal for future action, thus also maintaining her involvement in the decision-making process.

As the sequence progresses, the decision-making process becomes a more complex, and therapist-guided, interactional accomplishment. The extract below is an extension of Extract 1a.

figure b

Here, the client’s answer to the therapist’s information-soliciting question, which occurs over lines 3–4, is highly qualified. The client produces three suggestions including maybe jus’ watching a d-v-d maybe, jus’ goin’ out the ba:ck, or maybe jus’ spending some time with Holly, her dog. All of these suggestions display a rather weak commitment to engage in an activity. The therapist responds to the client’s response with some additional work of negotiation. Instead of waiting for the client to finish her turn at line 11, the therapist comes into complete the client’s turn at a point where it is projectable that the remaining component of the turn will be a suggestion of an activity. The client’s intra-turn pause provides the therapist this opportunity to enter. Rather than beginning a new turn, the therapist produces a continuation of the client’s current turn-construction-unit (TCU) (Sacks, Schegloff, & Jefferson, 1974). Lerner (2004) has demonstrated how such “anticipatory completions” work to achieve a heightened sense of affiliation between participants in interaction, and this is what appears to happen here. Therapist and client are not just sharing ideas; they are sharing turns. Although the therapist has suggested the idea of going to the beach, this has occurred as a completion of the client’s turn. Additionally, going to the beach is something that the client had said she enjoys doing 10 minutes earlier in the session. The therapist thus draws on her knowledge of the client (based on the client’s telling earlier in the session) to make a proposal for the client’s future action. The therapist’s deontic right to make this proposal is thus based on her shared knowledge that the client would enjoy going to the beach.

The client responds to the therapist’s candidate completion with a loud confirmation, YEAH. She then partially repeats the completion, reinstating her epistemic authority over the turn’s talk (Lerner, 2004; Stivers, 2005). The client then adds that she might even do that. The use of even in this formulation highlights that the therapist’s completion had not been exactly what the client had intended but that it is accepted anyway. In using an anticipatory completion, the therapist has also been able to achieve affiliation with the client. Rather than making an independent suggestion, it is as though the therapist has “read the client’s mind.” And the client is able to produce a third-position acceptance of the idea, to reinstate her contribution to the suggested course of action.

In sum, across this sequence, it can be seen that both therapist and client join in the decision-making process. Even when the sequence becomes more therapist-guided (than at the outset with the initial open-ended information-soliciting question), the therapist’s proposal for the client to go to the beach is generated from prior knowledge of the client’s preferences. The client and therapist remain affiliated across the sequence and there are no signs of client resistance to the therapist’s actions. In fact, it would be difficult for the client to show resistance to the proposed courses of action given that she either proposed the activities herself or mentioned the activity as something she enjoyed only 10 minutes earlier in the interaction.

Therapists’ Proposals for Future Action and Clients’ Resistance

The above sequence can be compared with other, more common, sequences in the corpus where it was the therapist who initially proposed a course of future action without first inviting the client to propose an idea (as seen in Extract 1). Therapists’ proposals in this CBT corpus were typically followed by localized, active client resistance. Clients drew on knowledge from previous experience, and of the current troubling situation, to produce reasons for their resistance, thus displaying their primary epistemic access to the situation under discussion. By indexing their superior epistemic authority, clients invoked their deontic right to reject the therapist’s proposed course of action.

Extract 2 provides an example. Preceding this extract, the focus of talk for most of the session had been around the problematic behavior of the client’s youngest daughter, Leah. The client had claimed that her daughter’s behaviour was contributing to her own depression. Following this sequence of troubles-telling from the client, in the extract below, the therapist launches straight into a proposal for future action for the client: that the client have some one-on-one time with Leah, and ask her partner, Pete, to keep her elder daughter, Alison, occupied so that this can happen.

figure c

The therapist’s proposal (lines 1–10), framed as an interrogative, is downgraded with an epistemic marker (do you think), the use of a low-modal operator (could), and downgrading devices (quite, just). In designing the proposal this way, the therapist highlights its contingent nature (Curl & Drew, 2008), inviting the client either to accept or reject it in response. In making her suggestion in this way, she thus somewhat shares the deontic right to make the decision with the client (Stevanovic & Peräkylä, 2012). However, the therapist is the one who provides the initial proposal for future action here, rather than asking the client for her own ideas (as seen in Extract 1) and so, in this way, the therapist displays a higher epistemic and deontic stance in relation to the client’s future actions. Deciding to agree with a proposal for future action is not the same as choosing your own action (Pilnick, 2008).

The client’s response (across lines 11–37) suggests that she may not be satisfied with the therapists’ claimed deontic status, as she seeks to establish a stronger deontic position within the interaction. Her response draws upon specific knowledge about her daughters to resist the preconditions of the proposal. The client’s reason for resistance is framed as an inability to comply account (Heritage, 1984): She is unable to accept the proposal because she knows, from experience, that her other daughter (Alison) does not respond well to being left out (see also Chapter 5). The client draws upon several high-modality adjuncts to show that her account is not based on knowledge of one particular occasion, but on her knowledge of how things consistently are (Halliday, 1985; He, 1993). When describing her daughter’s feelings toward her sibling, the client uses the adjunct always several times across lines 14–17. Then, in her expansion at line 21, the client uses the adjunct every to describe occasions when she has tried to talk to her daughter about her younger sister. Again, at line 33, the client states that she always seems to leave her older daughter out. In producing her account as a factual and generalized description of what happens in her household, the client displays her superior direct access to knowledge of what everyday life is like in her house, and of how her daughter generally reacts to conversations such as the one the therapist is proposing. By indexing her superior epistemic authority in the domain of her experience, the client is able to invoke her deontic right to reject the therapist’s proposed course of action. It is easier for the client to reject the therapist’s proposal here, compared to Extract 1, because, rather than inviting the client to provide initial ideas for proposed future action(s), the therapist has instead proposed her own idea. In doing so, she displayed a higher epistemic and deontic stance in an environment where it is the client who has the ultimate epistemic and deontic authority over her own experience and future actions.

Another example can be seen in Extract 3. Prior to this extract, there has been a troubles-telling from the client about her visit to a financial planner who is investigating whether she can buy her own home and thus move out of her parents’ house as she is currently experiencing some conflict with her parents. The client has told the therapist that the financial planner will not be reporting back to her until closer to Christmas, which is six weeks away. The fragment follows on from the therapist having delivered a gist formulation (Antaki, 2008) of the client’s trouble, which the client has confirmed.

figure d

The therapist first delivers an account for her proposal based upon the client’s preceding troubles-telling. Pre-proposal accounts set up the delivery of a subsequent proposal by first stating the problem to be solved (Houtkoop-Steenstra, 1990; Waring, 2007). The therapist’s pre-proposal account orients to the delicate nature of delivering the subsequent proposal. The therapist gets acceptance from the client of the candidate stated problem, which provides her with a go-ahead to deliver her proposal. Although the interrogative is grammatically structured for a “yes” response, in launching it with the phrase is it worth (line 10), the therapist downgrades her epistemic authority over the issue, relative to the client. However, the therapist is the one proposing the idea here for the client to accept/reject. This sets up a different sequential response to instances in which the client is invited to propose her own ideas for a future course of action.

The therapist completes her incremental turn at line 14, and there is a gap of 0.4 seconds where the client does not respond. When the client takes her turn she begins with Well which may be indicative of either a forthcoming complex response or a dispreferred response to the interrogative proposal (Schegloff & Lerner, 2009). In this instance, the well preface appears to be a marker of the upcoming dispreferred response given the silences from the client at lines 12 and 15, and the Uh preface of the client’s abandoned turn at line 13. The client responds to the interrogative with a softened “no” response: not really. The client’s subsequent account for not accepting the proposal involves a specific reason from her own life as to why she is unable to accept the proposal: in the six weeks I’m off I don’t get pai:d. This account draws on an aspect of the client’s life that the therapist could not have known, and is something that the client has no control over. With this resistive account, the client thus displays her epistemic authority over her experience. The reason provided for the resistance is also a factor that is essential for the proposal to be acted out: the client cannot find new accommodation outside of her parents’ home without money. The client thus uses her epistemic authority over the situation to invoke her deontic right to reject the therapist’s proposed future action.

In Extracts 2–3, the therapist proposes a suggested course of action to the client without first inviting the client’s own ideas. Although the therapists use several interactional resources to downgrade their epistemic and deontic stance within their proposal turns, they still claim epistemic and deontic authority by being the one to propose the future action. In each case, the clients resist the therapist’s proposal by providing an account that displays their authority in the epistemics of experience, and thus their ultimate deontic right to reject the proposed course of action.

Discussion

Joint decision-making can be a complex interactional task for therapists as they face the dilemma of needing actively to guide the trajectory of the therapeutic interaction, as well as offering clients the opportunity to be involved in decision-making by conceiving their own idea(s) for change. Although therapists carry professional authority, clients have expertise in relation to their own life experiences. This distinction has been referred to as the “epistemics of expertise” in coordination with the “epistemics of experience” (Heritage, 2013b; Lindström & Weatherall, 2015). When implementing joint decision-making, therapists could risk losing sight of the therapeutic goals of the session. This chapter has shown, however, that there can be consequences for the interaction when therapists propose courses of future action to clients without inviting clients to propose their own ideas in decision-making about behavior change.

This chapter has shown how therapists sometimes use information-soliciting questions to provide clients with an opportunity to suggest behavior change. Such questions positioned the client as the knowledgeable party in the interaction; as the one who would know how to change her own behavior. They thus set up a shallow deontic gradient, allowing the client rights to suggest a behavior change. This type of turn structure has important resonances with the theory underlying CBT (Wright et al., 2006), where joint decision-making is accomplished by therapists engaging clients through the use of Socratic questioning (Wright et al., 2006). Socratic questioning involves asking a series of inductive or open questions in a form that does not provide answers to which the client can respond, but which requires the client’s direct input. Extract 1 provided an example of such questioning. Although the therapist guided the negotiation, the client was first provided the opportunity to suggest ideas for future actions. In this way, both client and therapist proposed suggestions for the client’s future action, participating in joint decision-making across the sequence. Therapist and client appeared aligned and affiliated throughout these sequences, with little or no hearable signs of client resistance.

It was, however, more typical for therapists to propose their own suggestions for clients’ future course of action without first asking for the client’s own ideas. In doing so, therapists set up a relatively steeper deontic gradient for the decision-making process, implying that they had the right to tell clients what they should do. Even when therapists attempted to frame their proposals in a way that somewhat shared deontic rights (by designing their proposal as a question, or including epistemic markers), they still claimed a higher epistemic and deontic status by proposing a future course of action based on their own thoughts/ideas, rather than the clients’. In this way, the therapists’ proposals set up a more unilateral decision-making process, rather than joint decision-making (Collins, Drew, Watt, & Entwistle, 2005). The therapist had not attempted to acquire the client’s opinions or preferences before proposing a suggested course of action. These proposals were typically resisted by clients in ways that asserted their epistemic and deontic authority over the situation. Clients made it clear that they were the ones who had the ultimate right to decide what behavior changes they would implement. As Stevanovic and Peräkylä (2012) have previously shown, second speakers may not be satisfied with deontic symmetry; they often seek to establish a stronger deontic position within interaction. These findings support Pilnick’s (2008) idea that, for clients in institutional settings, deciding to agree with a proposal for action is not recognized as the same, and not responded to in the same way, as choosing their own action.

Clients drew on knowledge from previous experience, and of the current troubling situation, to produce reasons for their resistance, thus displaying their primary epistemic access to the situation under discussion. In looking at the detail of clients’ resistive accounts, we also identified several resources that were repeatedly used to display this epistemic stance. These resources included high-modality terms and generalized clauses (e.g., I’m sure, always, every) (Halliday, 1985; He, 1993), and direct reported speech (Clift, 2006). In examining clients’ resistance to therapists’ proposals, we can see the interplay between claims of epistemic and deontic stance. Clients’ resistive accounts were grounded in their superior knowledge of their own experience, and this knowledge allowed them to invoke their deontic right to reject specific future actions. In this environment, clients’ epistemics of experience could trump therapists’ epistemics of expertise. These findings build on developing work on the interplay between deontics and epistemics in interaction (e.g., Antaki, 2012; Heritage, 2013a; Landmark et al., 2015; Lindström & Weatherall, 2015) by illustrating some ways in which epistemic and deontic stance can be managed by parties in second-position resistance responses.

Demonstrated patterns in clients’ resistance of therapists’ proposals for behavior change have important implications for CBT practice. Analysis has demonstrated that when clients resist, they are not only concerned with rejecting the specific proposed change but also with claiming their epistemic and deontic stance in relation to the matter. Such resistance can obstruct the progression of therapy goals, minimize the degree of success in implementing behavior change, and create a poor relationship between therapist and client (Beutler, Moleiro, & Talebi, 2002; Muntigl, 2013; Safran & Muran, 1996). Therapists might therefore benefit from understanding the subtle implications carried by proposals for behavior change, as clients appear to be sensitive to such issues in the way that they frame their responsive turns.

There are similarities, here, to other healthcare settings such as doctor–patient interactions. Although patients in healthcare settings may be willing to defer to a doctor’s specialized medical authority (e.g., see Landmark et al., 2015; Toerien et al., 2013), patients may also sometimes draw upon their deontic authority to resist treatment recommendations, for reasons grounded in the lifeworld of the patient (Lindström & Weatherall, 2015). The findings in this chapter similarly show how the therapy session is a complex epistemic and deontic environment where both parties must manage their own knowledge and rights in relation to the activities being accomplished. Close analysis of the present corpus has shown how the different interactional ways in which therapists structure the client’s involvement in the decision-making process for future behavioral change can have significant consequences for the trajectory of the therapy session.