By John Rowan

Recently I have become involved with a couple of problems with therapists in relation to boundaries, and it seemed worthwhile to share some of my discoveries with others. All therapists are agreed that boundaries are important in therapy. It is the job of the therapist to hold the ring, to maintain the frame, to secure the temenos (the sacred space), to keep the vas (the alchemical vessel) closed.  We need to keep the energy in, on the one hand, and to keep distractions out, on the other. It is within the boundary that the relationship takes place. As Petruska Clarkson (2003) has told us so brilliantly, there are at least five relationships all going on at the same time in therapy. The therapeutic alliance is where we make a contract with the client, make sure that time, place and money are taken care of, that the client is motivated and responsible – all the practical things which have to be secured if the process is to be protected. The transferential relationship is where we pay attention to the unconscious part of the transaction, where both the client and the therapist may be subject to influences from the past which provoke reactions not under conscious control. The reparative relationship is where the therapist has to be willing to be what the client’s fantasy needs.  The therapist may turn into a mother, a father, a sibling, a teacher or other authority figure, and the client may regress into a much earlier age state. The real relationship is where the therapist is authentic, relating to the client in an I-You way, not in an I-It fashion. This is particularly important in humanistic work, where authenticity is a prime value, and must be actually available in the session. The transpersonal relationship is where the therapist is paying attention to the spiritual dimension of the work, and where the client’s spiritual aspirations and spiritual problems are taken seriously for what they are, and not reduced to something else. More details about all of these are of course to be found in the book mentioned: here we are simply admitting that they exist and can only be ignored at our peril.  In all of them the question of boundaries is important.

We have become more aware since about 1985 of the boundaries of sex and anger.  It has become crystal clear that sexual relations between therapist and client are always harmful, because they are always based on a difference in power, such that the therapist is the abuser even when they feel like the seduced. There is a good discussion of this issue in Russell (1993). This does not mean that the therapist has to be cold – indeed, Rogers has told us that nonpossessive warmth is an essential quality for any therapist.  But it does mean taking care of the difference between acceptance and approval.  Acceptance is fine – taking clients as they are, not as they should be. But approval is a kind of warmth that goes too far. We do not tell clients that they are OK, that they are good-looking, that they are well-dressed, that they are intelligent – none of these things or others of the same kind. That is reassurance, that is our judgement, that is manifesting our superiority. And we take care not to let our private opinions about the client become a part of the therapy, unless there is clearly some therapeutic gain to be achieved by doing so. There is a good discussion of this point in Chapter Two of Windy Dryden’s book (1985).

This leads to the whole question of dual relationships. It used to be held that all dual relationships were harmful. However, it has been recognised in recent years that there is an important distinction between boundary crossing and boundary violation, particularly well explained in the book edited by Lazarus and Zur (2002). Boundary crossing can be quite innocent and even valuable (such as giving the client a lift, for example), while boundary violation is to be condemned. The difference is that in boundary violation there is a suspect motivation on the part of the therapist, who is often exploiting the client in one way or another. There is a good discussion of this also in Gabriel (2005). The ‘slippery slope’ argument (that any crossing will inevitably lead to worse violation) is now not generally respected.

Probably the most important boundary issue is confidentiality. This is something which few people outside the world of therapy understand. Social workers, nurses, teachers – mostly don’t understand it. It means that the words of the client are sacred. They are not to be repeated or quoted to anyone else. The obvious exception to this is the supervisor, but in the best practice even the supervisor does not know the client’s name or address.  In supervision it is best to give the client a pseudonym or a letter or a number. So the therapist does not gossip about the client, the supervisor does not gossip about either the client or the supervisee. I use the word gossip to describe any conversation about someone when that person is not present.

Still further, the therapist attempts to ensure that the client does not talk about what happens in their therapy to a third party. This is particularly important when that third party is involved in the therapy, often being mentioned or referred to – partners and parents are the obvious examples. It is all too easy for someone to refer to their therapy in terms which bring forth a defensive reaction from the hearer. Also there are times in therapy when strong negative feelings are expressed towards someone, and if the client then meets that person and tells, the hearer may take the message too strongly and finally, as if it were the last word on the subject. If the client, later in the therapy, sees that some of those negative feelings were exaggerated, it is hard, and may be impossible, to undo the damage. Even the most definite conclusions may turn out to be not the final conclusions after all. So confidentiality is important in this way too.

Finally there is the issue of competence. It is wrong to take on a client if there are clear indications that such a client is beyond the expertise of the therapist.  The most obvious example is that of a therapist who takes on a child in therapy when they have had no training in child psychotherapy. Another example would be the case of someone who wanted to work through material pertaining to their birth, if the therapist had no training in or experience of this kind of work. Another would be to take on a client experiencing a spiritual emergency, when the therapist had not worked through their own material in this area.  There are plenty of examples of where an overoptimistic therapist has taken on the wrong client, and both of them have ultimately had a bad experience. Given all this, there are still great differences between different schools of therapy about the details. The psychoanalytic school is the most strict on the question.

A basic text by the highly respected psychoanalyst and teacher Robert Langs (1982) says that the following things are part of the essential framework for psychoanalytic psychotherapy, and that ignoring them will have a very negative effect on the whole course of the therapy itself:

1. The therapist’s chair should be quite different from the patient’s chair (Langs 1982:364).

2. There should be a table between the therapist’s chair and the patient’s chair (Ibid: 364).  3. There should be a table between the therapist’s chair and the couch (Ibid:364).

4. There should be a desk present to provide a sense of professionalism, but there should be nothing on the desk which is in any way self-revealing (Ibid: 364).

5. The therapist should not respond directly to requests to change the times of sessions, and should either remain silent, or suggest that the patient continues to say whatever comes to mind (Ibid: 421).

6. The therapist’s fee should be maintained. If it is suggested that it be reduced, sufficient exploration will usually point to the need to keep the fee as stated (Ibid: 421).

7. Any reduction in fees, or the proposal of a fee well below the therapist’s usual range, is highly seductive and infantilizing (Ibid: 435).

8. The patient should pay the fee by cheque, which should be endorsed by the therapist so that it is clear that the money has gone into the therapist’s professional bank account (Ibid: 439).

9. If an error is made with a cheque, it should be put out on the desk or on a table, so the patient can see it upon entering the next session (Ibid: 439).

10. The filling in of a form for the therapy to be paid for by an insurance company precludes the patient’s experience of separation anxieties in his relationship with the therapist, while reinforcing his own perverted tendencies. It is a disturbance in the therapist’s therapeutic commitment to the patient and a means of driving him away (Ibid: 486).

11. There will be no physical contact between therapist and patient other than the handshake with which the patient is greeted at the time of the initial consultation, and perhaps a handshake at the time of an extended vacation or at termination (Ibid: 429).

12. If the patient becomes bored and wary of the therapeutic relationship because of the therapist’s passivity, the therapist has no choice but to continue it (Ibid: 431).

13. The therapist has three options: silence, interpretations and the management of the ground rules; anything else is ruled out (Ibid: 466).

14. If the patient finds out directly that the therapist is in supervision, the therapist must give up supervision, even if this means that the therapy has to come to an end, because total confidentiality has been violated (Ibid: 478).

15. The prerogative of cancelling hours and taking vacations belongs to the therapist, and is one of the necessary inequities (hurtful aspects) inherent to an optimal framework (Ibid: 438).

16. The therapist should be careful to end the session within 30 to 60 seconds of the appointed time (Ibid: 449).

17. If the therapist cannot make a particular hour, there should be no charge for the session, but there should be no offer to make the session up (Ibid: 451).

These would be seen by many other schools as being over-strict, but once they are left behind there are many differences of opinion as to what to have instead by way of rules.  On the question of whether the therapist may touch the client, for example, there is now an excellent book (Smith et al 1998) which quotes a good deal of research on the subject. For example, Lawry says: “However, it does appear that there are times when touch can provide unique therapeutic communications, such as ‘You are touchable’, ‘Touch can feel safe and nurturing’ or ‘You can say no or yes to touch.’ These communications are particularly important to a survivor of sexual trauma” (Lawry 1998: 205).  It is clear from this work that physical contact with a client cannot be ruled out as always wrong. There are many considerations here, and there is no rule that should always be applied. In Chapter 2 of her classic book The Therapeutic Relationship (2003), Petruska Clarkson warns against a trend towards ‘defensive psychotherapy’ where therapists may be overcautious to avoid any physical contact even where it is indicated. Holding and hugging are contra-indicated when the client has strong and unfulfilled dependency needs (Imes 1998).  In other cases it depends a great deal on context, and there is no handy rule of thumb to go by.

Something not mentioned so far is a good example of a disputed area: how much does the therapist reveal of her or his inner processes, particularly as related to what is taking place in the here-and-now of the consulting room, but also of her or his everyday life as well? The engagement of the therapist presumes that the self of the therapist is kept hidden so as not to interfere with what the client wants to say, or with what the client is experiencing. Of the many ways in which the therapist more actively uses the self, it is obvious that empathic interventions only become empathic when they are openly shared. It would probably be agreed by every therapeutic approach that other aspects of the therapist’s thought processes – identification, intuition, imagination, will also be shared when it seems appropriate, and made relevant to the client without necessarily revealing anything about the therapist: ‘I can imagine how you might be feeling . . . if I were in your position, I might consider  . . . etc.’ But there is much less agreement across orientations as to whether the therapist might say: ‘Just at this moment I’m feeling (such and such) in myself’;  or  ‘I remember when I went through a similar experience’; etc.

First impressions also suggest that if we are going to look for any disclosure by the therapist that is at all personal, it is to the humanistic and existential therapies that we might expect to find support and examples.  The more cognitive and behavioural approaches seem to find little point in self-disclosing or temptation to do so, except perhaps in giving feedback about how the client seems to the therapist – hardly self-disclosure. Intersubjectivists (who are now found in many of the traditionally different schools) share with their clients their own conflicts of how much of their feelings to reveal.  Hycner, for example, mentions an instance when:

sensing her greater openness I took a major risk and I told her how I had often felt like a phantom to her.  The potential for meeting outweighed the risks…  In fact, I believed that it was essential for her to deal with this issue, since most likely she was dealing with other people in a similar manner.  Though my comment hurt her feelings initially, it became a turning point in the therapy.  (1993: 72)

It is also in the humanistic and existential tradition that we find discussions of different types and levels of self-disclosure. One of the most basic aspects of this is drawn out by an exchange of views between Carl Rogers and Rollo May. Rogers is well known for his insistence on genuineness and congruence in the therapist, but in practice (for example, as shown in the book edited by Farber et al. 1996) he appears never to have expressed any negative feelings. May criticizes this:

Rogers has, of course, been in the forefront of those insisting on respect for the patient. But is not respect best and most profoundly shown by openly admitting anger, hostility and conflict with another, but at the same time not withdrawing one whit from the relationship? Indeed, such ‘inclusion of the negative’ normally can make a relationship…more solid and trustworthy (May 1980: 216).

Once this is stated, it seems rather obvious that a therapist who believes that to be genuine is important (and humanistic practitioners typically do) must be willing to express negative emotions when this could be therapeutic. This will often involve self-disclosure.

Shadley (2000) has laid out a range of types of self-disclosure, from her research:

Intimate Interaction. Where the therapist tends to open up through verbal and nonverbal expressions of therapeutic responses, often with references to present or past personal issues are likely. Sometimes this is inevitable, as for example when the therapist gets pregnant. But some therapists, perhaps more women than men, rely on it a great deal.

Reactive response. The expression of verbal or nonverbal responses revealing emotional connectedness within the therapeutic relationship, without revealing personal experiences of the therapist’s outside life. For example, the therapist may cry at something the client has said. Hill’s research (1989) found that this kind of self-disclosure is particularly valuable in relation to outcome. It goes well with an ‘I-Thou’ type of emotional connection.

Controlled response. The therapist maintains a slight distance by limiting self-disclosures to past experiences, anecdotes, non-verbalized feelings or literary parallels.  The therapist chooses carefully which stories or other elements will be most valuable.

Reflective feedback. The therapist offers impressions of client issues, or asks questions that reveal a point of view, but seldom shares personal information or strong emotional reactions.  This is perhaps the standard behaviour taught on courses.

These are stances that the therapist may cultivate more or less consciously. But it has to be said that this set of choices is quite uncharted territory for most therapists of all persuasions. A huge effort of education is needed to let therapists know that all these possibilities are used in the field.  An unpublished study of self-disclosure concludes:

The data indicates (sic) that whilst therapists do choose when and how to self-disclose, the reasons for doing so are carefully and cautiously considered.  Additionally, although disclosing information to the client does betray the assumptions relating to their training and theoretical approach, the rationale is congruent with how they view the fundamental working philosophy of emphasising the ‘person-to-person’ relationship (King 2000).

Within psychodynamic traditions the Jungian position offers a wide range of interest in self-disclosure, with those at the more Kleinian (psychoanalytic) end of the spectrum probably being the least likely to say much about themselves. Sedgwick comments that for Jung (and Searles, with whom he compares him) therapeutic anonymity is almost as unnecessary as the pretension to neutrality. If the therapist experiences affects, and uses himself or herself in the work with the client, this implies that there will be some expression of this. Jung believes in spontaneity (Sedgwick 1993:123). There would be little point in the analyst discovering the archetype of the wounded healer if he or she never shared this knowledge with the patient.  ‘If the analyst has been moved by his patient, then the patient is more aware of the analyst as a healing presence. This frees the patient to have communication with his inner healer, identified as his potential’  (Samuels 1985:189). Another Jungian writes this:

I don’t conceal myself.  I frequently use myself as therapeutic instrument, and I don’t try to be anonymous.  Often I share my feelings with analysands – if the feelings are relevant …  I think an analysand is entitled to reactivity from his analyst.  Of course, I may temper my reactions.  It may be that the ego of the person I am working with is fragile, and if I get angry, for instance, I will hold back.  But I would surely sooner or later be able to say, ‘You know, ten minutes ago I was terribly angry at you, and the reason was . . .’. But if the patient is not too fragile, I might say, ‘The hell you say, you goddamn fool’.  Or I might laugh, or cry – any of these reactions is possible.  (Wheelwright 1982:117)

Family therapy, which often involves live supervision, may well mean obvious self-disclosure of therapist’s feelings and thoughts to the family group. Indeed in group therapy, even of the psychoanalytic sort, the therapist is more transparent, although there is a major difference between on the one hand the Foulkes (Foulkes and Anthony 1965: 62) and the Institute of Group Analysis model where greater openness by the therapist, owning personal feelings, acts as a model for the group members as to how to relate. On the other hand the Bion (1961, 1962) and the Tavistock Institute of Human Relations model tends towards a somewhat distant and esoteric group consultant, who makes group interpretations and is more concerned with the actual group than the relationships of the individual members to one another within it, except inasmuch as they are indicative of basic assumptions in the group. Here self-disclosure is almost non-existent, although of course Bion (1961: 29-40), in his early experiences in groups, was ready to disclose that he did not know what was happening, and was hammered for it!

The humanistic tendency is to self-disclose, with a humanistic psychologist, Jourard, writing the classic work on self-disclosure as a general form of interaction. Jourard (1968, 1971) found in his research that self-disclosure is the optimum way of building up a human relationship of the ‘I-Thou’ kind. For the person-centred therapist congruence is one of the main core conditions, which is likely to involve self-disclosure in order to be fully responsive to the client. Perhaps one of the most extreme examples of this in the British literature is the person-centred therapist Thorne’s naked embrace with a particular client (Thorne 1987), where he claims this event was both at a stage of ‘deep mutuality’ and was a ‘direct response to her overwhelming fear of fragmentation and of corruption’ (Dryden 1993: 115). Elsewhere this is held up as a ‘striking example’ of congruence (Mearns and Thorne 1988: 88). It is an instance of self-disclosure or self-exposure indeed, which extends to the physical. But we also have to recall the even more bold experiments in the 1960’s, following the existentialist psychoanalyst R.D.Laing’s innovative ideas in the Kingsley Hall community in London (e.g. Barnes and Berke 1973). Questions of such extreme self-exposure raise questions about boundaries, but similar questions about boundaries exist in milder instances of self-disclosure too, where therapists of different orientations have the same regard for the importance of boundaries, but interpret the precise point of them differently. Spinelli (2001) quotes the example of one therapist, who had great success with a difficult patient using a treatment which included talking about his childhood, his marriages, his despair with certain patients, the youth culture, TV and movies, extending sessions, going on walks together, inviting her to his home, providing her with meals and so on.  This seems quite a radical take on boundaries, but Spinelli appears to be quite taken with it, and certainly to regard it as legitimate. Similar issues are raised by Gale (1999). Smith et al. (1998) is a useful text on these questions. It discusses the whole question of boundaries in some detail, with a good deal of research evidence. It is particularly oriented towards the question of touch, but its findings are highly relevant to other boundary questions as well.

The core condition of congruence in the person-centred tradition seems to imply self-disclosure, although Mearns and Thorne (1988: 81-2) make it clear that congruence is not the same as self-disclosure: they mean by the latter term other elements of the therapist’s life. Congruence refers to the therapist’s experience of and with the client, and yet even these feelings are only shared when they are relevant to the immediate concern of the client (1988: 82). While masking dislike of a client is called an incongruent response, it is not clear exactly how a therapist would share this aspect congruently. For comparison we note the example of the analyst Winnicott faced with a similar situation, who waited until it was safe to share such a feeling. A patient of his was ‘almost loathsome to me for some years’. He recounts that ‘it was indeed a wonderful day for me (much later on) when I could actually tell the patient that I and his friends had felt repelled by him, but that he had been too ill for us to let him know. This was also an important day for him, as tremendous advance in his adjustment to reality’ (1975: 196). But, again from within the analytic orientation, Searles writes: ‘I have become much freer to express and make therapeutic use of previously suppressed scornful feelings toward patients, finding that these do not destroy, but rather help to activate the therapeutic relatedness’ (1965: 26). Lomas is the most obvious proponent of therapist honesty in British psychodynamic writing, who, while recommending that it is unwise to ‘indulge thoughtlessly in self-revelations’ (1999: 65) provides countless examples in his books of speaking his mind, answering patients’ questions, and sharing his own dilemmas (see also 1973, 1994).

When we come to examine psychoanalytic writing more closely we find quite a number of indications that self-disclosure, of various kinds, is not as unusual as might initially be thought. Freud addressed the issue of self-disclosure as he worked out the most useful techniques to use, but cited various problems that are likely to arise if the therapist responded to the natural wish of the patient that the analyst should say more about him- or herself. He concludes that ‘experience does not speak in favour of [a] … technique of this kind’ (1912a: 118). Yet if we read some of his case histories, and reports of those in analysis with him, it is obvious that he spoke of himself and his family.

Against this, again in the early years of the development of psychoanalysis, Ferenczi proposed mutual analysis, and we might also refer to the example of Anna Freud as just one of many analysts who clearly have not hidden behind a blank screen. As Greenberg writes: ‘On the one hand we have the testimony of the founder himself that self-disclosure demonstrably undermines our attempts to conduct an analysis. On the other hand we have the testimony of the man who was widely acknowledged to be the foremost clinician of his day that refusing to reveal ourselves demonstrably undermines our attempts to conduct an analysis’. He adds that ‘the same arguments that Freud and Ferenczi made, putatively based on the same empirical observations, are regularly repeated in contemporary discussions of the issue’   (1995: 195).

There is a strand of thinking in psychoanalysis that suggests that just as fostering of the transference will be jeopardized by self-disclosure by the analyst, so towards the end of therapy, self-disclosure, or more participation through what is sometimes known as ‘real relationship’ will help to weaken or dissolve the transference, in preparation for the end of the relationship. Similarly, as Kramer observes, in four-times-a week analysis where an intense transference is encouraged, self-revelation is minimal, whereas in once-weekly therapy, where a dependent, regressive transference is not desirable, ‘judicious revelation of personal material minimizes’ the risk of such extremes (Kramer 2000: 75-6). This appears somewhat manipulative of the therapist, and says that the therapist has the power to regulate the client’s feelings. Experience suggests otherwise: transference can be intense, even when there is self-disclosure. Sharing a personal difficulty in order to try to temper a patient’s idealization can lead to even greater idealization, because such honesty can make the therapist into an even more admired figure.

Nuttall (2000: 28-9) argues that this shift in relating takes place in Kleinian psychotherapy too, so that although it may start in a one-sided way, with the transferential relationship pre-eminent, as therapeutic change occurs the relationship becomes much more person-to-person; although he makes it clear that this cannot be contrived: ‘Physical contact, self-disclosure and social conversation are anathema to a Kleinian therapist’ (2000: 29). However, Hill (1993), who describes very different experiences with three Kleinian therapists, only one of whom fits the stereotypical picture of a tightly boundaried analyst; the effect on him ‘was not only personally therapeutic in itself but I became aware of my own analytic work both improving and becoming more enjoyable with the whole of myself becoming engaged in a more interesting exploration’  (1993: 469). This is another good example of what moving to the authentic position is like.

An interesting position is set out by Greenberg (1995), who suspects the wrong question is being raised, when it is asked whether analysts should self-disclose. Ferenczi said that it is inevitable that personal information will be revealed. So, firstly, ‘asking a particular question (instead of another that could have been asked); making a particular interpretation; decorating one’s office in a particular way; greeting (or not greeting) the patient; wearing a certain kind of tie; cutting one’s hair’ (1995: 195), and indeed everything that a therapist does or does not do reveals something of the therapist to the perceptive patient. Indeed, Singer (1977: 183) writes: ‘what analysts so fondly think of as interpretations are neither exclusively nor even primarily comments about their clients’ deeper motivations, but first and foremost self-revealing remarks’. Secondly, everything the therapist does also conceals something. ‘Even in moments when we are telling our patients about ourselves we are, consciously or unconsciously, deciding what not to say’ (Greenberg 1995: 196). Thirdly, ‘whatever is revealed is simply one person’s understanding at a given moment—never (despite the patient’s and sometimes also the analyst’s hopes) the last word on the subject’ (1995: 197). Greenberg concludes that:

it is not particularly useful to attempt to come up with any sweeping statement about self-disclosure.  I do not see any advantage to covering a wide range of situations with a one-size-fits-all technical prescription.  Rather, our task requires coming to grips with an endless flow of decisions, each made by a particular analyst, with a particular patient, in the context of a particular moment in their relationship.  In general, talking about how we arrive at decisions strikes me as more interesting than the particular conclusions we reach, especially when those conclusions are idealized as the only ones that are acceptable. (1995: 197)

This does seem like the kind of wisdom I should like to urge here, because it illustrates the openness so characteristic of the authentic self.

Burke and Tansey are more circumspect: ‘Although we have attempted to provide clinical guidelines for establishing a “judicious balance” (Tansey and Burke 1989: 3–150), disclosure…remains nessarily ambiguous and uncertain’ (1991: 379). Renik, writing in 1995, observes that:

the whole trend of the past ten years or so toward a theory of technique based on an intersubjective conception of the analytic situation has begun to treat analytic anonymity as a myth and to address the idealizations promoted by the myth . . . the assumption that an analyst can be anonymous and can function as privileged interpreter of a patient’s experience (‘realistic’ versus ‘distorted by transference’) is rejected. Instead, the patient is recognized to be as much a legitimate interpreter of the analyst’s experience as vice versa.    (1995: 480)

He cites Aron as illustrating the clinical implication of this view:

I often ask patients to describe anything that they have observed or noticed about me that may shed light on aspects of our relationship . . . I find that it is critical for me to ask the question with the genuine belief that I may find out something about myself that I did not previously recognize . . . in particular, I focus on what patients have noticed about my internal conflicts. (Aron 1991: 37)

Freud and many of his followers may have rejected Ferenczi, but ‘mutual analysis’ appears to have survived! Renik is clear where he stands. After quoting Hoffman’s (1994: 198) candid admission that ‘the magical aspect of the analyst’s authority is enhanced by his or her . . . anonymity. There is a kind of mystique about the analyst that I doubt we want to dispel completely’, Renik adds: ‘We may not want to dispel it, but I think we should!’ (1995: 481). The wish to dispel mystique is the typical authentic aim, and very characteristic of those therapists I have been mentioning here. The latest word on this topic is the book edited by Andrea Bloomgarden and Rosemary Mennuti (2009), which has some very moving examples of where the boundaries were challenged with good results.

Finally, I would just like to say a word about boundary maintenance versus boundary awareness.  Boundary maintenance is all about holding fast to the strict boundaries specified in the particular type of therapy being offered; boundary awareness is more about keeping a sense of boundaries, with the help of the supervisor, even when they are being varied or modified because of the needs of this particular client.  In my opinion boundary maintenance is important, but much more important is boundary awareness.

John Rowan, PhD, has been a psychologist since 1963, and a psychotherapist since 1980. He has been meditating every morning since 1982. His next workshop, on the AQAL approach to psychotherapy, is in March 2012.


Aron, L. (1991) ‘The patient’s experience of the analyst’s subjectivity’, Psychoanalytic Dialogues, 1: 29-51.

Barnes, M. and Berke, J. (1973) Mary Barnes: Two Accounts of a Journey Through Madness. London: Penguin Books.

Bion, W.R. (1961) Experience in Groups and Other Papers. London: Tavistock Publications.

Bloomgarden, A and Mennuti, R B (eds)(2009) Psychotherapist revealed: Therapists speak about self-disclosure in therapy  Hove: Routledge

Burke, W.F. and Tansey, M.J. (1991) ‘Countertransference disclosure and models of therapeutic action’, Contemporary Psychoanalysis, 27: 351-84.

Clarkson, Petruska (2003) The therapeutic relationship (2nd edition)  London: Whurr

Dryden, W (1985) (ed) Therapists’ Dilemmas  London: Harper & Row

Dryden, W. (1993) Questions and Answers on Counselling in Action. London: Sage.

Farber, B.A., Brink, D.C. and  Raskin, P.M. (eds) (1996) The Psychotherapy of Carl Rogers: Cases and Commentary. New York: The Guilford Press.

Foulkes, S.H. and Anthony, E.J. (1965) Group Psychotherapy: The Psychoanalytic Approach (2nd edition). London: Penguin Books.

Freud, S. (1912a) ‘The dynamics of transference’, Standard Edition, Volume 12. London: Hogarth Press.

Gabriel, L (2005) Speaking the Unspeakable: The ethics of dual relationships in counselling and psychotherapy  London: Routledge.

Gale, D. (1999) ‘The limitations of boundaries’, in C. Feltham (ed.) Controversies in Psychotherapy and Counselling.  London: Sage.

Greenberg, J. (1995) ‘Self-disclosure: is it psychoanalytic?’, Contemporary Psychoanalysis, 31: 193–205.

Hill, J. (1993) ‘Am I a Kleinian? Is anyone?’, British Journal of Psychotherapy, 9: 4, 463-75.

Hoffman, I.Z. (1994). ‘Dialectical thinking and therapeutic action in the psychoanalytic process’, Psychoanalytic Quarterly, 63: 187-218.

Hycner, R. (1993) Between Person and Person: Toward a Dialogical Psychotherapy. Highland, N.Y.: The Gestalt Journal Press.

Imes, S (1998) ‘Long term clients’ experience of touch in Gestalt therapy’  in E W L Smith, P R Clance & S Imes (eds) Touch in Psychotherapy  New York: Guilford.

Jourard, S.M. (1968) Disclosing Man to Himself.  Princeton: Van Nostrand.

Jourard, S.M. (1971) The Transparent Self (2nd edition). New York: Van Nostrand Reinhold.

King, M. (2000) ‘Examining therapists’ rationale for self-disclosure’. Research Dissertation,  Surrey (UK) University.

Kramer, C.H. (2000) ‘Revealing our selves’, in M. Baldwin (ed.) The Use of Self in Therapy (2nd edition).   Binghamton: The Haworth Press.

Langs, R. (1982) Psychotherapy: a Basic Text.  New York: Jason Aronson.

Lawry, S S (1998) ‘Touch and clients who have been sexually abused’ in Smith, Clance & Imes (eds) Touch in Psychotherapy  New York: Guilford.

Lazarus, A A & Zur, O (eds)(2002) Dual Relationships and Psychotherapy  New York: Springer.

Lomas, P. (1973) True and False Experience. London: Allen Lane.

Lomas, P. (1981) The Case for a Personal Psychotherapy. Oxford: Oxford University Press.

Lomas, P. (1994) Cultivating Intuition: an Introduction to Psychotherapy. London: Penguin Books.

Lomas, P. (1999) Doing Good?: Psychotherapy out of its Depth. Oxford: Oxford University Press.

May, R. (1980) Psychology and the Human Dilemma.  New York: W. W. Norton.

Mearns, D. and Thorne, B. (1988) Person-Centred Counselling in Action. London: Sage Publications.

Nuttall, J. (2000) ‘Modes of therapeutic relationship in Kleinian psychotherapy’, British Journal of Psychotherapy, 17: 1, 17-36.

Renik, O. (1995) ‘The ideal of the anonymous analyst and the problem of self-disclosure’, Psychoanalytic Quarterly, 64: 466-95.

Russell, J (1993) Out of bounds: Sexual exploitation in counselling and therapy  London: Sage

Samuels, A. (1985) Jung and the Post-Jungians. London: Routledge and Kegan Paul.

Searles, H.F. (1965) Collected Papers on Schizophrenia and Related Subjects. London: Hogarth Press/Karnac Press.

Sedgwick, D. (1993) Jung and Searles: A Comparative Analysis. London: Routledge.

Shadley, M.L. (2000) ‘Are all therapists alike?  Revisiting research about the use of self in therapy’, in M. Baldwin (ed.) The Use of Self in Therapy (2nd edition). New York: The Haworth Press.

Singer, E. (1977) ‘The fiction of analytic anonymity’, in K. A. Frank (ed.) The Human Dimension in Psychoanalysis. New York: Grune and Stratton, pp. 181-92.

Smith, E.W.L., Clance, P.R. and Imes, S. (1998) Touch in Psychotherapy: Theory, Research and Practice. New York: Guilford Press.

Spinelli, E. (2001) The Mirror and the Hammer: Challenges to Therapeutic Orthodoxy. London: Continuum.

Tansey, M.J. and Burke, W.F. (1989) Understanding Countertransference: From Projective Identification to Empathy. Hillsdale, NJ.: Analytic Press.

Thorne, B.J. (1987) ‘Beyond the core conditions’, in W. Dryden (ed.) Key Cases in Psychotherapy. London: Croom Helm.

Wheelwright, J.B. (1982) ‘Termination’, in M. Stein (ed.) Jungian Analysis.  Boston: Shambhala.

Winnicott, D.W. (1975) Through Paediatrics to Psychoanalysis: Collected Papers. London: Karnac Books.