AN OVERVIEW OF SUPERVISION OF PSYCHOTHERAPY AS A COMMUNICATIVE RELATIONSHIP (Part 3 of 3)

A Proposal for Change

A frequent response in our society to criticism, when persons who lack or have had limited formal training have limitations in their work, is to prescribe more training. Formal training programs for supervisors are a possibility. Another option, more congenial to persons who have reached a position in society where they no longer have to submit themselves to pedagogical tutelage (esp. certificatory evaluation) is for supervisors to organize peer self-study groups. These possibilities do not, however, address the question: Teach what? Study what?

It may be worth recalling here once again the deutero-proposition that any social process which is not self-destroying is likely to be able to be not only self-perpetuating, but even to be able to interpret its ability to reproduce itself as evidence of its rightness and goodness. Any large social project gives rise to a proliferation of subtasks in such a way that the purposes for which the activity was instituted can fade from awareness in persons endeavoring to make headway with the subtasks; this distraction can go so far that, in their sense of accomplishment in solving one subtask after another, the persons fail to notice that no real progress is being made toward the original purposes, and that the project as a whole has proven itself to be unfeasible through the very fact that subtasks continue to proliferate instead of eventually moving toward closure (Ravetz, 1971, chap. 13). There is an admonitory story (Rev. Howard Waterhouse, 1969, personal communication) about a certain kind of insect, called "processionary caterpillars," because, as they go through their environment looking for food, each butts its front end up against the back end of its predecessor, to form a long chain which breaks up when the leader finds food. Researchers are able, by connecting the head of the leader to the tail of the "caboose" caterpillar, to get the little creatures to creep around in a circle until they collapse from exhaustion, even though, right at hand, are large quantities of food.

This dissertation makes one "simple" proposal. I am proposing one precisely focused intervention in the psychotherapy establishment's "circle" of its institutionalized self-reproduction. I want to urge supervisors and supervisees to pay more attention, in a disciplined way, to their communicative interaction. (The remainder of this dissertation will focus on elaborating and, especially, operationalizing this schema.)

Psychotherapy cannot be reduced to a formula because that destroys its existence as communicative action between therapist and patient. Nonetheless, (to borrow Seymour Papert's and Sherry Turkle's phrases:) "powerful ideas," "evocative ideas" can enrich the empathic (and sympathetic) process when they are appropriated by the participants in an imagination-enriching way (as opposed to the ideas appropriating the participants, or at least, to return to Bateson's notion (1979, p. 107) of the self-confirming nature of social beliefs, the participants acting that way with significant – and often unrecognized – real world consequences). I believe the situation in supervision is analogous, and I wish to contribute my idea as a candidate for supervisors and supervisees (and others) to consider.

In psychotherapy itself, one such "powerful," "evocative" idea is: trauma. This idea has roots at least as far back as Freud and Breuer's Studies on Hysteria (1893-1895/1955), and was being vigorously pursued by Sàndor Ferenczi at the time of his premature death (1933). In the past decade [ca. 1983-93], it has been coming to increasing prominence and popularization, starting with Alice Miller's The Drama of the Gifted Child (1979/1986; originally published in English in 1981, under the title: Prisoners of Childhood). The idea that persons who seek out psychotherapeutic help are often seeking relief from aftereffects of hurtful events which happened to them in childhood – events which often they do not remember and, even if they do remember, the impact of which they do not understand – has energized and inspired many therapists (and patients), not in a programmatic way, but rather as a "Gestalt shift" in general orientation: they see things they didn't see before and they see things they did see before in a different way. Of course this idea is not congenial to everyone. And it would be a cruel irony if this idea finds its way into psychotherapy training programs in such a way that students are pressured (traumatized!) into studying too-long syllabi about it....

In a similar way, I hope the idea of supervisors and supervisees paying attention to their communicative interaction in a disciplined, self-accounting way may be generative. To do this, it has to be something that "clicks" with persons. For this reason, I shall present detailed examples in dramatic form, to maximize the possibilities of ["these here"] stolid written words engaging with (your, my reader's) living imagination. My dissertation shall succeed in this its purpose if, after reading it, supervisors and supervisees (and, by analogical extension, persons in other communicative situations in life) find themselves thinking: "Whoa! Here's one of those situations McCormick's dissertation was talking about." If, further, the person having this thought feels safe in the relationship, hopefully he or she will then go on to think: "I'm going to stop here, call our attention to this, and see what comes of it." --The communicative interaction will then become an exploration which (among other potentialities) tests my proposal. If, on the other hand, the person does not feel safe, and therefore decides not to speak up, at least they should be able to orient themselves more lucidly to how their situation is unsafe, and therefore be able more effectively to protect themselves.

What I am proposing supervisor and supervisee endeavor to do is: Whenever either person senses that maybe "something is going on" in their present communicative interaction, the person stops and calls both conversants' attention to it, and they examine the interaction in which they were a moment before engaged (along with the new turn the conversation has taken in examining that now suspended interaction). The "trigger" here will often be a conflict or uncomfortable feeling, but it need not be: the observation can also be about aspects of the constructively synergistic flow of interaction. Another "trigger" is precisely when "nothing unusual" seems to be going on, to try to explore what processes are transpiring under cover of un-event-fulness (the plenum of the customary – the "cultural unconscious," etc.). The main defect in the advice that the more obvious something is the more questionable it is likely to be is that it provides no clues how to detect the inconspicuous (I shall provide some examples – which I hope may function as "templates" and maybe even as models for making other templates – below).

Supervisor and supervisee can cultivate this attentiveness in their interaction, and cultivate their skills in cultivating this attentiveness (including by doing relevant reading and discussing this interactional process in other situations, such as peer study groups, professional association meetings, training institute classes, etc.). Furthermore, they can attend to remembering (including possibly taking notes for possible later use in lectures and publications) and reflecting on the vicissitudes of this process, to enrich their appreciation of it as process, i.e., as investment of energy the value of which may last beyond the possible disconfirmation of its occasional results, and exceed the value of those results even if their validity stands up "forever."

I do not expect the foregoing description will give the reader any sense of what I am talking about if the reader doesn't already have it. That is a compelling reason for the examples which follow, and for presenting them in what I hope will be an imaginatively evocative way. There is another reason for the examples, however, which is that – at least at my present point in the spiraling process of theory shaping experience testing theory... – I have more confidence in what I am trying to do (both in this dissertation and also in trying to live it in my own life) than I have in the theory in terms of which I interpret it.

The way I employ the examples is to present a vignette from a supervisory interaction as it in fact happened (as reported in the literature of theory of supervision), and then "rescript" it: that is, I have the participants in the interaction say something different from what they in fact said. The participants' new words are instances of what I wish to encourage supervisors and supervisees to explore in their relationship (and for persons to try in life in general). I hope that these rescripted interactions offer models which have value for persons to try out applying in their own communicative relationships, even if my theoretical arguments which I hope further support these proposals prove inadequate. Indeed, I think an excellent way to improve the arguments is by trying to actualize the kind of interactions presented.

I am not here trying to "weasel out" of defending my arguments. I am saying that, insofar as a latter-day Socrates tries to get me to trip over my own tongue and I find myself headed for logical self-contradiction, if I can confront my critic in person, I shall try to undercut my critic's method of argumentation and make them see they are engaged in producing a performative (a.k.a.: "existential") self-contradiction, before myself being reduced to paralyzed babble. Insofar as the reader already understands what I am trying to accomplish in this dissertation, he or she will see that this conversational tactic itself constitutes an instance of what I am here proposing, and would thus serve as evidence on the basis of which to try to recover my argumentative (theoretical) position vis-à-vis my hypothetical [and the applicability of this label is part of what I would investigate in the situation:] conversant. Furthermore, I propose this as a humanely constructive maneuver, because, to borrow a phrase from Paul Feyerabend [albeit with modification, but a modification of which I think he would approve]: it has the beauty of being ad hominem, in the sense of approaching [ad] the person [hominem], i.e., offering (not imposing upon!) him or her a potentially rewarding opportunity of social life (Feyerabend, 1987, p. 78).

Parallel Process and Interactional Process

There is a substantial literature on a phenomenon that takes place in supervision which may look similar to what I here am talking about: "parallel process." For the reader who is familiar with this concept, it may consequently be helpful if I briefly indicate how parallel process as generally understood differs from my focus of interest. Parallel process is an important phenomenon, and studying and making use of it are valuable, but it is not what I am here concerned with.

Parallel process is a phenomenon whereby the supervisee unwittingly shows the supervisor certain aspects of a patient's pathology by the supervisee unwittingly ("unconsciously") enacting similar (i.e.: "parallel") behaviors toward the supervisor. For example, a supervisee may exhibit obtuseness toward the supervisor's instruction not because, in general, the supervisee is obtuse, but because the supervisee's patient is being especially obtuse in therapy without the supervisee him or herself clearly recognizing this.

Such difficulties get resolved by the supervisee coming to see that the problems do not really arise from the supervisory relationship. Supervisor and supervisee mutually come to realize that their relationship has been "invaded" by the patient's material. Both may greet this realization with relief and even positive satisfaction, for now a problem concerning the patient's therapy has been clarified. Their relation – the relationship between supervisor and supervisee – can now return to its status quo ante, i.e., its "obvious" condition, unchanged. If supervisor and supervisee are inclined to literary allusions, they might see themselves as having come through a test, like Odysseus when he had himself lashed to the mast of his ship in order to be able to hear the song of the Sirens. Both participants are reassured that what at first looked like "their" problem really had nothing to do with them (i.e., with their self-identities).

The supervisor may sometimes judge that inexperience or personality problems in the student resulted in the situation escalating to parallel process instead of the student nipping it in the bud and correctly interpreting it in the therapy situation. In such cases, the student may suffer some damage to his or her self-esteem (e.g., if the supervisor tells the student to take it up in his or her personal analysis). Often, however, all parties concerned are satisfied that the patient was sufficiently "difficult" that the therapist, even an experienced one, would likely have needed help of something like the parallel process interaction which in fact dealt with the problem (e.g., in a peer supervision group). Whether or not the supervisee comes away from such an interaction feeling (or being instructed) he or she has some need to change, the supervisor is unlikely to have any such thoughts about him or herself, especially since the supervisor successfully managed the supervisory intervention which solved the problem.

This phenomenon surely has a place in supervision, but my focus here is on unwitting aspects of the supervisory interaction which are eventually resolved by recognizing their origin in the supervisory interaction – their, so to speak, iatrogenic nature.

For purposes of clarification, I have surely drawn too sharp a dichotomy here. There can be contributions from both sides in the same supervisory situation: the supervisee's (and supervisor's) interaction with each other can be influenced both by unwitting aspects of themselves and their interaction, as well as by what is transpiring in the supervisee's therapy with the patient which they are considering. But this is only a particular instance of the fact that anything going on in any part of a person's life can affect other aspects of his or her life.

The important distinction between what I am focusing on and what is usually called parallel process is that in the kind of interaction in which I am interested, the real relationship between supervisor and supervisee is changed as a result of their coming to see more clearly what is going on between themselves (due to whatever factors, including but not biased toward presumptive attribution to stresses in the supervisee's therapeutic work). Many of the dynamics which I am encouraging supervisor and supervisee to attend to would exist no matter who the supervisee's patients were, and would still exist even if the supervisee had no patients at all (which, of course, would altogether eliminate the possibility of parallel process since there would be no patient process to unwittingly reproduce in the supervision). Also, many of these dynamics would not exist with a characterologically different supervisor (not just a characterologically different supervisee) and/or a sociologically different organization of the psychotherapy "profession": they are genuinely inter-actional, context-dependent phenomena.

There is another supervisory phenomenon which warrants mention here because it too has affinity to that with which I am concerned. Unlike the previously described kind of parallel process in which the student functions like a telephone wire which cannot be blamed for the message it conveys, and which, consequently, if not deeply illuminating, is at least relatively benign, this other kind of parallel process puts the supervisee in the bad role of broadcasting trouble to everybody. It has much greater potential for being used to intimidate the supervisee (i.e., it is a method easily abused by bad supervisors). In its mildest variant, the supervisor may say to the supervisee something like: "You seem to be making the same mistake [or whatever] with me as you're making with your patient."

Such instruction, especially if "on target," can effectively cut through a supervisee's defenses. The problem with it is that it also cuts down the supervisee's self-esteem, and thereby raises other defenses, so that the result will likely be ambivalent. Its truth tends to get in the way of its effectiveness by failing to deal with the broader truth of the supervisee's total life situation qua recipient of the message. If the supervisor does empathically and sympathetically address and discuss these contextual issues with the supervisee, then their interaction may actually be an instance of what I propose; the less sensitive the supervisor is to the supervisee's experience of the criticism, the further the experience will diverge from this.

A bad supervisor, faced with an impasse in which the student refuses to obey him or her, may be able to get the student sufficiently frustrated that the student "loses his or her cool" (the supervisor may also interpret the student's stubbornness to the same effect, even if the student remains entirely calm and courteous). In such a case, the supervisor can employ a potentially devastating version of this second kind of "parallel process," which goes more like:

If you, supervisee, act this way with me, I can only wonder what bad things you do to your patients, which, of course, I can't be sure about, since I can't count on somebody with your serious problems to tell me what's going on.

At this point, the supervisee has only two choices: (1) to beg for mercy ("Please give me another chance to try to mend my ways") and hope the supervisor will not report him or her to the Student Evaluation Committee, or (2) to not beg for mercy and wait to see what the supervisor and the institute administration will do to him or her (e.g., expel the student from the institute; make the student go for more personal therapy; etc.). Here, the supervisor makes use of the epistemological difficulty in determining what actually happened in the therapy situation (due to its being an instance of an inaccessible referent of discourse) to produce grounds for speculating beyond the evidence at hand that things may be even worse than they seem. Here the supervisee has been reduced to an object of strategic communication concerning his or her disposition in that part of the social world in which the supervisor has role-defining power.

In brief, and to anticipate, the key difference between this kind of unilateral imputation of a problem to the supervisee (even in its non-abusive form) and what I propose, is that, in the proposed procedure, the supervisor asks himself and asks the supervisee (knowing that one can never count on the weaker person telling the stronger anything in consequence of which the weaker fears the stronger might hurt him or her):

Maybe this trouble that's happening here right now is all due to some problem in you, A. But, often, even when it seems like one person is the source of all the trouble, he's really just the bearer of the symptoms for a whole dysfunctional system. So I've got to stop and wonder, and ask you, if you will take the risk, to help me see how I may be contributing to making all this happen. The very fact of your being a student and me a supervisor, this asymmetrical relationship between adults in which you and I find ourselves, is itself an instigating factor, you know.

My Proposal is an Extension to Supervision of Something Already Familiar in Therapy ("Transference" Interpretations)

Actually, what I am proposing is an extension to supervision of a long familiar therapeutic notion, which is summarized in the following quote from Merton Gill's classic text on the subject of "transference." By a "transference" interpretation, is meant, essentially, an interpretation in which the therapist makes a point to the patient for which the patient's behavior in the present interaction with the therapist serves as evidence. An "extratransference" interpretation, on the other hand, has as its subject matter something outside the patient-therapist interaction. An instance of a transference interpretation would be if a patient shows up late for an appointment and the therapist proposes it is because the patient is angry with the therapist. An instance of an extra-transference interpretation would be if the patient reports having been late getting to a business meeting and the therapist proposes that was because the patient was angry with another participant in the meeting.

The paper analysts are most likely to think of in considering the centrality of the transference is Strachey's "The Nature of the Therapeutic Action of Psycho-Analysis" (1934). It is in this paper that Strachey introduces the phrase "mutative interpretation," meaning an interpretation that brings about a change. Strachey confronts the question directly by asking: "Is it to be understood that no extra-transference interpretation can set in motion the chain of events which I have suggested as being the essence of psycho-analytic therapy [i.e., insightfully motivated deep personality change]?" (p. 34). He replies in the affirmative, adding that one of the main intentions of his paper is to make explicit "the dynamic distinctions between transference and extra-transference interpretations."

Actually Strachey retreats a bit from this unequivocal position that only transference interpretations are mutative. Instead of saying that extra-transference interpretations cannot be mutative, he notes they are not usually "given at the point of urgency" (p. 34). Similarly, he states that, in extra-transference interpretation, it is not impossible but only more difficult for the patient to see clearly the distinction between what is real in the situation and what is fantasy. In these respects, then, "an extra-transference interpretation is liable to be both less effective and more risky than a transference one" (p. 34). (Gill, 1982, p. 52)

Everything in this quote, mutatis mutandis, applies to the point I am proposing to make for supervision: that the most pedagogically effective supervisory interactions are those in which the supervisory interaction itself exemplifies the lesson to be learned, i.e., where what is to be learned is the experience being lived through, and in which the evidence for it is self-evidence. The reason for the preeminent effectiveness of such interactions in both these (or any other) relationships can be expressed simply by such idioms as "being caught red-handed," "being caught with one's pants down," etc.

Strachey goes on to claim that the majority of the interpretations that are made are extra-transference interpretations. Yet he has some difficulty justifying this, presumably since he regards only transference interpretations as mutative. In order to clarify this imbalance, he resorts to an analogy of the battlefield. He sees "the acceptance of a transference interpretation" as "the capture of a key position." In contrast, "extra-transference interpretations correspond to the general advance and the consolidation of a fresh line which has been made possible by the capture of the key position" (p. 38). Just as in war the general advance is eventually met by a new "check," leading to "the capture of a further key position," so, too, in the analysis there is an "oscillation" between the consolidation of insight through extra-transference interpretation and new insight through transference interpretation. (Gill, 1982, pp. 52-3)

Again, all this applies to my position regarding self-reflection in supervision. I hope my examples will demonstrate the power of self-reflective interactions to make major breakthroughs in the supervisee's (and sometimes also the supervisor's) therapeutic orientation, and even their general orientation in life. But such breakthroughs need to be worked through and elaborated, including connecting them with whatever external phenomena. When, in supervision, only the external material from the therapy interactions is intentionally pursued, so that self-reflective insight is unlikely or, if it happens, its potential is not appreciated and therefore largely unrealized, then, just as in therapy which deals only with the patient's external life, the outcome is likely to be relatively ineffective and also more risky (of misunderstanding, etc.). When the lessons of the here-and-now supervisory interaction are vigorously pursued for self-reflective insight, then, as in therapy which focuses on so-called "transference" interpretations, not only are the insights and, more important: existential changes which result from these interventions themselves exceptionally fruitful, but they also make consideration of external material more rewarding. Thus, my proposal "includes" all the good techniques and attendant results of existing supervisory practice, but subsumes them in a new, richer and more productive context.

A question to which I do not have an answer is why nobody seems to have focused on this before. It seems "obvious" (which, of course, does not mean it "is" obvious! [--like Heidegger's example of the person looking for their eyeglasses and not being able to find them because they're on the person's nose and the person is looking through them as well as looking for them...]). In practice, such reflection does sometimes occur, often without being explicitly recognized, and, a fortiori, valued as such. Surely there are many times when a supervisor or supervisee may have an intimation that what they are saying about a patient "applies here too," and vice versa. But these interactions, left unattended to, are no more useful than Enrico Fermi's recollection (ref. lost) that he would have discovered nuclear fission ten years earlier than he did, had he paid more attention to a certain photographic plate from that time which, at the time.... [I find this phenomenon of retrospective recognition of the value of something one "missed" when it initially happened interesting. Clearly, one didn't altogether miss "it" at the time, or else one couldn't remember that one missed it. Yet, equally clearly, the person didn't "capitalize" [on] the experience at the time. Could this kind of self-interference be a result of traditional childrearing which vitiates the child's autochthony?]

Various authors writing on theory of supervision mention that the best way to address certain difficulties which occasionally arise in the supervisory relationship is for supervisor and student directly to talk out the issues involved. When it is intentionally undertaken, such reflection is often seen as something which has to be done in supervision instead of supervision when the learner has problems with supervision which interfere with his benefiting from supervision (see, e.g., Rubin, 1989). Therefore, one reason my proposal has not long since become commonplace may be therapists' (both supervisors and supervisees) narrowly functionalistic orientation to their work. Optimally, from a narrowly functionalistic perspective, one shouldn't have to engage in digressions from talking about the patient's therapy; next best is to attend to and get done with the precipitating problems as quickly as possible, to be able to turn attention back to "real supervision," focused directly on the student's work with his or her patients outside the supervisory situation. (This, of course, relates to the issue[s] I raised above, concerning the difference between a skilled craftsman's having a merely technical orientation in life, versus a more humanly rich "political" [in the classical Greek sense] appropriation of his or her skill for its contribution to social – i.e., conversational – living[, and also the issue of the process [of conversation] being of deeper and more enduring value than its product ["truths", etc.]].)

Another possible reason that the self-reflective aspect of the supervisory communicative interaction has not been studied is that the notion that psychotherapy is an application of communication theory is still not widespread, or, to be more precise: however far it has spread, its roots have not yet sunk deep – despite the work of Ruesch and Bateson, Thomas Szasz, Sullivan, [even] Langs and others.^[1] It is possible to theorize about communication ad infinitum without ever examining one's own communicative processes – neither the communication acts which constitute the theorizing, nor the theorist's communicative life as a member of society. In addition, so long as the "medical model" guides therapists' orientation in their world of work even when they deny it does, by their interpreting functional problems of living as "psychopathological" entities to be "treated" (nosology covering over noesis), therapists will continue to see communication primarily as a means of doing therapy, rather than as a primary, and an ultimate object of therapeutic concern. Living by such [deutero-]ideas will not likely lead a person to the notions proposed herein ("What you do not take the trouble to look for, you are unlikely to find" (Toulmin, 1990, p. 46)).

Some Basic Assumptions of My Proposal

My approach to supervision depends on the categorization of the supervisor's role as one in which power is wielded only as trusteeship, and only to the extent that the supervisee is still, as a psychotherapist, an "infant," in the etymological sense of one who cannot yet speak for him or herself. What I mean by this is that a supervisor can try my approach only if the supervisor chooses to use his or her power over the supervisee only in this facilitative way (and not, e.g., to assuage the suffering he or she underwent as a student by inflicting "parallel" suffering on his or her students, or to recruit disciples for his or her theoretical and/or professional-political position, etc.).

Conversely, for a supervisee to try out my approach requires the student to judge his or her supervisor according to whether the supervisor aims to help the supervisee find his or her own voice and unilaterally intervenes only – but then effectively! – when the supervisee is "at a loss for words" (i.e., unable to orient him or herself effectively in a situation).

My approach to supervision depends on seeing the supervisory process as truly interactional and conversational. For supervisors to use my approach requires their being willing to look closely at all aspects of the sociological structure of the supervisory relationship (including the involvement of the training institution) and to continually look for their own personal[ity] contributions to the interaction, including issues that might threaten their own self-esteem, in trying to understand what goes on between themselves and the supervisee. [The supervisor can interpret his or her advantage vis-à-vis the supervisee in terms of having a deeper cushion under him- or herself, rather than as raising him or her over the supervisee; and the supervisor can thus see his or her opportunity as being able to provide a cushion for the supervisee to fall on while the supervisee still lacks his or her own....] In order to foster the possibility of inter-action (as opposed to dominance and submission exchanges) with the supervisee, supervisors need freely to endeavor to compensate for their advantage, by imaginatively placing themselves in the supervisee's "shoes" and even speaking up against themselves when they think the supervisee might be afraid to:

I just said [whatever] to you, A. If my supervisor said that to me, I think I would be {angry|disappointed|hurt...} . But I also think I might be afraid to tell him. So you don't have to respond to what I've just said, but I'm letting you know how I think you might feel, and I'm apologizing to you anyway.

Conversely, for supervisees to use my approach requires them judging their supervisor in terms of how much the supervisor extends him or herself to help the supervisee overcome the asymmetry of their relationship, not by falsely pretending the two are in all ways equals, but by always taking the inequalities into consideration in such a way as to minimize both their extent and their negative impact on the supervisee. For one example, the supervisor knows the "tricks of the trade," and its "Gottcha!"s; the supervisor can certainly tell the supervisee, without making the latter feel "small":

You've just started in this work, and I suspect you probably have never encountered one of these [whatever] before. I may be wrong, but it looks like [such and such] may be about to happen with this patient [or with the training institute, etc.]. Be on the lookout for this, and, if it does happen, here's a way you might be able to save yourself some trouble [etc.]....

<a name="fneph">The foregoing are "enabling" conditions.</a> They are certainly not anything for which I claim originality. They are the way good parents and teachers have perhaps always related to their "charges" – at least within the constraints and distortions of their "cultural unconscious" which can make of any father an Abraham and any child an Isaac. Ferenczi wrote, in an essay evocatively titled "The Adaptation of the Family to the Child":

I am reminded of an incident with a little nephew of my own, whom I treated as leniently as, in my view, a psycho-analyst should. He took advantage of this and began to tease me, then wanted to beat me, and then to tease and beat me all the time. Psycho-analysis did not teach me to let him beat me ad infinitum, so I took him in my arms, holding him so that he was powerless to move, and said: "Now beat me if you can!" He tried, could not, called me names, said that he hated me; I replied: "All right, go on, you may feel these things and say these things against me, but you must not beat me." In the end he realized my advantage in strength and his equality in fantasy, and we became good friends. (Ferenczi, 1955, p. 75)

Another condition for using (or at least trying) my approach to supervision is not at all natural or obvious. Perhaps it could only derive, in any focused way, from the Husserlean idea of "phenomenological reduction." The various kinds of "parallel process" described above are cognates and adumbrations, but their employment in supervision does not seem to have led to anyone trying my approach. This condition is not enabling, but substantive: that supervisor and supervisee shall make it one of the enduring themes of their interaction to reflect upon that interaction with the aim of making their conversation account for itself within itself, i.e., that supervisor and supervisee shall strive to make their communicative interaction self-reflectively self-accountable. They will try to the greatest extent possible to make all the evidences employed in their interaction be self-evidences (or, stated the other way around: they will try to avail themselves as fully as possible of the evidences their interaction provides). They will try to learn from their own communicative interaction, as well as learning about externalities (e.g., the supervisee's therapeutic work) through it. They will try to formulate interpretations of their own interaction which, as messages occurring in that interaction, will constitute evidence for their own truth: performative speech acts whose manifest content is illuminating insight into the act's performative context, or, in Bateson's terminology, learning in which proto-learning and deutero-learning teach the same lesson.^[2]

Of course not all of supervision can be this kind of self-reflective, self-accountable and self-grounding conversation, even in its broader aspect of applying the insights gained from exploring the supervisory interaction to the wider social contexts of therapy (and life in general). Especially, case management issues will often be entirely "transcendent" (i.e., concerned with issues for which the validity grounds are somewhere other than in the supervisory conversation itself). Calling Child Protective Services concerning a client whose family appears to be abusing him or her (obviously) need have no correlation with the qualities of interaction between the supervisee who makes the call and the supervisor who advises the supervisee to do it. It was precisely in order to be able to exclude from consideration all these aspects of supervision which are not tightly relevant to (or at least not highly revelatory of) the notion of "self-accountable conversation" that, above, I carefully delimited what I would be considering, namely, supervision of psychodynamic therapy in the narrow sense, i.e., the vicissitudes of meanings which are the "dynamics" of the "psyche," and which are immanent, in the patient-therapist interaction in psychotherapy, and in the supervisee-supervisor interaction in supervision.

I shall now offer examples^[3] in which I hope the reader will be able to see how and also be inspired to want to operationalize this proposal.