Therapeutic Alliance, Focus, and Formulation: Thinking Beyond the Traditional Therapy Orientations

Therapeutic Alliance, Focus, and Formulation: Thinking Beyond the Traditional Therapy Orientations

by Robert-Jay Green
The main elements of successful therapy include a positive therapeutic alliance, a clear focus, a coherent problem formulation, and improvised techniques—not a particular theoretical orientation.


Get Endless Inspiration and
Insight from Master Therapists,
Members-Only Content & More


I recently attended the 40th annual retreat conference sponsored by the Association of Family Therapists of Northern California (AFTNC), and this year's renowned speaker, Monica McGoldrick, showed many videotapes of her own work. Throughout the two-day conference, I was continually impressed by her ability to relate so very closely and easily with such a racially and culturally diverse group of clients. I also kept noticing how much of the connecting seemed to take place through nonverbal behaviors and tone of voice (conveying warmth, comfort, directness, engagement, confidence, being collaborative, and indicating genuine interest in the details of people's life histories). As president of AFTNC, I also received lots of spontaneous comments afterwards about Monica McGoldrick being such a "lovely person," a "great human being," so "down-to-earth," "easy to relate to," "willing to show her mistakes," and so on. At the same time, I heard many therapists in the audience make comments about how they work so very differently from McGoldrick in terms of not using extended family genograms as a main vehicle for connecting nor using traditional Bowen coaching methods with one family member. I also heard a few complaints that McGoldrick's presentation did not allow for easy note-taking in that it was not structured with specific teaching points or generalizable techniques one could use with almost every case.

The Therapeutic Alliance

The combination of these comments and my own reactions reminded me of that old Marshall McLuhan phrase, "the medium is the message." That is, the main message for me in McGoldrick's way of working was not about genograms, or Bowen theory, or coaching methods, or concepts and techniques that could be written down. The meta-message was that she modeled a way of relating to people of diverse backgrounds that is very rare among therapists of any theoretical orientation but lies at the heart of therapy. This way of close (but still professional) relating is extremely difficult to describe and teach in a didactic format. Partly, this is because the emotional aspects of the therapist/client relationship, which are so much anchored in nonverbal behavior, are hard to communicate in words. There are no simple directives one can give therapists for generating warmth toward or comfort with clients in general. Simply telling therapists to reflect feelings or give compliments can backfire if the nonverbal aspects of these communications are not congruent with the content of what is being said.

Telling therapists to "be warm" or "be genuine" simply puts them in a "be spontaneous" paradox because true warmth and genuineness must, by definition, be guileless and uncontrived.
Telling therapists to "be warm" or "be genuine" simply puts them in a "be spontaneous" paradox because true warmth and genuineness must, by definition, be guileless and uncontrived.

In fact, few contemporary publications in the field of couple and family therapy have focused in any depth on the process of alliance building, the most notable exceptions being Bill Pinsof (see Integrative Problem-Centered Therapy, 1995) and Sue Johnson (see The Practice of Emotionally-Focused Couple Therapy, 2nd Edition, 2004). Ironically, however, most clinical supervisors believe that therapist relationship skills, rather than techniques or theoretical orientations, are the more important aspects of effective treatment (Blow & Sprenkle, 2001). After 30-plus years in the field, I also have come to believe that each of the most popular family therapy orientations is too limited and too focused on specific change-oriented interventions, rather than on the vicissitudes of the therapeutic alliance. I don't think treatments succeed or fail based primarily on the particular theoretical orientation of the therapist. Rather, I believe that treatments succeed mainly on the therapist's ability to develop and maintain an emotionally positive therapeutic alliance with all members of the system in treatment.

A positive therapeutic alliance requires:

  1. giving sufficient emotional validation and support to the individual or members of the couple/family,
  2. successfully managing negative emotions within the couple/family so that members are not hurting each other in sessions, and
  3. regulating the therapist's negative emotional reactions to couple/family members (i.e., managing "negative countertransference").
Too many of the current approaches to family therapy either take for granted or neglect to adequately address the importance of the emotional bond between therapist and clients.
No array of clever, change-oriented techniques is effective in the absence of a positive emotional connection between therapist and client.
No array of clever, change-oriented techniques is effective in the absence of a positive emotional connection between therapist and client. First and foremost, clients need to feel that the therapist is caring, collaborative, trustworthy, fair, reliable, knowledgeable, and that she/he understands and appreciates their feelings. This is true regardless of the therapist's theoretical orientation. Within each theoretical orientation, there are therapists who are very skilled at forming close and collaborative relationships with clients and those whose relationship skills need improvement. Yet the majority of current family therapy theories and training programs deal with these common therapeutic factors in a cursory fashion if at all, preferring instead to focus on abstract theories and very specific change-oriented techniques.

When therapy breaks down, it almost always is because of some emotional rupture and failure to repair (or failure to establish in the first place) a positive therapeutic bond with family members. In fact, therapists' and clients' negative emotional reactions to each other, even when not directly expressed, are among the major factors predicting early termination or poor outcomes. These negative emotions tend to be expressed indirectly in terms of behaviors like the therapist not returning phone calls promptly; not inquiring how the client is reacting to interventions even when the therapist observes the client's disapproval; the absence of spontaneous mutual warmth, humor, and responsiveness between therapist and clients; therapist passivity in sessions; or the therapist's unequal responsiveness to family members in sessions.

Although it would be comforting to assume that all therapists are inherently capable of forming positively toned therapeutic relationships with all clients, there is an enormous range of skill among therapists in this respect. Also, certain kinds of clients (e.g., those traditionally and pejoratively labeled "borderline" or "narcissistic") can challenge any therapist's relationship skills. The very best therapists tend to be those who can easily establish and maintain positive therapeutic alliances with the widest range of clients, both in terms of clients' cultural diversity and in terms of managing negative emotionality.

Training to Become a Therapist

Given the above, 

a major focus of training should be on how a particular trainee can use her/his emotions in forming therapeutic alliances and in preventing those alliances from breaking down.
a major focus of training should be on how a particular trainee can use her/his emotions in forming therapeutic alliances and in preventing those alliances from breaking down. This training focus should include the special issues in intercultural therapist/client matches (race, gender, social class, sexual orientation, age, religion, etc.), which are more likely to dissolve for lack of a strong positive emotional connection, or to be negatively tinged and lead to experiences of oppression for the client. In my view, a major focus of all training programs should be the development of therapist alliance-building skills, especially emphasizing cases in which the client displays a lot of negative emotionality and cases where the therapist and client are from different sociocultural groups.

Therapeutic relationship skills are best acquired through close observation of the therapist's behavior in role plays and in sessions (i.e., via live, videotape-, or audiotape-based case supervision). Such skills are not as easily learned—or maybe impossible to learn—through "delayed verbal report" supervision because the latter's effectiveness is constrained by what the therapist was aware of during the session, remembers after the fact, and can report in words to the supervisor. Verbal report supervision always loses emotionally relevant information because a lot of what transpires between therapist and clients is automatic and not necessarily within the therapist's awareness, especially when the therapeutic alliance is not functioning well. Paradoxically, the very areas where supervisees may need the most help are areas about which they are unaware and cannot articulate the relevant emotional information to their supervisors.

The analogy I like to use here is that learning to do therapy is like learning to swim. You could describe to your swimming coach, after the fact and in words, how you moved your arms and legs while in the pool. You could even demonstrate your movements while sitting in your chair or lying on a table. But to actually learn how to swim, it is far better to have your coach by your side, preferably in the water with you, watching you perform the new strokes, and giving you immediate feedback on your movements based on the coach's immediate observations rather than on a delayed report. The same is true of therapy. Describing a difficult session in words for one's supervisor will not convey the same information as having one's supervisor directly observe the "drowning" noises one makes in an actual failed session! Supervision based on delayed verbal reports limits the supervisor's ability to accurately visualize and give input about the therapist/client alliance.

There simply is no substitute for live supervision (or video- or audiotape-based supervision) in learning to become a therapist. In my opinion, the fact that many mental health professionals still become therapists never having had their work observed is a major shortcoming in our training programs. Just as one would not want to be under the care of a neurosurgeon whose work had never been observed by other neurosurgeons, we should require extensive observation-based supervision for all therapists in training. Only then can we be more certain that our trainees have learned to build and maintain effective therapeutic alliances with a wide array of clients.

Maintaining a Focus

Although I am emphasizing its centrality here, I think of a positive working alliance as a necessary but not sufficient condition for therapy to be effective. That is, I think of the relationship as a partial intervention in itself (a "corrective emotional experience") as well as being the necessary groundwork for other, more specifically change-oriented interventions to have an impact. The other main ingredient in effective therapy I would call "focus": successful therapy requires establishing relatively clear collaborative goals with clients and using interventions that are relevant to those therapeutic goals throughout. For example, in consultations for "stuck" cases, I frequently have found that a clear sense of direction was never established at the outset of a treatment; or, once having been established, the therapy conversations meandered or avoided dealing with the main presenting problems.

By using the word "goals" I don't mean to imply some superficial, limited purpose, but rather that the client and therapist should share the same vision of what a desirable outcome might be in terms of changed behaviors, affects, or cognitions, as well as the path for getting there. As the old aphorism holds: "If you don't know where you are going, you are not likely to get there." This also reminds me of something that a hapless character in a Lily Tomlin play once said: "I always wanted to be someone, but now I realize I should have been more specific."

In therapy, you can't get there with the client if you don't have a strong positive emotional relationship and if you haven't established together where you are going. I think that effective therapist/client systems have reasonably explicit and clear (albeit evolving) goals, and the conversations in the sessions are always brought back to those main foci. This point may seem obvious, but a lot of ineffective therapy sessions end up veering off into conversations relevant to the therapist's theoretical orientation or to the client's momentary preoccupations rather than to the clients' main presenting problems. Therapists need to keep their eye on the ball—the shared goals—and help their clients do likewise. This requires a client-specific formulation of the problem and goals, but it does not require that the therapist use only one particular theory.

I often feel that the specific interventions chosen by therapists (e.g., unique outcome questions, sandtrays, enactments, genograms, dream interpretations) are much less important than whether the interventions are explicitly related to the client's presenting problems and the established goals. That is, the client and therapist should readily understand what the connection is between the agreed-upon goals and what is happening in the sessions. Too often, clients seem not to understand that connection, and they are reluctant to ask for explanations, even when they think the therapy is not really addressing their main concerns and is going nowhere.

To summarize this viewpoint: therapy is effective to the extent that it is characterized by a strong positive emotional alliance, relatively clear goals, and conversations and tasks whose relevance to the goals is obvious to all parties. By contrast, therapy that does not consistently maintain a positive emotional alliance, has extremely vague goals, and consistently includes conversations and tasks that are tangential to the client's experience of the problem(s) is unlikely to be effective. I believe these statements hold true regardless of the therapist's preferred theoretical orientation.

Problem Formulations Rather than Theoretical Orientations

I question whether it is even necessary to have a specific theoretical orientation. Obviously, one cannot not have an idea about what is causing the client's problem, and one cannot not use techniques. Virtually anything a therapist says or does can be considered an aspect of technique that is based on some theoretical concept. But how much does a therapist need to adhere to a specific theoretical orientation in order for therapy to be successful? Or to put it differently, how eclectic can one be and still be effective?

I believe that successful therapy requires that the therapist have a workable problem formulation—a hypothesis about what is causing the problem in a given case—and a general treatment plan—an evolving path toward the goals in that case—in order for the therapist to select meaningful things to say and do in the sessions. The problem formulation and treatment plan should be explicit (able to be articulated verbally or in writing) and tailored to the client's and problem's uniqueness. However, I don't think a therapist's formulations and plans have to derive from a single theoretical orientation in order for that therapist to be effective. The formulation for a given problem will typically involve the relationships between two or more levels of functioning—biological, psychological, interpersonal, and/or sociocultural factors.

Without digressing too far, I'd like to suggest that given the enormous variety of client belief systems and presenting problems in the world,
each theoretical orientation may be best suited for certain kinds of clients and presenting problems and not as applicable to others.
each theoretical orientation may be best suited for certain kinds of clients and presenting problems and not as applicable to others. It also seems that less experienced therapists may have a greater need to adhere to a single orientation in order not to lose their focus over the course of treatment, but I think they do so at the risk of reducing their flexibility to respond in the most helpful way.

The Integrative Therapist and Emergent Design

The task of the "eclectic" or "integrative" therapist is to co-develop with the client a set of achievable goals, a coherent problem formulation (an explanation for why the problem exists or what is causing it), and a treatment plan tailored to the client's specific problems and situation, all the while maintaining a positive therapist/client alliance. The process of therapist and client co-creating these therapeutic elements is what I once called an "emergent design" (Green & Herget, 1989b). No two therapists would be able to develop the same design with a client. The emergent design process can best be conceived as a mutually catalytic reaction between a particular therapist and a particular client evolving together at a particular point in time. In Darwinian fashion, if all goes well, the ideas and behavior patterns that "survive" over the course of therapy will have therapeutic utility whereas other ideas and behavior patterns will become "extinct." It is for this reason that some cases that start out, for example, in a seemingly psychodynamic mode may become increasingly behavioral over time, or some individual therapy cases may surprisingly turn into couple therapy cases by the end. If therapist and client are sufficiently flexible, this evolution happens more quickly and progress is likely to be quicker. If therapists adhere too closely to traditional theoretical orientations after the beginning phase of therapy, the essential evolutionary nature of the treatment relationship is more likely to be impeded and progress stalled.

For the therapist's part, after establishing a mutually acceptable formulation of the problem, the remaining therapeutic task involves improvising a treatment plan and techniques based on that formulation. The therapist as participant-observer in this process must be both emotionally engaged and purposeful in ensuring that a consistent focus is maintained during the sessions. Effective therapy, in this view, does not require that the therapist adhere to a single theoretical orientation with all cases or even across all problem areas within a given case. It does, however, require reasonable consistency and narrative coherence regarding each specific problem focus and formulation established with a given client.

This kind of meta-theoretical approach to therapy is not entirely new and is partially an outgrowth of comparative psychotherapy research and the movement toward psychotherapy integration. It constitutes a new kind of "theory of therapy," emphasizing the "common factors" that make the implementation of any therapy either effective or ineffective regardless of the "brand" of therapy practiced by a therapist (e.g., psychodynamic, cognitive-behavioral, or systems-interpersonal). Versions of this meta-theoretical approach were proposed in the 1970s by individual psychotherapy researchers such as Jerome Frank (1973), Edward Bordin (1979), and Hans Strupp (Strupp & Hadley, 1979); and by family therapy researchers in the 1980s such as Leslie Greenberg and William Pinsof (Greenberg & Pinsof, 1986). Mary Herget and I also used it to some extent at the Redwood Center in our small-sample research on Milan teams in the mid-1980s (Green, & Herget, 1989a, 1989b, 1991). Most recently, the meta-theoretical approach is best represented in books such as Psychotherapy Relationships That Work: Therapist contributions and responsiveness to patients (Norcross, 2002) and The Heart and Soul of Change: What works in therapy (Hubble, Duncan, & Miller, 1999).

Having participated in many of our field's fads and fancies since 1970, this meta-theoretical framework represents my current personal list of the "eternal verities of therapy"—the essential components of effective treatment. Although I will never use Bowenian techniques to the extent that Monica McGoldrick does, the AFTNC annual conference reminded me once again of the centrality of therapist relationship skills, which cannot be acquired through books and lectures. The indelible images of McGoldrick relating so closely, comfortably, and confidently in sessions—even with computer in hand and genograms as the focus—was the take-home message for me, and I hope for students learning the practice of therapy, as well.


Blow, A.J., & Sprenkle, D. (2001) Common factors across theories of marriage and family therapy: A modified Delphi study. Journal of Marital & Family Therapy, 27, 385-

Bordin, E.S. (1979). The generalizability of the psychoanalytic concept of the working alliance. Psychotherapy: Theory, Research, & Practice, 16, 252-260.

Frank, J.S. (1973). Persuasion and healing (2nd ed.). Baltimore: Johns Hopkins University Press.

Greenberg, L.S., & Pinsof, W.M. (Eds.). (1986). The psychotherapeutic process: A research handbook. New York: Guilford Press.

Green, R.-J., & Herget, M. (1989a). Outcomes of systemic/strategic team consultation: I. Overview and one-month results. Family Process, 28, 37-58.

Green, R.-J., & Herget, M. (1989b). Outcomes of systemic/strategic team consultation: II. Three-year followup and a theory of "emergent design." Family Process, 28, 419-437.

Green, R.-J., & Herget, M. (1991). Outcomes of systemic/strategic team consultation: III. The importance of therapist warmth and active structuring. Family Process, 30, 321-336.

Hubble, M., Duncan, B. & Miller, S. (Eds.) (1999). The heart and soul of change: What works in therapy. Washington, DC: American Psychological Association.

Johnson, S. (2004). The practice of emotionally focused couple therapy: Creating connection (2nd ed). New York: Brunner-Routledge (a division of Taylor & Francis Publishers).

Norcross, J. C. (Ed). (2002). Psychotherapy relationships that work: Therapist contributions and responsiveness to patients.

Pinsof, W.M. (1995). Integrative problem-centered therapy. New York: Basic Books.

Strupp, H.H., & Hadley, S.W. (1979). Specific vs. non-specific factors in psychotherapy: A controlled study of outcome. Archives of General Psychiatry, 36, 1125-1136.

Note: An earlier version of this article appeared in the January 2004 issue of the Association of Family Therapists of Northern California Newsletter. For further information about AFTNC, you may visit its website at

Published on with written permission from the author.

Copyright © 2004 Robert-Jay Green. All rights reserved.
Order CE Test
$15.00 or 1.00 CE Point

CE points are a great way to save if you need multiple CEUs. Get up to 45% discount when you buy packages of 10, 20 or 40 points. Your CE points will be redeemed automatically at checkout. Get CE packages here.

Earn 1.00 Credits
Buy Now

*Not approved for CE by Association of Social Work Boards (ASWB)

CE Test
Robert-Jay Green Robert-Jay Green is Professor and Associate Director of the Clinical Psychology Ph.D. Program, California School of Professional Psychology at Alliant International University, San Francisco campus. He has a private practice in San Francisco specializing in couples therapy, and is currently President of the Association of Family Therapists of Northern California (AFTNC). He is author of more than sixty publications about couples/family interaction and therapy. You may contact him at 415-749-0100 or by email: His personal website is at:

CE credits: 1

Learning Objectives:

  • Describe factors that contribute to a positive therapeutic alliance
  • Explain the significance of direct observation of training therapists

Articles are not approved by Association of Social Work Boards (ASWB) for CE. See complete list of CE approvals here