Other Articles by Chris Walsh

No Paradigm

This article displays something about Chris Walsh's attitude to psychotherapy.

Therapeutic Alliance

The Therapeutic Alliance with Those Having Both Substance Abuse & Major Mental Illness.

Mindfulness In Individual Cognitive Therapy

Taking advantage of the recent acceptance of mindfulness meditation by cognitive therapists, Chris presented this paper to the 28th National Conference for the Australian Association for Cognitive and Behavioural Therapy in April 2005.

Chris's Mindfulness Site

Carmen's Dream

A case study integrating contellation work with ongoing therapy.


Bert Hellinger

Constellations for Organisations

Print this Page

<< Back

The Therapeutic Alliance with Those "Having It Both Ways"

By Dr. Christopher Walsh

This paper was presented at the first Australian National conference on treating people with both major mental illness and substance use disorders in 1998. At the timeChris Walsh was Consultant Psychiatrist to the Substance Use & Mental Illness Treatment Team (SUMITT) at the Western Health Care Network in Melbourne Australia.

When I started working in psychiatry over thirteen years ago, I often saw therapists giving patients lectures about their drug taking. As the client was hearing how bad cannabis or speed was for them, I noticed one of two reactions. Either their eyes glazed over with their whole demeanour defiantly stating "Oh no, not this again", or they looked shamed and humiliated. Either way there did not appear to be any meaningful discussion. I wondered about the drug from the client's point of view. I knew most of their friends probably used drugs without any obvious ill effects. I knew they often did not have many other meaningful activities in their life apart from drug taking.

Yet issues like these did not seem to be taken into account. The client was blamed for their behaviour, there was not much attempt at empathy and very little sense of therapeutic alliance was established. At that time I didn't have a suitable paradigm of understanding to discuss my vague intuitive insights in a meaningful way. Although much of what I have to say today may be intuitively obvious, I hope this brief introduction to the topic will provide a framework for understanding and communication.

It has been stated that the therapeutic alliance is the single most important tool available to any professional working as a therapist. It is the foundation of all therapies. It has been shown to be important in cognitive therapies and even in non-psychological therapies such as prescribing medication.(1) Indeed the research shows a largely consistent finding that across diagnostic groups and therapeutic approaches, using a number of self report measures, the alliance, especially as assessed by the client between the third and fifth sessions, was one of the most promising within-treatment predictors of positive treatment outcome. (2)

In this paper I will be outlining the concept of therapeutic alliance, as well as beginning to address particular challenges involved in forming a therapeutic alliance with clients with dual diagnoses.

These challenges include:

* Difficulty discussing highly stigmatised issues such as substance abuse and mental disorder.
* The ability of the therapist to empathise with a client undergoing psychotic processes or continuing to take substances despite the harm they cause.
* Fear both in the therapist and the client.
* Different life experiences and worldviews of the therapist and the client.

The term "therapeutic alliance" has its origins in psychoanalytic theory, evolving out of the concept of transference. Hence early definitions were similar to:

"A Non neurotic rational and reasonable rapport which a patient has with his analyst and which enables him to work purposefully in the analytic system."

This has evolved to a broader and more pragmatic definition, even within psychoanalytic circles ie:

"The way two minds, the patient's and the analyst's work together. The alliance has an affective as well as its intellectual side." (3)

I think of the term more broadly still as meaning:

A client and therapist having a sense of working together as a team towards common therapeutic aims directed towards improving the client's quality of life.

It is this definition I plan to use in this talk.

Clearly then, this definition is applicable to a number of different approaches, from different types of psychotherapy, to advocacy, to directive problem solving, and even the narrowest of medical models.

The establishment of a therapeutic alliance has several prerequisites that are largely the therapist's responsibility:

  1. Empathy
  2. Agreed Objectives
  3. Reasonable boundaries, including appropriate limit setting

* Empathy generally precedes the others. Even when dealing with someone who is psychotic or violent a skilful practitioner can often get a much better outcome by attempting to establish rapport as a first step.

* An empathic approach allows the therapist to negotiate objectives that are actually meaningful to the client. When the client does something that appears to sabotage the objectives, an empathic exploration of the situation will usually produce good results. Either the goals will be found to have been inappropriate or the client will gain insight into his/her behaviour.

* Empathy stops the establishment and maintenance of appropriate limits and boundaries degenerating into punitive behaviour on the part of the therapist. This is similar to the idea of "holding" the client as described by Winnicott. (4)


Although much more can be said about negotiating objectives and establishing reasonable boundaries I will now concentrate on empathy.

Rogers described empathy as:

"The state of empathy or being empathic is to perceive the internal frame of reference of another with accuracy and with the emotional components and meanings as if one were the person, but without ever losing the "as if" condition. Thus it means to sense the hurt or pleasure of another as he senses it and to perceive the causes thereof, as he perceives them, but without ever losing the recognition that it is as if I were pleased or hurt or so forth. If this "as if" quality were lost then the state is one of identification." (5)

Or as the other great empathy theoretician, Heinz Kohut rather succinctly puts it empathy is "vicarious introspection". (6)

This appears to go a little further than just putting yourself in the client's shoes. When we do that we often imagine ourselves in the client's current situation with our own life history rather than the client's. This means we may look at the client's situation with an ability to tolerate frustration the client doesn't have. We may also substitute our own emotional and cognitive frameworks for those of the clients. So we put ourselves in the client's shoes without taking into account the state of the client's feet so to speak. Going this one step further often goes a long way to undercutting subtle judgementalism that effectively sabotages empathy and the therapeutic alliance.

It can therefore seem that empathy is the inverse construct of "expressed emotion" as defined by Vaughan and Leff. (7) "Expressed emotion" in families of people suffering schizophrenia is assessed by measuring:

* Criticism, as assessed by content and voice tone
* Hostility, of the client himself, not just his actions
* Emotional overinvolvement.

Clients living with families with a high level of "expressed emotion" have a much higher rate of psychotic episodes. The same has also been found to be true with relapses of mood disorders and eating disorders. One can extrapolate from this that high levels of empathy in the client's home environment are protective against relapse. As therapists we should also try to be aware of signs of high "expressed emotion" with our clients. When we find we are reacting to clients in a hostile, critical or overinvolved way it should be cause for self-reflection. Are we, over stressed or perhaps the client is somehow activating our own unresolved issues? For example are we overidentifying with the client?

Sometimes with dual diagnosis clients even the most empathic therapist feels stuck, unable even to get to first base. This usually means something very important is being missed. Is the client intoxicated or in withdrawal? Is he acutely psychotic? Is there a covert agenda that is not being addressed, such as the client having to appear drug free so he can regain access to his children? Once any of these issues are recognised they can be dealt with in an empathic way that builds the therapeutic alliance. Again it is the avoidance of judgementalism; as well as clinical astuteness that allows these matters to emerge.


Psychotic symptoms may be diagnosed simply by the client telling you, or by the observation of odd bizarre behaviour or speech. In terms of building the therapeutic alliance the next step is to ascertain what the significance of these psychotic symptoms is for the client. If they are hearing voices for example: Are they friendly companions, threatening and menacing or meaningless babble? The meaning of these symptoms is going to affect how the client feels about them as well as how they are going to respond to any offers of help. If the help is both appropriate to and consistent with the client's understanding then a therapeutic alliance is likely to be formed. Often this requires a process whereby the therapist is always trying to better understand the client's point of view whilst balancing this with a professional understanding.

Traditionally psychiatric nurses have been taught not to collude with psychotic client's delusions. Whilst this is a sensible policy, it can be overapplied. It is often more skilful to work with the element of truth found in the client's belief system. An illustrative example is when a highly distressed psychotic patient is admitted to an inpatient unit believing that aliens are trying to kill him. It may be beneficial to all concerned not to argue with him but simply to empathise with how distressed he must feel and to reassure him that he will be safe in the unit. Moreover we can offer the patient some medication to calm him down and help him to sleep. Incidentally this same medication will help cure the psychosis. This example illustrates the point that when a client is paranoid it is preferable to make oneself the client's ally if possible.

Several different situations could cause a similar agitated state, such as: Schizophrenic psychosis, hypomania, an organic delirium, drug intoxication or withdrawal. There are a number of principles that are useful to follow when dealing with an agitated client regardless of the cause.

* Firstly the therapist must be aware of his/her own attitudes and feeling state.
* It is useful to be aware of your feeling state prior to the interaction with the client. It may be useful to try to clear the mind and to calm oneself prior to talking with the client. One way to do this is to empty your mind of thoughts and concentrate on your breathing. This can be done even in the presence of the client. This allows the therapist to be more present to the client without clouding the situation with his or her own issues.

* It is important that the therapist feels reasonably safe. Adequate backup should be available if necessary.

* Secondly awareness of one's own nonverbal communication is invaluable. Unless you develop a habit of self-awareness this can easily be forgotten, as nonverbal communication is usually unconscious. It should however be a priority because when there is a conflict between verbal and nonverbal communication, the nonverbal is what is usually believed.(8) Awareness and control of nonverbal communication is largely a physical skill. As such it is best learnt by physical practice, such as in an experiential workshop. Suffice to say that the main areas that are valuable to monitor in oneself are:

  1. Voice Tone
  2. Body Position
  3. Body Movement
  4. Eye contact

I have deliberately omitted facial expression as attempts to control facial expression often convey the impression of insincerity.


* The verbal component of communication is more representative of our conscious attitudes. It is therefore amenable to a conscious examination of our own attitudes. The following attitudes are more likely to enhance a therapeutic alliance:

* Interest in the client and their point of view thus predisposing the therapist to real listening.

* To present information in a non-provocative and non-judgemental manner.

* To present information at a level which the client can comprehend. To do this you need to take into account:

  • The client's vocabulary and educational level
  • The client's life experiences which may suggest appropriate metaphors
  • The client's mental state, as those who are significantly psychotic, depressed, anxious, substance affected or in severe withdrawal are unlikely to take in or remember much information.

* To show respect for the client and their interpretation of events.

* To avoid early interpretations

* To demonstrate that you have control of the situation without adopting an overly authoritarian stance.

* Do not make promises you can't keep.

If the client regards the therapist suspiciously or has incorporated them into his delusions then the therapist should adapt a relaxed style that is not overly assertive. Another therapist may need to be recruited temporarily

It is important that therapists do not put themselves in any unnecessary danger. When the therapist feels afraid, is an indication of potential danger that must be respected. It is often accurate and can also precipitate aggression in a paranoid and fearful client. Therefore the establishment of a therapeutic alliance complements rather than contradicts the principles of management of aggression. In fact patients who had a poorer therapeutic alliance at the time of admission to a psychiatric ward were significantly more likely to display violent behaviour during hospitalization. (9)

Fear can also be an important component of therapist attitudes toward the client, which can interfere with empathy. Fear of patient aggression is dealt with by paying due attention to therapist safety. Other fears are often less conscious and therefore more insidious. Fear of insanity can cause therapist to place artificial boundaries between clients' experiences and their own. This process makes some of our most powerful empathic tools unavailable. For example we all use highly illogical and symbolic primary process thinking. This clearly occurs in dreams and meditative states. It is also the stuff of which much great poetry and art is made. Sometimes when we are tired and stressed we can become paranoid. Many psychotic experiences clearly are not so far removed from normal experience. We can use these normal experiences to bring us a little closer to an understanding of these people. Instead the fear of the insane causes many therapists to not only miss this opportunity but also cuts them off from simpler empathic experiences. For example it is important to remember that our clients, as human beings, share a great deal of normal human experience with us. They have sexual urges, they enjoy music and they feel frustration and pain. It is relatively easy to empathise with these things.


It is useful for therapists to examine their own attitudes for elements of judgementalism regarding people who take drugs or participate in associated activities such as prostitution or theft. It is quite reasonable for the therapist to have their own judgements about these activities in their own right. However, empathy is about understanding these activities from the client's point of view, without judging them as a good or bad person.

This non-judgmental attitude allows clients to then explore their own attitudes without feeling they have to defend themselves against the therapist. For example, when a schizophrenic client has his third amphetamine induced psychosis it can be very tempting to give him a lecture like: "How many times have I told you John that you shouldn't take speed. It just makes you sick." A more useful alternative is to encourage the client to explore the whole situation. One way of doing this has been outlined by Miller (10) with his technique of motivational interviewing. This would include: Exploring what led to the drug taking, what alternative ways of coping may be available and what were the consequences of drug taking (both good and bad).

Finally the client's fear can also interfere with the therapeutic alliance.

* The stigma felt generally in the community toward mental illness leads clients to be fearful of labels such as schizophrenia or even psychiatric patient. Therefore these labels should be given judiciously and tactfully. Opportunity should be made for the problem of stigma to be discussed. However, it is important to bear in mind that many clients find the label quite useful as it explains the previously inexplicable and can give them a sense of control.

* The fear of being judged can be allayed by the therapist's attitude.

* Fear of the stereotypical malevolent therapist.
Many people fear being "psychologically raped" by psychotherapists. This fear is often justified when unskilled therapists impose their own neurotic projections on the client or simply make poorly timed interpretations of clients' words or actions. Skilful empathy will minimise these tendencies as well as allaying the client's fear.

Basically being empathic and sensitive to the client's feelings allows a space within which these feelings can be brought out in the open and dealt with.


In this paper we have had a theoretical introduction to the concept of therapeutic alliance. We have demonstrated its practical value. The therapist's contributions to the therapeutic alliance have been outlined. These are empathy; negotiating therapeutic objectives with the client and establishing appropriate boundaries, including limit setting. Empathy was discussed in detail, as it is the most fundamental of these skills. Practical issues in regard to empathy such as verbal and nonverbal communication were presented. Special points in regard to establishing the therapeutic alliance with dual diagnosis clients were introduced. These included:

  • How the therapist's own attitudes and life experiences in comparison with the client's can affect the empathic process
  • The importance of a non-judgemental, empathic approach in creating a safe space in which stigmatised issues such as mental illness and substance use can be discussed.
  • Dealing with agitated clients who may be psychotic, intoxicated or in withdrawal
  • Dealing with fear in both the therapist and the client.



As I wrote this paper I realised that thirteen years ago when I first started in psychiatry I was arrogant and ignorant enough to believe I knew all about empathy and the therapeutic alliance. This sad assumption was fortified by the fact that the people I worked with that did know more, did not have the language necessary to hint at that deeper knowledge. Since then I have discovered as I learn more, that the possible degree of subtlety and the depth of understanding seems to be limitless.

Now I also realise that much of the real understanding of these issues occurs at a nonverbal level. The skills involved are practical skills like learning to sing or dance or play tennis or even driving a car. Like any of these practical skills it is ludicrous to think you can learn them just through words. The understanding has to literally get into our muscles and our nervous systems. Just as we know through practice when to swerve our car in advance to avoid an accident without even thinking about it, we can develop a similar ability to read the direction of the interaction with the client and take appropriate action.

The fact that this sort of understanding exists can be debated forever, but can only be proven by allowing oneself to experience it. As experiential training seems to have gone out of fashion in the 80's and 90's we have lost an arena within which we can share our experiences and can acknowledge our deepening insights in these matters. I believe that as a result of this the very basis of therapeutic effectiveness ie the therapeutic alliance has been taken for granted or even worse, forgotten all together.

Today I have attempted for a short moment to bring the therapeutic alliance back into the foreground, I have attempted to highlight its importance and hopefully I have introduced a clear and practical way of thinking about it. I would like to make it clear that I believe this topic can usefully be explored in much more depth, both intellectually and experientially. I will be attempting to do this in my position as consulting psychiatrist for the newly formed Dual Diagnosis Team. In fact if the Dual Diagnosis Team can contribute to a culture in both Drug and Alcohol Services and Psychiatric Services whereby the therapeutic alliance achieves a place of prominence then I for one will be extremely pleased because I will know we will have then achieved something very worthwhile.


Unfortunately, I believe the above aim represented too dramatic a cultural shift for these services. I resigned from all public drug & alcohol and psychiatric services early in the year 2000. I stayed away from these services until August 2004, when I cautiously accepted a very part time position at Turning Point Drug and Alcohol Service.

My hope now is that work such as constellation work might one day be able to create the shift necessary in the system so that staff, clients and their families can all find the system more supportive. Perhaps then love, energy and respect will finally start flowing here too.

Interestingly it was partly my feeling of having failed in my position as Consultant Psychiatrist to the
Substance Use & Mental Illness Treatment Team (SUMITT) that left me open to discover Hellinger's constellation work.


(1) Krupnick JL, Sotsky SM, Simmens S, Moyer J, Elkins I, Watkins J, Pilkonis PA
The role of therapeutic alliance in psychotherapy and pharmacotherapy outcome: Treatment in the National Institute of Depression Collaborative Research Program.
J Consult Clin Psychol 1996, 64:532-539
(2) Alexander L.B. & Coffey D.S.
Understanding the therapeutic relationship
Current opinion in psychiatry 1997, 10: 233-238

(3) Stewart, in Kaplan & Sadock
Comprehensive Textbook of Psychiatry/IV 1992 pp1342

(4) Davis M & Wallbridge (1987)
Boundary and space. An introduction to the work of D.W. Winnicott pp 98
Brunner/Mazel Publishers
New York

(5) Rogers Carl (1980)
A way of being
Houghton Mifflin
Boston, Massachusetts

(6) Kohut H. (1984)
How does analysis cure? pp82
Ed A Goldberg
University of Chicago Press

(7) Vaughan C and Leff JP (1976)
The influence of the family and social factors on the course of schizophrenic illness.
Brit J Psychiatry 129, 125

(8) Argyle M. (1983)
The psychology of interpersonal behaviour.

(9) The utility of initial therapeutic alliance in evaluating psychiatric patients' risk of violence
Beauford JE, Mcneil DE, Binder RL
Am J Psychiatry 154/9 (1272-1276) 1997

(10) Miller W.R. & Rollnick S 1991
Motivational interviewing, Preparing people to change addictive behaviour
New York