Over the last few years many analysts and therapists have been surprised to find an increasing shift in the way that supervision is understood in documents concerned with regulating practice. These may come from UKCP or from organisations which set themselves the task of specialising in questions relating to supervision. If there is an overriding tendency here, it is to construe supervision as a process of transmitting knowledge and skills from one party to another. The supervisor, on this model, is more mature than the supervisee, who will benefit from the transfer of knowledge and improve his or her handling of the clinical work deemed to be 'under supervision'. The aim of supervision is defined as the enhancement of the supervisee's professional development, taking place in a supportive context of education and 'coaching'. This benevolent atmosphere, it is also claimed, should include a 'monitoring' function of the supervisee's work.

This conception of supervision is certainly not without its precedents in the history of psychoanalytic and psychotherapeutic theory, yet what is startling is the way in which all the other variables once seen as central to supervision have disappeared. Transference and so-called counter-transference have become more or less absent, and supervision is understood as a knowledge-based endeavour similar to those common in non-analytic fields such as business consultancy and marketing where an expert imparts methods and skills for a fee. This redrafting of the supervisory process may be seen, in part, as a consequence of political change. Now that the government has recognised the value of talking therapies and set itself the task of guaranteed service provision within a set time period for those judged to be in need, the only mechanism which would make this goal possible is a deskilling of the workforce. As psychology graduates and even students are brought in to offer rudimentary 'therapy', so the role of the more fully trained - and much more expensive - therapists becomes subtly pushed towards group supervision rather than direct practice. This is arguably one aspect of the future for therapists working in many NHS trusts.

It also implies that therapists will spend more and more time supervising trainee clinicians who have had no analysis or therapy themselves, a problem that is currently being widely debated. This is crucial to the question of supervision since, once we assume that the supervisee has not embarked on the self-exploratory journey of a talking therapy him or herself, supervision starts to swing in precisely the direction that the regulatory literature has taken. Once unconscious dynamics are deemed peripheral, all that's left is the knowledge and skills model of supervision in which the supervisor becomes almost responsible for the trainee's clinical work. This suggests that any critique of this model of supervision must also insist on the centrality of the experience of personal analysis or therapy for the trainee. And isn't this, after all, exactly what defines our field of the talking therapies?

Some History

The parsimony of Freud's comments on supervision is well-known. In the 1919 article on the teaching of psychoanalysis in universities, he refers to the 'teaching of psychoanalysis' as part of analytic training, adding that the analyst can - not 'ought to' or even 'should' - get supervision and guidance from "recognised psychoanalysts". Beyond a few similar remarks, there is no theory of supervision in Freud, although there are several anecdotes about Freud's actual supervisory practice. Freud would meet with students like Eduardo Weiss to discuss clinical cases, just as many of the first generation analysts would consult with each other about problematic aspects of their work. Rado would go to talk with Ferenczi, for example, and such discussions appear to have been initially informal and unstructured. They also appear to have been conducted outside any kind of financial contract, more like a variety of collegial meeting.

These informal beginnings were to change with the systematisation of analytic training after the establishment of the Berlin Institute. As the regulations of the Berlin training became progressively formalised during the 1920s, supervised practice became not only an obligatory component of analytic formation but also the subject of theoretical debate. In the early 20s, if a supervisor had doubts about the competence of the clinical work of a particular trainee, they were advised by the Berlin rules to actually take over the case being discussed in supervision. By 1930, trainees had to have completed two years of supervised case work. In Vienna, where training requirements were often at odds with those in Berlin, each trainee had to have two supervisors, with a minimum of four supervised cases. The analyst was separated here from the supervisor, whereas in Budapest, many trainees were encouraged to start their practice with supervision from their own analyst.

By the mid-30s, a real debate about the rationale of such practices had begun. Helene Deutsch, writing in 1935, proposed that supervision aimed at removing the "sediment" from the trainee's work, allowing the supervisor to "see" the patient. This was in fact a quite literal aim for Deutsch, who claimed that supervision was almost impossible unless she had seen or spoken to the patient at least once. Supervision would allow her to "observe" over the "candidate's shoulder", a preference echoed many years later by the French analyst Serge Lebovici who noted that he required "a physical description of the patient - his clothes, his habits" so he would be able to visualise him (Lebovici 1983). One wonders what could be achieved by this, apart from the minting of fresh prejudices. Deutsch's supervisory method, likewise, might seem peculiar in terms of the later history of supervised practice. She would ask her supervisees to write down their ideas about a case, and then free-associate. Theoretical intervention, she thought, was best avoided and the dynamic principle of her supervisory sessions was analytic, relying on speech and association.

Deutsch also implemented a practice in the Vienna training that had its own institutional logic. A supervision seminar was introduced, in which two cases would be tracked over time, one presented by a trainee, the other by an experienced analyst. This would have the effect of allowing not only new perspectives on the cases in question, but also challenging the received separations of analytic generations. The experienced analyst, it might turn out, was committing errors just as staggering as the neophyte, while the neophyte could be showing a clinical agility that was nowhere to be found in the senior's practice. The seminar, indeed, put dialectic rather than hierarchy in the Institute's training programme.

In Budapest meanwhile, Vilma Kovacs was developing her own theory of supervision. This pioneering analyst - an agoraphobic successfully analysed by Ferenczi and the mother of Alice Balint - believed that the first supervisor should be the trainee's analyst. "If the candidate", she wrote, "continues his own analysis when he begins to analyse patients, the two parallel pieces of work bring to light those sides of his personality which have hitherto received too little attention or none at all, or at least could not manifest themselves in so expressive a fashion. All his good and bad qualities and also his weaknesses, are revealed: for example, his incapacity for objectivity; his impatience; his vanity; his inability to bear criticism; the tendency to observe only what is in his favour and the failure to note the serious accusations which the patient is bringing against him but dares not express except in a disguised form; the tactlessness which ministers to those sadistic or masochistic instincts in himself which he has failed to master; his callousness or, on the other hand, his exaggerated fellow-feeling and excessive tolerance". This sobering list immediately renders the apparent transparency of supervisory aims problematic, and points to the central place of unconscious desires and drives in the analytic project itself. "All this", she concludes, "provides an opportunity for showing the student the right way to handle the counter-transference, which is one the most important factors in analytic work". Without this, the supervisory relation will languish in a false ideal of "pure humanitarianism".

Kovacs provides a solid argument for the non-separation of analyst and supervisor in the early days of a trainee's practice, although we know that her advice was not followed widely in the Budapest group, contrary to many historical accounts. The 'London Standard' of 1947 decreed, in fact, that the supervisor had to be distinct from the analyst, yet without providing any solid arguments against Kovacs' polemic. Her emphasis on the counter-transference, however, was shared by many other analysts who did not agree with the non-separation principle. The focus on the trainee's unconscious dynamics, and how these were operative in clinical work, would become the hallmark of one major current in the literature on supervision. Rather than assuming that the goals of a treatment were obvious and universal, each analyst would have to decipher in each case what exactly was being sought beyond the myth of the well-being of the patient. This would involve, by necessity, a questioning of their own tendencies to respond to human distress.

In contrast to Kovacs' position was the more pedagogic stance of Bibring. On this model, the aim of supervision is primarily didactic. The supervisor instructs the supervisee, in particular about the timing and form of interpretations. Counter-transference is not taken to be the key issue, and supervision is seen as an educative process. This stance is of course quite close to what is being claimed today, and the tension between the position of Kovacs and that of Bibring is often dubbed the 'Treat or Teach' problem. Should the supervisor treat the supervisee or teach them? Should the supervisory encounter focus on analysing the supervisee's unconscious baggage or simply transmitting knowledge and skills? We will see later on why this formula poses a false and fundamentally misleading problem.

The debate on supervision would shift during the 1950s to the question of institutional process. Where the earlier debates had tended to privilege the unconscious dynamics at play in supervision, it was now reframed as a question about the role of the supervisor in a training programme. If the trainee believed that the supervisor would report anything of negative note to the training committee of the institution, surely the freedom of the supervisory relation would be compromised? The supervisor would be little more than a judge. In England, Paula Heimann and Michael Balint pointed to the dangers of such an approach, and insisted on the essential privacy of supervision. This question is still quite pertinent today, as the following vignette, reported by Estelle and Morton Shane, demonstrates.

"One candidate wrote up her case report for graduation (and, ultimately, for certification by the American [Psychoanalytic Association]), showing it to her supervisor. Her supervisor called attention to the fact that the candidate had included an exchange between herself and her patient during the last hour of the analysis, which had gone on to a successful termination. The supervisor noted that in the exchange, the analyst answers her patient's question, indicating that, as the candidate knew well, questions coming from patients are best handled by interpretation. The candidate responded that over the course of her work with the patient, this particular question had been well analysed, and in fact was dealt with many times over the years. In the context of the patient's last hour, however, and under the pressure of the patient's urgings and reasonable importations ("After all", the patient argued, "this is the last hour, and the last time we can go over this ground"), the candidate provided a response. The question had centered around a work of art that had been admired frequently and asked about often by the patient, and the question itself had many meanings in different contexts over the years. In the last hour, the patient had simply asked: "Where did you but that statue?" When the analyst replied simply that she had purchased it in Singapore, the patient was relieved and enlightened, saying that he, too, hoped to travel now that he was finishing his analysis, feeling himself to be both wealthier and more adventurous. It appeared to the analyst to be a meaningful exchange, and the supervisor was convinced, as well, that it was important enough to let the incident stand as reported, even when presented for certification to the American. The outcome was unfortunate. The candidate was, while graduated from her institute in excellent standing, turned down for certification, attention being paid especially to this incident in the last session where a question was answered rather than analysed".

This curious story highlights both the question of the transmission of information and the rigidity of technical rules. It also demonstrates a rather simplified view of what it means to have a question answered. Telling the patient the statue came from Singapore can hardly be said to be answering the patient's question, if we assume that beyond every demand there lies another dimension of human subjectivity such as, for example, a desire. Replying to her that he too would travel due to his new wealth and adventurousness is also indicative: on the one hand, there is the nuance of an identification, on the other, of an aggression. Without knowing anything else about this incident, we could ask why the trainee wished to include it in the report: if it was so important, it could hardly be the innocuous exchange described, and if it was so meaningful, to merit inclusion, what exactly did it mean? The vignette tells us more, perhaps, about the style of the therapist and her wish to not allow the patient to leave without giving him something.

The story, then, illuminates two questions about supervision. Firstly, the relation to the institution and all the problems this raises. And secondly, given the apparent innocence accorded to the exchange, the idea that there can be aspects of the relation between two parties working clinically that are somehow exempt from the effects of the unconscious. The enquiry about the statue and the therapist's response are presented as if now, finally, they can be deemed value-free or unconscious-free, as if a margin of speech could be separated off and understood as testifying to a real, authentic exchange, uncontaminated by the demons that Freud had set at the heart of the analytic experience.

Treat or Teach?

These different questions would be debated with some frequency in the 60s. The search to remove the 'sediment' from speech would find its apotheosis in the work of Lawrence Kubie, who tried, with his colleagues, to tape-record both analytic and supervision sessions and then compare the results. Observers would be asked to listen to the different sets of material and assess the agreements and disagreements between the content of the analytic sessions and that of the supervisions. Had everything been reported? What bias had been introduced? How faithful was the record of the patient's speech? This method relied on a profound misconception not only of supervision but of speech itself, as if words could be reduced to some transparent content and their interpretation and relevance understood by a third party. As for supervision, it supposed that its aim was simply to transmit as accurate a record as possible of what went on in the session. This strange idea is echoed in the practice of many supervisees today who try to 'write up' complete sessions, as if to include everything and convey to the supervisor exactly what was said.

As for the relation of supervision to the institution, this was the subject of much debate, particularly in the States, together with a return to the older problems posed by the Budapest group. Should the analyst and the supervisor necessarily be distinguished, and should supervision include therapeutic effects? This so-called 'treat or teach' dilemma was discussed with sensitivity by the American analyst Daryl Debell in a 1963 paper which shows that the basic questions of supervision were actually pondered with greater subtlety then than they are now. Reviewing recent literature, Debell noted how for many authors, too much emphasis on counter-transference risked obscuring the patient's narrative, whereas for others, too much emphasis on the patient threatened to marginalize the crucial issue of, precisely, counter-transference. However, if the aim of supervision was defined simply as the effort to increase the clinician's ability to analyse, the 'treat or teach' question becomes meaningless. It would suggest, in addition, that the effects of the unconscious in speech had suddenly vanished. A supervisor might say something to a supervisee that would have unconscious effects, whether intended or not, just as the question about the statue we discussed earlier - and the reply to it - might be the form of a host of other questions beyond the conscious grasp of either of the parties. To say that supervisory intervention either treats or teaches is to ignore the most rudimentary facts about human speech that psychoanalysis rests on.

It often happens that a supervisor feels, for example, that the supervisee is too immersed in the patient's material. They report it with great gusto and feeling, as if they themselves were the speaker of the patient's words, and their affective reaction may suggest that they have little distance from the material. In such circumstances, familiar to anyone with a practice of supervision, it may be prudent to respond to the supervisee not with an interpretation - 'You're too involved' - but simply with the very distance that seems absent in the supervisee's approach. As the supervisee enthuses about the material, the supervisor might show no interest, and this lack of interest may then allow the supervisee to start thinking about their own implication in the material. Likewise, a supervisee's distance from and boredom with the material could be met with a parallel response: the supervisor shows absolute fascination with the case under discussion. This might encourage the supervisee to start asking themselves questions about the case, and so on. The underlying assumption here is that speech is a dialectical and dynamic relation rather than simply a vehicle of meaning, and that people do not change if you present meanings to them (eg 'You're too involved').

Let's take two clinical examples. In the first supervisory encounter, the analyst reports to the supervisor that the patient has been speaking about the daydreams she has before going to sleep: she marries a famous actor and has a child with him. The supervisor responds: 'Did you point out to her that that sort of daydream is very common?' This minimal exchange opens up a range of questions. Does it, for example, treat or teach? At the level of knowledge, it might be asserting that yes, indeed, such daydreams are very common. But the crucial fact here is that the supervisor has said this to the supervisee. This changes everything. It is no longer a communication of knowledge but a dynamical act, putting in question the supervisee's response to the material. It implies, at one level, that the supervisee was perhaps too impressed by this obviously Freudian material. At another level, it questions why this material was brought to supervision. For the supervisee, it trivialises the material, perhaps undermining an idealisation. And, in terms of the analysis itself, it might have suggested to the patient that such a wish is, indeed, common, and this may in turn have an effect of deflation. The supervisor's comment thus works at a number of levels simultaneously, touching not only the supervisee's relation to the supervisor but also his relation to the patient and the patient's relation to her daydreams. It thus 'treats' and 'teaches' at the same time, touching both the supervisee and the patient.

In the next vignette, the supervisee has chosen to speak about a case of hysteria, in which the symptoms include hysterical paralyses. The supervisor asks 'Have you taken her reflexes?'. The supervisee is amazed: 'Nnnnnooo', he stammers. The supervisor: 'Do you have a reflex hammer?'. Supervisee: 'No!' The supervisor: 'Then go and buy one!', and then adds 'Along with some headed notepaper'. Do these interventions treat or teach? Once again the distinction has little purcahse, and the supervisor's remarks work at several levels. To start with, they question the place from which the supervisor thinks that his clinical work is being looked at. Expecting an analytic gaze, pleased with the classical symptoms of a textbook hysteria, he receives something quite different. The supervisor is taking the patient's symptoms, in fact, quite seriously, literally even. By suggesting the use of a hammer for reflexes, the supervisor is questioning the supervisee's self-image. And also, in terms of the case, there is an implicit questioning of the place that the patient is putting the clinician in: analyst or medical doctor? The final comment makes it clear that the ego of the supervisee is being questioned: he went to see his very prestigious supervisor, perhaps, not to interrogate his own practice but to consolidate his self-image as a 'psychoanalyst'. The absurdity of this project is brought out nicely by the advice to get some headed notepaper.

What these examples show so clearly is the way that supervision can be construed as a process designed to make the supervisee question what they are doing. This does not mean telling them what to do, but encouraging them to give an account of the rationale of their action. In justifying their action, many things will be brought to the surface and many questions opened up. Why, for example, has the supervisee accepted the demand for treatment of the specific patient they are working with? Beyond the possible financial incentive, this question must be answered in each case. What, likewise, are they aiming at in each particular case they work with? Why, finally, do they practice at all? Supervision, certainly, is not the only arena in which to open up these questions, and the countertransferential problems involved may be raised but, as Grotjahn pointed out many years ago, as a rule not settled there. Its function is less to educate and transmit skills than to give the supervisee a space in which they can - and must - justify what they are doing. Surely this is an ethical consideration which cannot be ignored in any serious analytic formation.

The contemporary perspectives which construe supervision as a procedure that can be administered from one party to another are therefore misguided. As a relational process, it relies on the dynamics of speech and the effects of the unconscious. Two bodies meeting in a room where there is a note on the door saying 'Supervision in progress' does not constitute supervision. As a dynamic relation, it may or may not occur, but it cannot, contrary to the new ideology of supervision, be applied like a drug or a plaster. The so-called 'parallel process' models of supervision which emerged during the 70s and are still popular in some quarters today might look like a viable alternative to the bandage model, but they are also beset with problems. Searles, Arlow and Grinberg all proposed some version of the idea that what goes on in a supervision will parallel what goes on in the analysis being supervised.

This view is highly optimistic It supposes that the unconsciouses of the supervisee and supervisor are so malleable that they act like a kind of photographic plate for those of the patient and the supervisee. What occurs in one place transmits itself to another in a vaguely paranormal process which functions via automatism. If we were to take the idea at all seriously, there would be consequences for today's suggestions that each supervisor should in turn be supervised him or herself. If the automatism did indeed function so blindly, surely the original relation between analyst and patient would be reduplicated in an endless hall of mirrors?

If supervision, as we have seen, involves unconscious dynamics and the unpredictable effects of speech, it would seem obvious that there would be attempts to master its indeterminacy. This is exactly what we have seen in the recent climate, in which supervision becomes the site of an ever-expanding host of regulatory interventions. Supervisors need to be specially trained and registered, and require their own supervisors. In the analytic field, this is echoed in the work of an analyst like Langs who has set out a stringent list of rules for supervision. Supervision, according to Langs, should be 'sterilised', and this includes strictures on the setting and even the requirement that analysts and supervisors have separate conferences so they don't meet each other. "Psychoanalysts", Langs writes, "desperately need secured frames within which to learn" (Langs 1994). Supervisees and supervisors, therefore, should have absolutely no contact outside their supervision sessions. Readers of Freud will recognise here the very structure not of analytic thought but of the obsessive symptom: rigid separations are appealed to only as a measure to undo what is known to be a prior contamination.

What Is Supervision For?

So what can we learn from these different conceptions of supervision? They all derive ultimately from very different conceptions of what psychoanalysis is about. On the one hand, there's the medical model of psychoanalysis, based on a belief in knowledge. Analysis is about learning and increasing the field of knowledge. When the beginner starts to work analytically with others, therefore, it is paramount to understand as much as possible. Hence they need help from someone more experienced. The aim of supervision, according to this view, is to help the supervisee understand the patient more fully, and the supervisor is then left with three basic positions: that of pedagogue (we can think of Grinberg's claim that the supervisor "offer himself as a model of identification" or of Casement's 'internalised supervisor'); that of the defender of the respectability of the analytic institution; and worst of all, that of the defender of the patient.

Are any of these compatible with an analytic position? The recent emphasis on the first and the third of these options suggests nothing less than that the supervisor substitute him or herself for the supervisee. One wonders what effect this pull towards identification might have on the treatment, and on the potential rivalry between supervisee and supervisor. Already in 1963, Debell had warned that "the supervisor's responsibility should be rather to insist that identification with the therapist is to be analysed along with other transference phenomena". Yet the very danger to be avoided then has become an imperative to achieve today. To take the logic of the identification model seriously, wouldn't the most coherent outcome be the last resort strategy of the Berlin training: that the supervisor actually step in and take over the case?

These misunderstandings provide fertile soil for fantasy. Supervisor and supervisee, we learn, work together at an "adult, sophisticated level" (Grinberg), graduating from the parent-child relation as "their ego systems face further surges of development" (Stockwell). The clinician who becomes a supervisor is like the adult who becomes a parent, and "the expression of their pleasure in their supervisees' growth…adds to the confidence of the supervisees". Through the "triangulation of supervision", another author writes, "both the trainee and the patient experience…a good containing parental couple able to work within the Oedipal situation of the supervisory dynamic" (Thomas). These glosses on the practice of supervision reveal fantasy positions which can only operate towards consolidating self-images for both supervisor and supervisee, rather than advancing the analysis of unconscious structure.

On the other hand, there is the Lacanian model of psychoanalysis which sees the analytic goal as a questioning of knowledge and its effects. Analysis is all about undermining the appeal of knowledge, and the aim of supervision is, in part, to temper the supervisee's passion for meaning. Contrary to the medical model, it is less the supervisor who knows something here than the supervisee, although this may be a knowledge not 'known' to him or herself. The supervisor, among other things, may encourage the supervisee to realise that they know more than they think they do, by careful and unexpected questioning about a case. Aspects of a case may come to light which the supervisee had not paid attention to, despite 'knowing' them. To take an example, the supervisee presented his interpretation of a case to the supervisor. The latter replied 'You're great. It's exactly that'. Three months later, the supervisee admits 'My construction must be false, things are going from bad to worse. But you told me that it was exactly that. It can't be that!' The supervisor: 'I'm pleased', adding a moment later, 'I'm pleased that you noticed too'.

The vignette illustrates once again a questioning of the place of knowledge. What did the supervisee want in presenting an elaborate construction in supervision? Was it to make himself be seen as an 'analyst'? And, in particular, as an analyst 'who knew something'. The first intervention could be taken as irony, unnoticed by the supervisee, while the second makes this a bit clearer. It shows how the search for meaning can be put in question, and how an analytic orientation involves trying to undo an exclusive appeal to meaning and knowledge, as well as engaging with their limits. Crucial to each supervision, then, is that supervision itself is not taken as a given: why, beyond institutional imperatives, is the supervisee there? And the supervisee is not expected to blindly bring material to each meeting, but to come, rather, with a question or a problem.

The Lacanian approach here does not spring from a vacuum, but has its roots in a practice of supervision that can probably be traced back to Freud, whose preference for Socratic methods is well-known. It was most notably articulated by the Viennese and subsequently New York analyst Otto Isakower with his theory of the 'analysing instrument'. Isakower was interested in the use of the unconscious as an instrument, and saw supervision as a laboratory for the practical application of the technique of psychoanalysis. The aim of supervision, for Isakower, was just to pursue psychoanalysis in the purest way possible. The emphasis was not on the 'treat or teach' cliché, but rather on what could be called a function, the function of the analysing instrument, an idea that resonates closely with Lacan's notion of the desire of the analyst. The aim of this instrument was simple: to analyse.

Let's take an example of Isakower's supervisory practice (Balter et al 1980). "The analysand, a man in his late twenties with marked feminine masochistic trends, had told some events of the previous evening. While driving his woman friend home, they passed through a long tunnel. During the drive, as frequently happened, his friend had been berating him for quite some time for what she felt were his weaknesses. On this occasion, his irritation became so great that he abruptly tapped the brake pedal, causing the car to jolt sharply. This had the desired effect of scaring the young woman into quieting down. That night he had the following dream. He was in front of a small hut in a grassy clearing in a forest. Suddenly, an animal looking like a deer stood before him; it had sharp, knife-like points at the tips of its antlers. In the dream it was called a 'yaki'. The dreamer was afraid that the yaki would attack him, and he fled into the hut. The analysand gave very few associations to this dream. The analyst, however, became intrigued with the name 'yaki'; he asked the analysand what came to his mind about the name, and nothing did. The analyst began to think about the issues of castration and aggression and their possible symbolic representation in the dream in the form of the animal. He also wondered to himself whether there might be a reference to the Tibetan beast of burden, the yak. The analyst asked the dreamer what his associations to the animal were, with the expectation that some reference would be made to his woman friend's 'castrating' behavior and to his own 'sharp' braking of the car. Neither this not much else occurred to him". Isakower, having listened to all this, broke his silence to say 'yakety-yak'.

This marvellous intervention works against the supervisee's rather ridiculous response to the material. He was interested, it seems, more in his own associations to the words used by the patient than to the patient's speech, and the hopes he'd set on the Tibetan reference illustrate this quite clearly. The analysand's material was not being listened to properly, yet Isakower's comment brought his attention back to the level of words, to the signifier rather than the imaginary meanings and associations he himself was adding to the material. The ideas of castration and aggression, although possibly of great relevance here, are constructions of knowledge, and the analyst is keen to bring them in. Isakower, with more prudence, returns the focus to speech, as well as deflating the supervisee's associations; yakety-yak, after all, refers not just to the woman in the car's banter but also to the analyst's response to his patient and to his presentation of the material in the supervision. It works at several levels at once, like the other supervisory interventions we have discussed.

What is the function of the supervisor here? He is hardly a pedagogue or a defender of the institution or of the patient. Rather, he is a defender of the analytic process as such, the analysing instrument described by Isakower. His aim is just to further analysis. This introduces what some might find a startling conclusion: that the responsibility of the supervisor is not to the patient whose case is 'under' supervision, nor to the supervisee, but to psychoanalysis itself. This ethical position is not going to be an easy one, and it implies that clinical responsibility lies with the clinician him or herself. The supervisee, according to his view, must give an account of what they are doing and be ready to engage with the consequences. They must be able to do so with rigour, and, rather than appealing to the supervisor to solve their clinical problems for them, they must endeavour to use the space of supervision to do it themselves. The supervisor's comments and questions may, of course, help them to do this. By bringing out what is implicit in the supervisee's work and thinking through the effects of their action, this can have significant ramifications for the direction of the treatment. Perhaps this is why Lacan included supervision under the rubric 'Pure Psychoanalysis' in the Act of Foundation of his School.

Supervision is less about providing knowledge for a fee than learning something about how a clinician works and why. This is what the supervisor must learn from his supervisee, and it is through this learning process that the supervisee's practice may change and develop. Since these hows and whys will be different for each clinician, it becomes imperative to resist the drive to make supervision the site for standardisation and external influence which inscribes supervision within the market of goods and services for sale. A clinical practice can never be taken for granted, and no amount of supervision can guarantee anything about it. Rather, it can allow a supervisee to question their desire to be a clinician, and to realise that their only reference points are the pathways that have constituted their own trajectory to the analytic position.

For those who have not embarked on this trajectory, supervision may be expected to alleviate the anxiety that the real involved in practice releases. Today, the unanalysed graduate or young psychiatrist dispatched to the therapeutic front are likely to turn to supervision for what only the experience of an analysis can really provide. The demands placed on the supervisor are now double: to supply knowledge and to alleviate the symptoms of the supervisee. In other cases, a clinician may hope to continue in their supervision what they have failed to address in their own analysis, returning again and again to the same point of impasse and invoking the image of their supervisor as almost a symptomatic emblem of their dilemma.

As for the supervisor, how can we avoid the question of why he or she would accept a demand for supervision? What would make someone who has done an analysis agree to work with a supervisee? If they subscribe to the pedagogic model, how could the position of transmitting knowledge be compatible with the questioning of knowledge that forms a central part of any serious analysis? If an analysis had taught someone the vanity of occupying the place of supposed knowledge, how could it be taken up lightly in the practice of supervision? What, in other words, could a supervisor want?

Darian Leader

August 2006

Michael Balint, 'Analytic Training and Training Analysis', International Journal of Psychoanalysis, 35, 1954, pp.157-162
Leon Balter et al, 'On the Analyzing Instrument', The Psychoanalytic Quarterly, 49, 1980, pp.474-504.
Emanuel Berman, 'Psychoanalytic Supervision', International Journal of Psychoanalysis, 81, 2000, pp.273-290.
Daryl Debell, 'A Critical Digest of the Literature on Psychoanalytic Supervision', Journal of the American Psychoanalytic Association, 11, 1963, pp.546-575.
Helene Deutsch, 'Control Analysis' (1935), in 'The Therapeutic Process, The Self and Female Psychology', New Brunswick, Transaction, 1992, pp.239-246.
Christine Driver and Edward Martin eds, 'Supervision and the Analytic Attitude', London, Whurr, 2005.
Claude Dumezil, 'D'un Discours a l'autre, l'Institution du Controle', Scilicet, 6-7, 1976.
Rudolf Ekstein and Robert Wallerstein, 'Teaching and Learning of Psychoanalysis', London, Imago, 1958.
Sigmund Freud, 'On the Teaching of Psychoanalysis in Universities' (1919), Standard Edition Vol 17, pp.169-174.
Leon Grinberg, 'On Transference and Countertransference and the Technique of Supervision', in Brian Martindale et al, 'Supervision and its Vicissitudes', London, Karnac, 1997, pp.1-24.
Martin Grotjahn, 'Problems and Techniques of Supervision', Psychiatry, 18, 1955, pp.9-15.
Paula Heimann, 'Problems of the Training Analysis', International Journal of Psychoanalysis, 35, 1954, pp.163-8.
Vilma Kovacs, 'Training and Control-Analyses', International Journal of Psychoanalysis, 17, 1936, pp.346-54.
Lawrence Kubie, 'research into the Process of Supervision in Psychoanalysis', Psychoanalytic Quaterly, 27, 1958, pp.226-236.
Robert Langs, 'Supervision in Training Institutes', Contemporary Psychoanalysis, 30, 1994, pp.75-82.
Robert Langs, 'The Framework of Supervision in Psychoanalytic Psychotherapy', in Martindale et al, op.cit, pp.117-134.
Serge Lebovici, 'Supervision in French Psychoanalytic Education', The Annual of Psychoanalysis, 11, 1983, pp.79-89.
Maud Mannoni, 'Supervision et Sensibilisation au Processus analytique, Apercu historique, Problemes actuels', in 'De la Passion de l'Etre a la 'Folie' de Savoir', Paris, Denoel, 1988, pp.99-135.
Moustapha Safouan, Philippe Julien and Christian Hoffman, 'Malaise dans la Psychanalyse, Le Tiers dans l'Institution et l'Analyse de Controle', Paris, Arcanes, 1995.
Morton and Estelle Shane, 'Un-American Activities and Other Dilemmas Experienced in the Supervision of Candidates', Psychoanalytic Inquiry, 15, 1995, pp.226-239.
Rose Sockwell, 'The Ego and Superego in Supervision', in Driver and Martin eds, op.cit. pp.98-114.
Sandra Thomas, 'Supervision and Training: Two Different Foci', in Driver and Martin eds, op.cit. pp.152-164.


Click here to go back to previous page
All content © 2009 - 2017 The College of Psychoanalysts - UK
a company limited by guarantee No. 4684476 (England & Wales)