Background

Beginning in 1982, Insoo Kim Berg, Steve de Shazer, and their colleagues at BFTC (and their clients) have been developing a solution-focused approach to doing therapy that is significantly different from other approaches (Berg & Miller, 1992; de Shazer, 1985, 1988, 1991). In comparison to other approaches, there are differences in theory, in assumptions, in practice and resulting differences in the way of conducting sessions; there are fewer sessions involved¹ and longer intervals between sessions (de Shazer, 1991). The approach encompasses a different view of therapy, a different view of clients, and thus a different view of therapists and their tasks.

Prior to 1982, we were using a problem solving model of therapy based on the idea that the therapist’s intervention needed to fit with the patterns of the client’s problem in a very particular and specific manner (de Shazer, 1982). Like most models of therapy, we focused on describing the problem or the client’s complaint, based on the assumption that the nature of the problem (or at least the description of the patterns of the problem) determined what the intervention should look like and, therefore, what the eventual resolution should look like. Like many other models, we saw the client’s complaints (including the patterns around them) as continuing to occur because they seemed to follow rules similar to the rules that can be seen to govern and pattern all human interactions.

While many therapists group problems into categories, i.e., 1) psychological problems with obscure origins, 2) psychological problems with obvious origins, 3) organic disorders, or 4) by standard diagnostic categories, we tended to group cases together according to their responses to our initial intervention, particularly the homework tasks we gave at the end of the session. The form the intervention took depended on behavioral sequences around the complaint and the meanings the clients gave to their situation.

From a Problem-Focus to a Solution-Focus

In mid-1982, we had a case that forced us to radically change our minds about therapy, clients, therapists, problems, complaints, and solutions (a process that has taken a number of years and is still continuing). After the therapist asked “What brings you in today?” the mother began to describe her concerns but, before the therapist could begin to elicit a description of the patterns around that complaint, the father interrupted. He had his own concerns but, before the therapist could either return to mother’s concerns or elicit a description of father’s, the eldest child interrupted with her own concerns. During the course of the session, the family members continued this pattern and by the end of the session, the five of them had listed 27 different concerns. However, none of these were well enough described for the therapist and his team to use as a basis for an intervention task. (Every therapist who has interviewed more than two or three families has had a similar experience.)

It was clear to the team behind the see-through-mirror that this family had a “problem”: They clearly had something to complain about and their coming to therapy seemed appropriate. However, none of their concerns had yet been put into a form that could be used as a guide for an intervention. Fortunately, some of the team had been influenced by the works of Milton H. Erickson (Haley, 1967) who suggests that if efforts made by the therapist to clear up the client’s vagueness have led only to more vagueness, then the therapist should do something different and be at least as vague as the client has been and, eventually, this will lead to at least some clarity. The team thus developed this task which came to be known as the Formula First Session Task (FFST):

“Between now and next time we meet, we want you to observe, in such a way that you can tell us about it next time, observe what happens in your life that you want to continue to have happen” (de Shazer, 1985).

Two weeks later, when the family returned, my colleagues and I were surprised when the family described 27 different things that had happened that they wanted to continue to have happen. 25 of the 27 were directly related to the 27 concerns listed during session one. When asked, the family members said that they thought the problem that brought them to therapy was solved and therefore no more sessions were needed. Six months later, 20 of the 27 things listed in session two were seen as continuing to happen.

According to our problem solving model, this intervention should not have worked since it was not related to the patterns of any of the complaints the family members listed. But work it did! Therefore, we were faced with a choice: a) We could pretend it did not happen, ignore it, and thus describe it as some sort of fluke, or b) we could investigate it and thus risk changing or even rejecting the model/theory we had been constructing for 15 years. We choose that latter course.

Research<>Practice<>Theory²

During the course of the next two years we gave this same task to hundreds of clients (individuals, couples, families) and found that, about 90% of the time, clients will report having had something happen between sessions one and two that they want to continue to have happen. (It was, in fact, our rule to give this task at the end of the first session unless we had very good reasons not to: We gave the formula first session task to two-thirds of our clients during this investigation. Most frequently, they listed from 7 to 11 things in the second session.) Frequently these events were unrelated to the client’s problems and complaints and yet are seen by the client to make things significantly better: We learned that what clients used to judge that things were significantly better are frequently very different from what professionally trained therapists or researchers would use or would think that the clients should use.

At this point we lost our highly valued connection between complaint and solution. Clearly, if this generic task can lead to descriptions of significant improvement regardless of the complaints involved, then the complaint does not determine the process of solution. The process of solution development can begin without the therapist knowing or the clients agreeing what the problem was or what the clients have to complaint about.

This investigation confirmed for us the idea that what the client does and what the therapist does while doing therapy, what goes on in therapy is more important to successful therapy than the problem or diagnosis or the client’s situation or personality.

Central Premises

The central philosophy of solution-focused brief therapy (de Shazer, 1985, 1988, 1991) can be summed up as follows:

These three premises may seem overly simple or perhaps even simple-minded when it comes to doing therapy: i.e., therapist and client working together (We doubt the usefulness of calling our work “therapy” but no viable alternative is readily apparent). Obviously, a tradition based on such simple premises is frequently going to be seen as deeply flawed. Surely the resolution of overwhelming, chronic human problems cannot be so simple. Interestingly, applying them in the “real world” of doing therapy certainly is not easy.

Premise 1: If it ain’t broke, don’t fix it

This premise is so central to life that it seems self-evident. In fact, there should be no need at all for this premise which, among other things, we take to mean the following: If something is not a problem for the client and, therefore, the client does not complain about it, then — no matter how obviously problematic that something might be in the eyes of the therapist or “society” — it is none of the therapist’s business. Unlike most, or at least many, other types of therapy, solution-focused brief therapy is non-normative and thus the need for explicitly stating this premise.

Premise 2: Once you know what works, do more of it

On the surface, this premise should be so self-evident that stating it seems stupid. At BFTC, we serendipitously found out that, when asked in the right way and/or at the right time, many, perhaps most, clients will report that there are times when the complaint/problem does not happen, even though they had every reason to expect it to happen! And that there are times when things, in general, go along quite well given the clients’ life circumstances. Whatever it is that the clients are doing at these exceptional times are exactly what the client needs to do more of: The client needs to continue doing what works.

Although the premise itself should go without saying, it is actually broader in its implications and applies to doing therapy as well. For brief therapists it means that a) if you know a solution that works, do not un-necessarily look for another way: Never forget what works; and b) if what you did in the previous session was effective in the client’s judgment, then you should do it again. Furthermore, the therapist should resist doing anything additional.

Premise 3: If it doesn’t work, don’t do it again: Do something different

Since the problematic situation has been simply and clearly described as “the same damn thing over and over again,” it also seems as if this third premise should be self-evident. However, given the folk-lore idea that “If at first you don’t succeed, try, try again,” this premise cannot be overstated primarily because it seems to run counter to common sense.

Clearly, problems have self-maintaining properties including (so-called) repeats of the same old failed attempts at solution. Obviously, things are not working and Premise 3 of the Central Philosophy suggests that — in such a problematic situation — someone needs to do something different if anything different is going to happen. In fact, when anyone in the problematic situation does something different, anything that cannot be seen as “the same damn thing over again,” then the problem is on the way to resolution, i.e., what was problematic becomes just another example of “the one damn thing after another” that in fact constitutes normal, every-day life.

If progress has not been achieved within a few sessions, our view is that the therapist has also become part of the problematic situation and also needs to do something different since studies from all forms of therapy have shown that if progress is not made by 5 to 8 sessions, it is unlikely that it ever will be made. Since we do not use several sessions to do an assessment, we feel that if progress is not acknowledged by the client by the third session then the therapist needs to pay attention to premise 3. Examples of the therapist doing something different include changing teams, changing rooms, changing who sits where, changing which therapist (from the team) is in the room, radically varying appointment dates and times, changing who is invited to the session, etc. Often the most effective difference is admitting to clients that we do not seem to be helping them so they need to help us.

Case Illustration

Assumptions influence what therapist does or doesn’t do. The following case helps to demonstrate how the assumptions that are held by the therapist direct his thinking and the questions that he asks. The client is a married woman with two grown children. She became blind 10 years earlier as a result of diabetes and recently learned that she has lupus.

Clients generally need some reason to come to therapy: a problem or complaint that they may have about themselves or someone else or a problem or complaint that someone else has about them. If they did not have some reason or justification for being there, they would be mere window-shoppers. However that complaint or problem is simply their ticket to get in the door. Likewise if they have a noise in their car, they have a good reason to go to an auto repair shop. They may not know exactly what they want fixed. While they’re there, they may want the mechanic to also look at the steering. Maybe they’re not sure if the two problems are related so they attempt to describe in detail the symptoms in order to help the mechanic. Clients who come for psychotherapy do much the same: They describe in great detail the problems that are bothering them enough that they feel they need help. Often the therapist assumes that the client is there for relief of their symptoms and becomes frustrated when that is not possible.

In this particular case the client complains about her symptoms from the recently diagnosed lupus. She can not continue certain activities in her life such as knitting and swimming. She does not have feelings in her limbs. And she does not want to go on living like that. Because she is obviously suffering greatly from the combination of blindness and lupus, any therapist with an ounce of compassion would want to help this woman. Many therapists would feel obligated to show that compassion and to make suggestions to help relieve her symptoms. But do we really know what she wants in the way of help from us? Often therapists feel the need to jump in right away to help the client. We might ask questions to find out more about her physical complaints. We might do a suicide assessment, etc. In this particular case, the therapist (Ron Wilgosh)⁴ did not explore those avenues. Instead he listened very carefully for any indications of what the client did want instead of what she did not want and, importantly, he did not make the assumption that he knew what she wanted. She had been focusing on what she already knew she was not able to do. Obviously that added to her frustration and to her limited hope for change. Continued attention to that by the therapist would only add to that frustration; if he were to have done that, he would be joining her in doing more of the same of something that is not working. In fact, the client does not think the therapist can help relieve the symptoms. So her therapist simply acknowledged that he understood what she was saying.

Finding out what the client really wants

Most therapists (at least in the U. S.) are required to have goals for specific cases if for nothing else other than insurance purposes. For the sake of accountability, these goals must be measurable which means that they usually need to be described in behavioral terms. However, few forms of therapy specify the procedure for obtaining goals or the qualities of well-formed goals (Berg & Miller, 1992; de Shazer, 1991). The insurance forms have a space for goals which client and therapist are supposed to agree on together. So why not simply ask clients what their goals are? It has been our experience that this is not very productive for several reasons. Clients are frequently so focused on the problem that the goals are expressed as the absence of that problem or the stopping of something. The former is hard to measure and the latter cannot truly be achieved until death. In fact, defining goals this way often makes matters worse because it causes the client to focus more on undesirable behavior(s) and/or the problem. Often these goals involve very big leaps from the present reality and they usually involve someone else having to change in order to reach the goal. Moreover, it is our belief that if the client could express what they wanted in a different manner, it would be much easier not only to reach that goal but also to know when they are done with therapy. This relates to our third premise in that the “it” that’s not working can be the client’s way of thinking about things.

How best to help the client think in a different way so that they can best accomplish this? We have found that the use of what we call the miracle question (de Shazer, 1985, 1988, 1991) frees the client to think of solutions in the future without being burdened by the problems of the past and present. It is most appropriately used when the client has had sufficient time to describe the reason(s) for being in the therapist’s office, i.e., the complaint or problem. We have found that, if the therapist tries to solve the problem too quickly (often joining with the client in doing more of what does not work), the client will continue to try to convince the therapist that he or she has good reason to be there by continuing to describe the problem. Similarly, if the therapist persists in asking lots of questions about the complaints, the client will continue to be helpful by answering more about the problems. If however the therapist just listens and acknowledges that they have heard what the client says, there usually comes a point at which the client starts to repeat or looks at the therapist for some indication of where to go next: i.e., Do more of the same or do something different? This is the point where we find asking the miracle question most appropriate.

The miracle question is not necessarily a substitute for the goal question; it is a way of helping clients think differently so that they can arrive at solutions that may have not been possible to imagine by focusing on the problems. The goal of the therapist is not to force the client to do the miracle they describe but to open many new possibilities that might lead to solutions. At this point the client and the therapist do not know which possibility might lead to a solution. They do not know what works so they do not know what to do more of; that will come later. So, the therapist does not latch onto any particular idea or attempt to establish priorities. The more numerous the examples, the more detailed the description, and the smaller the responses to the miracle question, the higher the probability that these behaviors or perceptions have already happened to some extent, might happen in the future, or might be noticed as having happened. In order to maximize the possibilities, it is necessary for the therapist to ask additional questions to assist the client in coming up with more and smaller possibilities. When clients come up with large or impossible responses (i.e., “I’ll get the feelings back in my legs and hands”), rather than challenge that possibility which can only bring the client back to the unsolvable problem, the therapist can simply accept that as a perfectly legitimate desire and proceed by asking “What else?”. As this conversation continues, the client is able to think about more and more because, after all, they are only pretending.

Because we are only looking for a small change anywhere in the client’s life, we do not limit our conversation to only the person in the room with us; after all, the client’s life is not limited in such a way. So we ask what we call relationship questions: What the client would notice other significant persons in their lives doing or what the other persons would notice the client doing. In some cases the identified client might not want or expect any difference on their part but want someone else to change. Again the more numerous and minute the responses, the more probable their reality. In this case the therapist continued to explore differences after the miracle that she would notice with her husband and her daughter.

Searching for Solutions

We talk of searching for solutions because the therapist’s goal is not to be able to define the one solution for the client. Rather the client will be the one to know when the solution is found because they will decide that there’s no further need for therapy. But since they usually do not think they have found the solution before coming to us, the question is “How can we help them?”

We believe it is important to help clients develop a rich picture of life without problems because no matter what means they choose to arrive there, they need to know when they are successful. In an analogous way if a person wants to get to Chicago, he has to have an idea how he will know it’s Chicago. If it’s by recognizing certain buildings, he might have to travel by car rather than by plane. The various responses to the miracle question provide a direction or vision toward which the client can focus. With this new focus, the client is better able to see past, present, and future examples of change. In other words, the client and the therapist do not have to decide to do anything new at this point.

So how does this help? Because the client now has a better picture of what they and others will be doing after the problems are solved, we assume that there is some probability that parts of this miracle picture are occurring or have already occurred. Unless the therapist asks, he will never find out.

In one study we asked the question: What have you noticed different about your situation since you called for the appointment? Two-thirds of the clients answered that something had changed in a positive direction and all of those said the changes were related to the reasons they came to therapy and were changes they wanted to see continue to happen (Weiner-Davis, de Shazer, and Gingerich, 1987). In some ways this finding is not surprising since other studies have shown that clients on waiting lists solve their problems without the help of therapists (up to 50% of the time). Knowing that pre-session change takes place so frequently helps us to listen for it and to ask about it in a different way and at a different time: After the miracle question, we ask “Are there times now that small pieces of this miracle happen?” By asking this question now we not only find examples of exceptions to the problems but also incidences of partial solutions: i.e., exceptions related to the goal. In this case, the client mentioned her previous adjustment to blindness.

Another way that the therapist can help the client search for solutions is to ask scaling questions in the first session: On a scale of one to ten, where 10 is when these problems are solved and 0 is when they are at their worst, where would you say things are today? Some therapists might be reluctant to ask this question during the first session for fear of hearing the client respond with a low answer. However the focus is not on the number itself; the actual number is not important. The emphasis is on the positive direction; whatever the client’s response, an inquiry is made as to how they got from zero to that number. (Even if the client’s response is “0” we would inquire about how come it is not “-1” or “-2”? How come things did not get worse? which implies that they did something to prevent things from getting worse.) Clearly, the indication is that they have already started to solve their problem. In addition, it’s helpful to find out what the next small step will look like. It is not necessary to ask or tell the client how to get to that step. Even with clients who say their life is at its absolute worst, picturing the next small step provides them with some hope because this small step is possibly closer to their grasp than trying to eliminate the presenting problem.

Again the usual tendency of therapists would be to try to help this woman by telling her what to do or getting someone else to help her. But is that what she really wants from the therapist? If one listens closely, the indication is that she wants to do for herself and is not real sure what she wants from therapy other than to talk.

So in the message at the end of the session after a short break, we ask her help in thinking about how talking can be helpful in producing the changes she wants. We also offer compliments on how she has tried to adjust to this and other events in her life. This message is no longer called the intervention message because the word “intervention” suggests that our aim is to intervene or interfere in our clients’ lives rather than assist them in finding their own solutions.

Evaluating progress

How do we know this seemingly minimalistic approach works? Since we maintain the assumption that clients have all the necessary tools to construct their own goals and solutions, we have to let them be the judge of their progress. If there has been progress toward their goals, we expect their lives to be better. So we start off second and subsequent sessions by asking them:

It is not unusual for a client to respond by saying “nothing” initially. Given their initial problem-focused view, this response is neither unexpected nor surprising. After all, for clients to think and to notice positive change is a rather large shift for them. It takes some time for clients to be able to think in this different way. Therapists can be helpful by being patient (interesting word since it is usually used to refer to clients in medical settings so maybe it suggests shifting roles) and by allowing the client time to think differently. Our experience has been that approximately half of our clients answer this question in the affirmative and another quarter will answer in the affirmative after the therapist listens for a while for an affirmative description to develop or asks an additional question such as “Which was the best day of the week?” In particular, this client’s initial response is also not unexpected since she was at a zero on the progress scale at the end of the first session. This means that, at that moment, she saw 100% of her life being consumed with problems. Thus, we cannot expect her to see the small relative change if she had been focusing on the problems: e.g., a change from 100% to 99% is only a 1% relative change. However, a change from 0% to 1% in the positive direction is mathematically approaching infinity. Again the therapist’s questions can encourage the client to think in a different way so she can see changes by eliciting more positive change talk, amplifying those changes, and reinforcing the changes.

We generally ask scaling questions after this opening discussion to determine whether or not the changes are related to the goal: i.e., life could be better, but is it related to solving the problems that brought them here. In this case there was a small but discernible improvement in both progress toward the goal (i.e., the client said she had moved from 0 to 1.5) and in her confidence about eventually reaching the goal. Scaling questions often help the client to remember even more positive change. In this case the client described how she made an appointment for a manicure and got there all on her own.

And the client generally knows what the next step will look like. The therapist does not need to overwhelm the client by making lots of suggestions but rather simply needs to support the client’s going in the direction she chooses. Of course, that direction may change in the course of therapy. For clients to know at the beginning of therapy exactly where they want to go is unrealistic; if they did, they probably would not need therapy. For this reason we do not find it necessary to contract with clients a) for a specific number of sessions or b) for specific goals or c) to measure progress on specific goals. To do so would again constrain and limit the possibilities for change.

Ultimately, clients will take enough steps toward constructing a solution satisfactory for them (solving their problems and reaching their goals) that they do not feel the need to continue therapy. In this particular case that happened in the third session. At that point she gave herself a 3.5 on the progress scale. The client said she was doing “better enough” and that she did not feel the need to continue therapy. This “better enough” included getting out and doing more things for herself. Although her “better enough” was only at 3.5 on the scale, the therapist needs to accept her judgment even if he might want more for her: Since only the client can make the judgment that things are “better enough,” any attempt to help her more would be intrusive and likely to meet with failure (premise 2). It is important to remember that, whereas her previous adjustment to blindness took a year just to get out of the house, this adjustment took only a few weeks. Experiencing the various changes she made gave her the confidence that she could be independent. Of course, “independence” for her meant not needing to be in therapy and going back to doing some of the things that had worked before. For instance, she had begun going regularly to a support group and had begun working as a phone volunteer on a local hotline.

Conclusion

Solution-focused brief therapy can be described as a way of developing solutions to human problems and there is an assumption (all too frequently implicit) that any individual in the same situation as the one the client describes would have the same problem. That is, problems are seen as primarily situational and as embedded in language, i.e., problems are more a result of the situation in which the clients find themselves and their definition of the situation, (i.e., how they are talked about) than they are a result of any causative underlying maladjustments or psychopathology or systemic dysfunctions (de Shazer, 1991) Furthermore, what qualifies as a “solution” is dependent upon the clients’ judgment, i.e., the clients saying that things are “better enough” for therapy to stop. This way of thinking suggests that we need to look at how we have ordered the world in our language and how our language (which comes before us) has ordered our world. That is, rather than looking behind and beneath the language that clients and therapists use, the language they use is all that we have to go on.

Just as the “problem” has different meanings from client to client depending on their different situations, the “problem” has different meanings for the same client because the situation is never the same; even contexts vary over time. Although our approach is sometimes criticized for not getting to and treating “the problem” and thus cannot be long-lasting, our view is that it is unlikely that any approach can cure “the problem” because “the problem” is never the same. Thus, the solution cannot always be same. Frequently our clients tell us exactly that when we ask what worked when they had this problem before; they give us reasons why what “worked before” can not work now. Our follow-up studies have told us that clients have gained the ability to think differently so they can continue to solve problems — whether they are recognized by them as the same ones or new ones. This difference is to us a measure of a long-lasting effect.