Strategies for working with resistant ambivalence

Our search for ways of working with resistant ambivalence was guided by this last assumption. We particularly sought less directive approaches that were based on empathy and support. We found two well-developed treatment approaches that seemed particularly appropriate, both of which have a supporting body of research associated with them: Motivational Interviewing (MI) (Miller & Rollnick, 2002) and Two-Chair work used in Emotion-Focused Therapy (EFT) (Greenberg, Rice, & Elliott, 1993; Greenberg & Watson, 2006)2. Engle and Arkowitz (2006) discussed other potentially useful strategies as well. However, because these are not as well-developed and lack a solid research base, they will not be discussed here.

One particularly interesting feature of both the MI and Two-Chair approach is the variety of ways that they can be employed: as a pre-treatment to other established approaches such as CBT to prepare clients for change by reducing resistant ambivalence, as a stand-alone treatment, and combined or integrated with other therapies.

In addition, both approaches are compatible with the integrative model that we have proposed. They both emphasize the importance of a therapeutic relationship characterized by acceptance, support, and empathy. They share a common emphasis on the importance of ambivalence and of client agency in change. Perhaps most importantly, they both use the therapist's role to tap into the client's inner resources to effect change, rather than taking the role of external change agent. In this sense, both are client-centered, seeking to make the client the advocate and agent of change, and avoiding therapist directiveness that may elicit resistant ambivalence.


Motivational Interviewing, developed by Miller and Rollnick (1991, 2002), began as a way of working with alcohol and substance abuse problems. However, it is now being expanded to a number of other areas (Arkowitz, Westra, Miller, & Rollnick, in press) including depression and anxiety disorders (Arkowitz & Westra, 2004), eating disorders (Treasure & Schmidt, in press), compulsive gambling (Hodgins, in press), and suicide (Zerler, in press).

Principles and Strategies of MI

Miller and Rollnick (2002) describe MI as a client-centered and directive approach. It is client-centered in its basic humanistic underpinnings, in how people and change are viewed. It also draws heavily from client-centered therapy including its emphases on reflection and empathy. It is directive only in a subtle sense. MI does not try to directly influence people to change. In fact, any therapist who adopts the stance of "change advocate" is not doing MI. Instead, the MI therapist seeks to increase intrinsic motivation and reduce ambivalence about changing. With these changes, it is assumed that behavior change will occur naturally, with the client perceiving the locus of change as internal rather than residing in the therapist. Miller and Rollnick (2002) describe four basic principles of MI: (1) express empathy, (2) develop discrepancy, (3) roll with resistance, and (4) support self-efficacy.

In MI, empathy is primarily communicated through reflective listening (or accurate empathy) as described by Carl Rogers (1951). Underlying this principle of empathy is a client-centered attitude of "acceptance," wherein client ambivalence or reluctance to change is viewed as a normal part of the human experience rather than as pathology or defensiveness. Reflective statements are more than a simple "parroting" of what the clients says. Instead, they are guesses at the client's meanings and experience. For example, a client might say: "My parents have been bugging me a lot this week." The therapist might respond with the statement: "So, you've been pretty angry at your parents this week." The therapist's statement is just a small step beyond the client's statement and makes a reasonable guess that the client feels angry when bugged by parents. But it's an attempt to deepen the client's experience and check the therapist's understanding of it.

Developing discrepancy, the second principle of motivational interviewing, is where MI begins to depart from classic client-centered therapy. A key goal in motivational interviewing is to increase the importance of change from the client's perspective. This is accomplished using specific types of questions, along with selective reflections that direct the client toward the discrepancy between his/her problem behavior and important personal values. For example, a drug-addicted woman may see, through the therapist's reflections, that such behavior conflicts with her strong value on being a good mother to her child.

The third basic principle of motivational interviewing is to roll with the resistance rather than opposing it. This involves accepting the client's concerns about changing as valid without trying to directly challenge them. A useful strategy in this regard is the use of a decisional balance framework in which the pros and cons of change from the client's perspective are both fully explored. During work on decisional balance, and throughout MI, the therapist tries to highlight behavior-value discrepancies and elicit, reflect, and reinforce talk relating to commitment to change (Amhrein, Miller, Yahne, Palmer, & Fulcher, 2003) in order to help tip the balance toward change.

The fourth guiding principle of motivational interviewing, therefore, is to enhance the client's confidence in his or her ability to cope with obstacles and to succeed in changing. This confidence, which Bandura (1997) has described as self-efficacy, is an essential element in motivation and a good predictor of treatment outcome. This too is done by the therapist evoking (e.g., asking about success in past change attempts), reflecting, and reinforcing statements relating to confidence in changing.

When the therapist thinks the client may be ready for change, the therapist elicits the client's thoughts about how to go about making that change. Throughout MI, the therapist is a consultant to the client's change program, but the client is always in the lead. As a consultant, however, the therapist can and should offer advice and suggestions about change strategies that may be helpful. The therapist asks to be "invited in" by statements and questions like: "I have some thoughts about what's been helpful for other people with similar problems that might be helpful for you. Would you be interested in hearing them?" The client is then free to accept, reject, or modify the therapist's suggestions.

While MI has been used as a stand-alone treatment to prepare clients for change and to work with them during the action stage, it has also been used extensively as a prelude to other more directive treatments like cognitive-behavior therapy and 12-step approaches (see reviews by Burke, Arkowitz, & Menchola, 2003; Hettema, Steele, & Miller, 2005). Several studies have found that MI pre-treatments enhance the outcomes of these therapies, even compared to other types of pretreatments. MI seems to increase motivation to change sufficiently that clients are "ready to go" and respond favorably even to more directive treatments once they are ready.

Case Illustration

This case illustrates the use of MI as a stand-alone therapy for depression (see also Arkowitz and Burke, in press). Brad sought therapy with the first author for depression and anxiety. He was a college junior who lived at home and who felt "lost" because he had no idea what he wanted to do with his life after he graduated. In the first few sessions, the therapist primarily utilized open-ended questions and reflections to hear Brad's story and to understand his frame of reference. In the process, we began to examine the decisional balance relating to his depression. Brad listed the many obvious disadvantages of being depressed including feeling sad, not being able to do things, having no interest in anything, etc. The advantages of being depressed were explored from Brad's point of view in a nonjudgmental manner, and without the therapist advocating for the advantages of change. Brad reported that if his depression and anxiety improved, he would have to deal with the difficult question of what to do with his life. He stated that this was one of the issues that may have precipitated the depression. In addition, if his symptoms improved, he expected that his parents would "be more on my case to get a job or do something, and I don't know what I want to do." Using reflections, value-behavior discrepancies, and elicitation and reinforcement of change talk, Brad began to engage in behaviors (e.g., socialize more) that he believed would help him reduce his depression.

Research on the Efficacy of MI

A meta-analysis by Burke, Arkowitz, and Menchola, (2003) reviewed 30 outcome studies of MI with problems that included alcohol and drug abuse, smoking, diet and exercise, and HIV-risk behaviors. They found that across problem areas, MI was more effective than no treatment or placebo, and as effective as other treatments to which it was compared. Interestingly, the effect size for MI as a prelude was greater than for MI as a stand-alone treatment. When MI efficacy for specific problem areas was examined, results supported the efficacy of MI for problems involving alcohol, drugs, and diet and exercise. Results did not support the efficacy of MI for smoking and HIV-risk behaviors, although it should be noted that the number of studies in each of these categories was quite small. A later meta-analysis of 72 studies (Hettema, Steele, & Miller, 2005) also found considerable support for the efficacy of MI.

More recently, Westra and Dozois (in press) compared CBT for anxiety disorders with and without an MI prelude. The MI-CBT group showed significantly greater reductions in anxiety, scored significantly higher on a self-rating measure of CBT homework compliance, and significantly increased scores on a measure of optimism after MI. A greater percentage of the MI-CBT group completed treatment than the CBT only group, but this difference only approached significance.

Overall, there is considerable support for the efficacy of MI in a variety of areas. However, the mechanism that accounts for this remains unclear. Although MI does work with ambivalence, we cannot conclude that it was the resolution of ambivalence that accounted for the observed changes since we do not yet have adequate measures of resistant ambivalence. Further research will be needed to ascertain this and to clarify why MI works.

The Two-Chair Approach

The Two-Chair method has its roots in the Gestalt therapy of Perls, Hefferline, and Goodman (1951). This therapy makes use of "experiments" that are semi-structured novel experiences related to the problem under discussion and constructed jointly by the therapist and client. They are discovery-oriented and designed to increase the client's awareness of feelings and change dysfunctional ways of thinking about themselves and others. They may take place during or between therapy sessions. For example, Perls and colleagues (1951) described the use of two-chair experiments in which the client engages in a dialogue between conflicting aspects of the self. The client may take one role (e.g., the critical self) in one chair and another (e.g., the rebellious self) in the other chair, with the therapist facilitating a dialogue between the two. In recent years, Greenberg and his associates have expanded on this work and developed Emotion-Focused Therapy (EFT) (e.g., Greenberg & Safran, 1987; Greenberg, Rice, & Elliott, 1993; Greenberg & Watson, 2006) that builds upon Rogers' (1951) client-centered therapy as well as Gestalt therapy. The part of their work that is most relevant to ambivalence is what Greenberg, Rice, and Elliot (1993) call "conflict splits." Here, there is a sense of struggle between the two selves that pull a person in different directions, e.g., "Part of me wants this, but another part of me wants that." We built on this work to develop a two-chair procedure specifically aimed at resistant ambivalence and its resolution.

In our version of the conflict split, one self advocates for change and another self struggles against change. The markers we employ for inviting a client to participate in this experiment are statements and behaviors that suggest ambivalence about change. The dialogue is structured so that the client takes turns speaking from each of two chairs that face one another. In one chair, the client is asked to speak from the perspective of the part of self that moves toward change (the "Change Self") and, in the other, the part of self that struggles against change (the "No Change Self"). Clients are usually fairly aware of the former, but much less aware of the latter. The experiment is more "discovery-oriented" than hypothesis-testing, and is aimed at both bringing the contents of both sides to full awareness, and resolving conflicts between them. The role of the therapist is to facilitate this dialogue and work from the client's perspective rather than imposing an external perspective. As a facilitator, the therapist does not side with either of the selves. More detailed discussions of the two-chair experiment can be found in Engle and Arkowitz (2006) and Greenberg, Rice, and Elliot (1993).

The dialogue is usually continuously evolving as the experiment proceeds. For example, we have seen experiments start with the Change side expressing desires to change and the No Change expressing fears of change. This then developed into a Should and Reactant dialogue with the Should side often evolving into a critical parent. It is important for the therapist to follow the client's lead in these experiments, but also to be attuned to underlying meanings and feelings and to help bring these into awareness.

The initial stage of the Two-Chair experiments emphasizes separation of the selves in each chair. Clients will often shift into the self in the other chair, while remaining in their current chair. When this happens, the therapist tries to establish separation by asking the client to move to the other chair so that the different selves remain coherent and relatively distinct.

Often, the first few minutes of the dialogue involve the two selves talking at rather than with one another. They are each staking out their territory without being particularly responsive to the other. At this point, the task of the therapist is to encourage contact between the selves. To accomplish this, the therapist might make suggestions like "Tell her (the self in the other chair) how she makes you feel when she says that."

As contact is made, the therapist tries to encourage expression of wants and needs, particularly in the self that is experiencing more emotion (e.g., the reactant self being criticized by the should self). For example, the self that was initially reactant may respond to the should self by saying "Your criticisms make me even more afraid and unwilling to change. I need you to back off from these criticisms. If you do I might be more willing to try changing."

Toward the end of a successful experiment, the client will often become aware that both sides are trying to help, but in different ways. The two sides begin to negotiate with each other in a genuine attempt to resolve the conflict between them and work together in the person's interests. Resolution of the conflict leads to an integration of the two sides.

Throughout the process, the therapist is particularly attuned to tacit emotions, and gently encourages the client to express them. Emotional expression facilitates greater contact between the two sides, and often leads to the awareness and expression of material that is deeply felt, but censored from expression in the nonemotional state.

Case Illustration

Sarah was an attractive single 20-year-old with a young son. She had been in a two-year relationship with a 36-year-old divorced man whom she described as emotionally abusive. He had other relationships with women, disappeared for days at a time with no explanation, stole money, and lied to her. He yelled at her and demeaned her for things that were not her fault. She made several failed attempts to end the relationship, always resuming it after a couple of weeks. Her parents and friends urged her to end the relationship. Although she knew they were right, she was unable to do so. Her statements during the interview clearly reflected her ambivalence about staying with him or leaving.

The Two-Chair experiment took place over several sessions. In speaking from the Change side, she described the boyfriend as abusive, unfaithful, and untrustworthy, leading to her unhappiness in the relationship and a desire to end it. The No Change side stated that "You know he loves you" and "... maybe things will change, maybe things will work out." As the experiment continued, the Change side became angry at the other side for keeping her trapped in the relationship, while the No Change side was sad, clinging to belief that he could change and the relationship could work. At this point in the dialogue, there were the beginnings of contact, but both sides were still firmly entrenched in their respective positions. She renamed the two sides the "emotional self" that wanted to stay and the "logical self" that wanted to leave.

At the end of the second dialogue there was an emerging understanding between the selves, and the anger was mostly gone. In response to a prompt for the logical self to tell the emotional self what she needed from that side, she said: "You're a part of my life that I need. I need your ability to trust. I need your ability to love. I need who you are. I would like for us to be able to work together and not be at such odds. I'm sure there is somebody out there that will fit both of our needs ... and not just yours, and not just mine." This was the first overture toward resolving the conflict between the selves and working cooperatively. The dialogue continued to reflect an increasing sense of cooperation between the two selves. In the following session, there were remnants of fear about "being a loser" if she left the relationship and if her boyfriend very quickly found someone else and she did not. However, she also expressed how tired both sides were of the endless struggle. A strong desire to meet in the middle emerged, with a bridge being a central image. Subsequently, the logical self moved more toward accepting the idea of her leaving and the emotional self asked for support and help from the logical self in leaving. The dialogue continued with mutual support as the theme and without conflict. The different selves were now working out how to give and receive the support needed to leave. Sarah subsequently did leave the relationship, and a year later had not gotten back together with him, despite frequent requests on his part to do so.

Research on the Two-Chair Approach

There have been several studies that bear on efficacy of the Two-Chair procedure. Early studies (Greenberg & Clarke, 1979; Greenberg & Dompierre (1981); Green-berg & Higgins, 1980) compared empathic reflection (derived from Carl Rogers' Client-Centered Therapy) with Two-Chair work for clients who were experiencing ambivalence about a decision. Both groups made considerable progress toward behavioral goals, but were not significantly different from one another.

Another study used volunteers who were having trouble making a difficult decision. Greenberg and Webster (1982) gave these subjects a six-week treatment consisting mainly of Two-Chair work. After the treatment, subjects were divided into "Resolvers" and "Non-Resolvers" based on whether they had manifested three components of a proposed model of conflict resolution during treatment: expression of criticism by one side; expression of feelings and wants by the other; and a "softening" in the attitude of the critic. After treatment, Resolvers were significantly less indecisive and anxious than Non-Resolvers.

Clarke and Greenberg (1986) employed subjects who sought counseling to help them resolve a conflictual decision. They were randomly assigned to two sessions of either a Two-Chair intervention, a problem-solving cognitive-behavioral intervention (CBT), or a no-treatment control group. The Two-Chair group improved more than the CBT group and the control group on one measure of indecisiveness. The two groups did not differ significantly on the other measure; both improved significantly more than the no treatment control group.

Arkowitz and Engle (1995) conducted a small single-group pilot study on the Two-Chair procedure for resolving ambivalence. People who were having trouble making an important change in their lives were recruited from advertisements in the campus newspaper. Seven respondents were deemed appropriate for the study. Their focal problems included: two women who wanted to leave what they considered to be bad relationships but were unable to do so; one smoker who wished to stop; one who was trying to lose weight; one who was messy to the point of embarrassment about having people visit her; one who was indecisive in her career choice, and a depressed man who was unable to move ahead on many of the goals he set for himself. Each subject received four half-hour sessions devoted almost entirely to the Two-Chair procedure applied to the focal problem. Of the seven, there were clear resolutions and behavioral changes in four, improvement but short of full resolution change in two cases, and no change at all in one.

Greenberg and Watson (1998) conducted a study of patients who met the criteria for Major Depression. They compared 15-20 sessions of either Client-Centered Therapy (CCT) or Emotion-Focused Therapy (EFT), which includes the two-chair method. The EFT therapy included a base of CCT in the context of which the therapist used several different Gestalt techniques including, but not limited to, Two-Chair dialogues for conflict splits. Overall, both groups showed considerable improvement with treatment. The effects seemed clinically significant when compared to effect sizes in other treatment studies of depression that employed a No-Treatment Control Group. At post-treatment, the EFT group showed greater improvements in self-esteem, interpersonal functioning, and symptom distress than did the CCT group. Further, EFT seemed to work faster, showing greater changes than CCT at mid-treatment. Treatment gains were maintained at six-month follow-up, but differences between the two treatments disappeared. A replication and extension of this study (Goldman, Greenberg, & Angus, in press) demonstrated that EFT was significantly more effective in the treatment of Major Depression that CCT, and this difference was maintained at follow-up. Another study by Watson, Gordon, Stermac, Kaleogerakos, and Steckley (2003) compared EFT and CBT in the treatment of Major Depression. Overall, both therapies were equally effective, but several measures showed an advantage for EFT.

Although results for the Two-Chair procedure are promising, the jury is still out on whether it is efficacious. As with MI, it is also not clear that this procedure effects change through the resolution of ambivalence.

Free Yourself from Panic Attacks

Free Yourself from Panic Attacks

With all the stresses and strains of modern living, panic attacks are become a common problem for many people. Panic attacks occur when the pressure we are living under starts to creep up and overwhelm us. Often it's a result of running on the treadmill of life and forgetting to watch the signs and symptoms of the effects of excessive stress on our bodies. Thankfully panic attacks are very treatable. Often it is just a matter of learning to recognize the symptoms and learn simple but effective techniques that help you release yourself from the crippling effects a panic attack can bring.

Get My Free Ebook

Post a comment