Language Usage and Social Action in the Psychoanalytic Encounter: Discourse Analysis of a Therapy Session Fragment

Introduction

With the accelerating breakdown of the hegemony of classical Freudian thought in the psychoanalytic world over the past three to four decades, post-modernist relativism rapidly became a cliché in psychoanalytic circles. No school could any longer lay claim to a monopoly on truth. Indeed, the notion of truth itself was been rendered highly problematic. The anti-realist views of philosophers of science such as Kuhn (1970) penetrated the psychoanalytic world via the work of psycho-analytic writers such as Spence (1982). In the latter's discussion of “narrative truth”, a correspondence theory of truth was abandoned in favour of a coherence view of personal narrative. Through the analytic process, the patient's initial narrative is, gradually replaced by a new, presumably more coherent or useful narrative negotiated with the therapist. However, as Gergen and Kaye (1991) pointed out, while this view may be a useful description of analytic practice, there are several problems with it as a description of or a prescription for the therapeutic process. This conceptualization still privileges the analyst's narrative account of the patient's experience over the patient's initial narrative account. The analyst still seems to have a pre-existing “truth”, albeit in quotes, a privileged story that s/he imposes on the “material” the patient brings. In the work of many narrativists, the view of the therapist as possessing a special knowledge of souls has still not been superseded.

Moreover, while the question of truth has been relativized, the problem of truth has been preserved in a narrativist framework that still relies heavily on a notion of representation.

What analyst and analysand are said to be doing is constructing better stories, i.e., better pictures or representations of something “out there”. To this conception, Gergen and Kaye posed a different view of therapy as enacting and creating meanings through dialogic acts. Psychoanalytic writers, such as Hoffman (1991), also went beyond relativism to the position of social constructionism. This view tries not to privilege the analyst's meanings but rather sees the meanings arrived at by both participants in the therapeutic dialogue as mutually and interactively constituted. Such a position must necessarily force a reconceptualization of the nature of the therapist's activity and expertise. If therapy is not simply the replacement of one narrative with another, what else is it? The trend of thinking in Hoffman's work, and in the work of non-psychoanalytic social constructionists such as Gergen, is to see the therapeutic situation as an arena of (potentially) creative dialogic enactments.

In what follows, I will attempt to address the question of the grammar and vocabulary of some of these enactments. Some very interesting prior work in this area is collected in Siegfried (1995). Most of the work in this volume is by non-explicitly psychoanalytic writers (one notable exception being M. Horowitz). A more recent collection of work on conversational analysis of psychotherapy edited by Peräkylä, Antaki, Vehviläinen and Leudar (2008), building on the pioneering work in this area by Labov and Fanschel (1977), analyses patient/therapist turn-taking sequences. Several of the authors in this book examine dialogic specimens of how therapists use linguistic strategies such as substituting single words or short phrases (i.e., Rae) or reformulations (i.e., Antaki) as a means of introducing new meanings or extending patients' meanings in what the therapist deems therapeutically/emotionally useful ways. Peräkylä in this same volume discusses intersubjectivity by means of analysing how patients signal various levels of agreement or disagreement with therapist's interpretations. Streeck (also in this 2008 collection) discusses enactment but principally from the non-verbal gestural side. Of course, much has been written about enactment from a psychoanalytic/relational point of view.

Tilly's (2006) sociological work on reason-giving for action is an interesting discussion of how reason-giving functions in the negotiation of roles and relationships, one very important subset of meanings. In the case example I give below, the therapist (myself) asks a series of questions. Questions on the face of it are seeking answers - reasons for action, explanations or accounts of motives, statements of rules of social behaviour, and so forth. In this sample of dialogue, role negotiation is a salient feature as is the attempt to make possibly implicit emotional meanings explicit.

As Hoffman has noted (1992), there are limits to the range of meanings that the members of the therapeutic dyad can co-produce. Neither participant just “decides” what to say, what move to make. Rather, in Hoffman's terms: “…among the limiting factors are the temperaments and resources of the participants as well as the analyst's and patient's unconscious interest in particular kinds of interaction and associated unconscious resistance to other forms (p. 294)”.

To these considerations I would add that the means for co-producing meaning are also socio-culturally limited. This is to say that in all our conversations, including the therapeutic conversation, we are following rules not necessarily of our own devising. We are playing “language games” as Wittgenstein (1953) first pointed out, which are in turn embedded in social “games”, practices, or, as Wittgenstein termed them, “forms of life”. It is with these tools that we socially generate meaning. Therapist and patient are at times

agreeing and at other times disagreeing about what it is they are doing, and much of this shared or divergent understanding of what they are about is unformulated. This, I think, is necessarily so for two reasons. First, there is the high speed of the analytic exchange, which, in this respect, is no different from many other forms of social encounter. There is no way we could become conscious, in the moment, of all the language games we are playing. But, secondly, embedded as we are within these forms of life, it is difficult for us to be aware of them as such. There is always some “horizon” of experience as Merleau-Ponty termed it (1962), which is difficult if not impossible to transcend or even comprehend. In part, this is because the rules of social and linguistic practice by means of which we constitute our worlds lie right under our noses. Thus, we take no note of them, or, taking note, we view them as so general and obvious as to seem trivial. We always can, but do not usually ask ourselves how we know what a speaker means by what he or she says. This field of inquiry is part of the discipline of speech pragmatics, which includes the study of linguistic forms, of reference, presupposition, types of speech act, relation of utterance to linguistic and social/action context, rules of turn-taking in dialogic exchange, and so forth. Although the formalized study of speech pragmatics is a technical specialized field spanning aspects of linguistics, philosophy (Austin, 1962), psychology, and sociology; the intuitive, implicit use of the rules and practical procedures of speaking and of understanding what another is saying is something that as naive laypersons we do every day. We do not have to be students of speech pragmatics to function in everyday social life. However, I would suggest that, as therapists, we are enhanced in our functioning by the more detailed understanding of dialogic process that inquiry into speech pragmatics provides. Indeed, knowledge of or, better, sensitivity to the subtleties of dialogic process could be said to constitute an essential aspect of our expertise.

But also, as noted above, linguistic practice is embedded in, if partially constitutive of, social practice. The forms of social practice include but go beyond the linguistic. If we ask the question of how we know what a person is doing extra-linguistically (in the sense of social action) by saying something, then we are starting to tread on somewhat broader but shiftier ground. We are entering the field where we can begin to talk about the underlying rules of communicative aims, i.e., the grammar of motives and purposes, intentional, quasi-intentional, or so deeply embedded as to be not easily formulable. This field includes the area of psychodynamic theory which we, as therapists, make use of to varying extents and in a more or less systematic way. It also includes the common-sense psychological and sociological theories of motivation that we all, therapists and laypersons alike, make use of, however consciously or unconsciously, in daily life. As applied to therapy, we might call this the field of receptive/passive or interpretive therapeutic speech pragmatics, insofar as we, at any given point, make use of a more or less systematic theory of social action (of which the dynamic metapsychologies are a subset) to understand the (speech) actions of each member of the therapeutic dyad. Finally, there is the question of what we do with our understandings of what we and the other are saying/doing in the therapy setting, however we arrive at these understandings. These involve questions of therapeutic aim and technique or, if we narrow it to linguistic activity, active or productive therapeutic speech pragmatics. Thus, there are at least three distinguishable levels of language use that come into play in therapeutic practice. First, are the everyday rules of language that enable basic communication; then there are the higher order social and psychological inferences about meaning and intention (which includes affect); and, finally, there are our conceptions of what we do as therapists with the meanings developed in the analytic encounter.

It might seem intuitively evident that these three levels of language use and understanding are hierarchically ordered. That is to say, one might suppose that the common-sense pragmatics of the rules of discourse are the basis of everyday understanding upon which the more arcane psychodynamic type of understanding is founded and that the theory of technique flows out of, or is entailed by, this latter kind of understanding. However, I would argue that no such simple hierarchy of knowledge of language use obtains in the therapy situation. Rather, we often jump from one level to another without acknowledging it to ourselves. This is to say that not everything we say or do in psychoanalysis is motivated out of the concerns of psychoanalytic theory. Rather, much of what we say and do as therapists is directly conditioned by culturally received linguistic and social common sense.

Of course, sometimes our psychological theory (our subcultural “common sense”) moves us to override the general common sense. Moreover, and more importantly, theoretical and common- sense understandings and action are not always conceivable as two distinct moments in the therapeutic encounter. Rather, what we do as speakers in our culture shapes our understanding of what we are doing as therapists. A parallel situation obtains for the patient. His/her understanding of what s/he is doing qua patient takes place against a background of implicit everyday social communicative action.

Also, and importantly, understandings of the situation do not arise in an interpersonal vacuum but are always created under the influence (including implicit and explicit demands and instructions) of the other. These understandings then constitute an interpersonal field of meanings beyond the individual's control, an “intertext” in Bakhtin's terminology (Todorov, 1984). The intertext is a constantly developing organic process and product. As the dialogic interchange unfolds over time new possibilities and choice points of meaning and action open up. These choice points are complexly contingent on many factors, among which our awareness of choices and possibilities is, I believe, primary. Our awareness of choices depends on a subtle textual and textural sensitivity to the communicative actions, i.e., the gestures, moves, meanings, language games, and social beliefs and expectations of ourselves and the other. In order to demonstrate some of these nuances and possibilities of language use and meaning in the therapy situation, I will turn to an actual example of therapeutic discourse and interaction.

Case Study

A woman in her early 30's has been in therapy for seven years through many life vicissitudes including a series of conflicted and often self-destructive love relationships, a long period of dissatisfaction with her work and the launching of a new career path, and many battles in an ongoing struggle to separate herself from her family of origin. In the year preceding the session I am about to describe, she had come to be in a basically satisfying relationship with a man, had gotten started in graduate school and been able to distance herself more than previously from the troubling influence of her family. She came to the session, having spoken in recent weeks about heading towards termination. Toward the end of the session, the following exchange ensued.

In reviewing this and other segments of therapeutic dialogue, I have recurrently been struck by how complex the therapy interaction can be when considered at any relatively fine level of detail. One can analyze what is happening in terms of the theory of psychoanalytic technique (relative to certain standard notions of therapeutic aim), at the level of conscious, pre-conscious, or unconscious motive and affect of both actors (often summarized as transference and countertransference), at the level of social action considered from a sociological point of view (e.g., assertions of the speakers as to their respective statuses, roles, and competences in these roles) and at the level of discourse, linguistically considered (what are the speech acts involved and the sequencing rules for these that are being followed in the discourse). If you really get into it, there is too much to say. So for the sake of staying focused, I will try to anchor the discussion on some linguistic features of the dialogue.

First, the patient's initial “I know that…” locution functions, to use a term of speech pragmatic theory, as a presupposition or propositional trigger. “I know that” embeds the proposition “I have to leave”, thus suggesting that the latter is a declarative proposition which could be true or false. But, this is already complex and contradictory because the sub-predicate “have to” is not, strictly speaking, declarative in its form. It is more like an imperative, a command given to oneself. To use a term first introduced by the philosopher, Austin (1962), its illocutionary force, i.e., what kind of speech act it is, is ambiguous. It trades on other uses of “have to” which seem closer to reports of states of affairs rather than expressions of mere want or obligation. So, perhaps, we might

understand the speaker to be saying that she feels she ought to leave. This would count as a statement with a truth value - either she does or does not feel she ought to leave. But this is our gloss on her text, and certainly not the only one warrantable by what she said. What seems evident, just from the form of her words themselves, leaving aside any higher order psychological theory, is that the speaker or rather the speech means more than she or it says. Thus, we are compelled to examine connotations of key words and how these connotations condition each other. Making no pretention at exhaustiveness, I will list a few senses of “know” that might apply here: “feel compelled to”; “resignedly accept (that)” (this is perhaps communicated by the preceding “sigh”); “believe that you think I should”; “hope that you think I should (or should not)”; “want to”; etc. Also, various combinations of these seem possible such as, “believe-that-you-think-I should-plus-I-don't-want-to”; or “hope-you-think-I-should-plus-I-want-to”; (or, alternatively, “plus-I-don't-want-to”). Any connotation or combination of connotations will carry a slightly different motivational force, capturing different issues in how the speaker feels about herself, about the therapist, what she wants the therapist to feel or desire in the situation, what she wants the therapist to believe she feels or desires in the situation, and so forth. The question of which reading is most accurate or useful will be conditioned by many things, including one's own theoretical predilections and the patient's own self-understanding, whether theoretically informed or not, as best as one can discern it. And, of course, one's choice of reading will partly condition one's response, and one's response as therapist will retroactively confer a particular meaning or set of meanings on the patient's initial utterance. Thus, interpretive pragmatics and productive pragmatics are inevitably intertwined. One cannot have an action orientation in a situation without some reading of the situation's significance, and, vice-versa, one cannot have any understanding of the situation outside of all possible action tendencies or intentions. In short, we are always choosing our meanings even as we, seemingly involuntarily, create them.

But, I want here to advance two simple claims. First, you don't need a very developed theory to tell that something of interest is going on in attending to the utterance that we have been discussing. You can tell that the words mean more than they say by paying a little attention to how they are being used and in what tone they are being said. The point of developing a sensitivity to, say, ambiguous illocutionary forces is not to be able to sit there and say, “Aha! An ambiguous illocutionary force!” Rather, it is to sharpen our linguistic and social alertness to the many possible meanings and action possibilities of the situation without a lot of theoretical prejudice about what these might be. For example, a knowledge of the different illocutionary forces of utterances in the therapeutic situation that are ostensibly functioning as questions (coming from either the patient or the therapist) can help us become more aware of the many dimensions of the processes we are engaged in with our patients (Schneider, 1991).

Second, sometimes a fairly content-less awareness of possible subtext based on linguistic sensitivity alone can enable the analyst to question or comment on the patient's utterance as a tool to get a therapeutic investigation and/or interaction going. In the above sample, the opening sigh tips us off to surplus meaning. It seems to say that the patient has been engaged in some internal struggle. Thus, apart from explicit meanings, verbal quirks in tone and grammatical form, departures from normal discourse rules, truncated utterances, etc. can be clues to important matters and prompts to therapeutic inquiry. Such commenting and questioning was what I tried at first in the above exchange. However, I quickly deviated from a close following of the patient's “text.” When she said, “I guess

it's time for me to leave”, I asked what made her guess, but I did not let her answer the question. Instead, I moved in with a pronouncement, “It's something you decide”. What might have been going on here was a counter-transference reaction. I believe I heard the patient saying that she did not feel “cured” but rather exhausted, out of meanings, and doubtful about whether further progress could be made in the therapy. Since this was damaging to my therapeutic self-esteem, I unconsciously diverted the exchange onto an existential point about taking responsibility for one's decisions. Valid as this may be, it may not have been the most important issue at that time. Thus, out of counter-transferential anxiety, I allowed one of my theoretical attitudes to intrude on the flow of meanings.

But, what one might ask was this counter-transference or theoretical position about? What was at stake here? The patient's reluctance to assume full responsibility for the decision to terminate is not merely an expression of anxiety or lack of autonomy on her part. It is also, and perhaps just as importantly, a construction of our respective roles and, more generally, our emotional significances to each other.

She is playing by the social rule of therapist as expert. In this case, my presumed expertise is to know when she is finished with therapy, i.e., “cured” or “cured enough”. It is this kind of expertise that I want to disown. The reasons for this would require a long discussion best reserved for another place, but let it suffice here to say that they involve values that are in some broad sense political. However, there is no way for me to escape from all attributions of expertise nor is it at all clear from the exchange that I want to do so. In saying, “It's up to you”, I am also asserting an expertise, to wit, my professional opinion that it would be best (healthiest?) for her to feel she can take on this responsibility. There is no way for me to avoid propounding a rule about our respective roles and attendant jobs and obligations. By calling oneself a therapist and setting up as one to whom another can come for help, one is necessarily claiming some kind of expertise, just as the other, in coming for help, is attributing expertise. Once one has chosen this role and meaning, many other meanings may follow, some of which may be unwanted by one or both parties and can become a matter of conflict or negotiation.

I do not know how many colleagues share my particular discomfort with the role of therapist as expert judge of another's mental health. But, I would venture to guess that at some time or other everyone in the field has experienced some role attribution that they want to disclaim. An interesting situation often occurs when a meaning (such as a role attribution) gets generated that neither party seems to want. This is the result of social and linguistic rules operating partly out of the control of either participant. It may be the case that something like this phenomenon was operative in the situation under examination here. Perhaps the patient was asking for my opinion about her leaving or staying not as an “expert” but merely as someone who is in a position to know because he knows her well. However, given that she is paying for her sessions and has come to me as one who presumably has some special skills for helping her, any request she directs toward me will take on the overtone of a request for my exercising professional knowledge. Everything she does or I do falls within this “contract”, as it were.

But, to return to the analysis of the dialogue, this part of the encounter ended at the point where the patient said “yeah, so I can decide…that's all…” and then paused. This pause constituted a choice point for the therapist to speak or not, but it was a strange choice point. Insofar as the patient ends her statement on a note of finality, it seems as if she

were saying that there is nothing to add. Perhaps she is saying, “OK, You shut me up”. But, in so far as the long pause seems to indicate that she is ceding her turn to speak, it seems I am being invited to say something. Or, perhaps, I am being both invited and disinvited to say something here. Again, you do not need a lot of psychological theory to see that this is the case. It just requires some sensitivity to the normal nuances of turn-signalling in conversation.2 I opted for saying nothing, in effect invoking a psychotherapeutic rule at this point, to wit, the therapist is allowed to remain silent even if “normal” socio-linguistic rules call for a response. I did this because the only thing I could think of to say looked to me like it could be received as having a presuppositional and prosecutory edge to it, i.e., something to the effect of asking her why she wanted to disown responsibility for her decision.3 This, I feared, might have destroyed any possibility for mutuality and collaborative investigation at this point. (By the way, I think that it is useful for us to think about the specific reasons for silence at any particular juncture rather than follow the blanket analytic rule.) As it happens, the patient had plenty to say on the subject of her difficulty in deciding whether to leave, which is what took up the next section of the dialogue. One might guess that she felt enough freedom or urgency about her own agenda, however conscious or unconscious, to proceed in spite of my previous intrusion. (It may be that my silence indicated to her some willingness on my part to let her concerns resume a center-stage position.) I use the word “agenda” advisedly here because I think it is a propos given the indirection of her remarks. She starts off with a topic statement about how hard It is to say goodbye, and then launches into a story about an analogous situation about saying “good-bye” to someone else, Ed, from the group (for which I was also the therapist). One theme seems clear from this story, although she never makes it completely explicit. That is, she has some kind of fond attachment to Ed and, by implication, to me that makes it difficult to separate. However, she does not draw out the analogy or the theme of attachment specifically. Instead, she leaves a space for this. I feel called upon to fill it, and I do with my series of questions about whether she finds it hard to say good-bye to me. Why do I do this here, i.e., say for her what she seems to find hard to say? There are a number of reasons, none of which were pre-calculated strategically on my part. Basically, her message that she has some fond attachment to me makes me feel good, and on some level, not completely consciously, I want to let her know that I know that. My acknowledging the feeling says that I like it and that I reciprocate it.

Her following statement “It trivializes…” takes up an aspect of this theme and has many possible meanings of which I selected one with my comment to the effect that perhaps she thinks our ending our sessions together is trivial to me. That statement advances a psychological hypothesis that is not arrived at strictly by attending to language. There is a linguistic cue, of course, in her previous statement. There is an ambiguity of reference in the phrase, “It trivializes”. The “it” seems to refer to the disembodied act of saying goodbye,

but it could also include reference to the speaker and/or to the listener.4 One might say that the “theory” or interpretation was a latent possibility available to the therapist and was triggered by the patient's saying “It trivializes…”. Her form of words gave me an opportunity to move in with this particular meaning, and I did so by pinning down the referent of the “it” in her statement to me, i.e., “you think (fear) it's trivial to me”. I made the choice to advance my hypothesis out of considerations again not completely explicit to myself at the time. Most therapeutic exchanges, like most social Interactions, take place at high speed. We do not have time to scan our assumptions and strategies. Just as an improvising musician does not have time to consult his built-in “knowledge”, i.e., the principles of technique or harmony, so, most often, we do not act from spelled-out strategic considerations. Rather, armed with our technique and our array of implicit theories, we take what the situation seems to give us, just as the jazz soloist improvises at the moment based on what he hears in the voicings laid down by the accompanist. One dynamic formulation of what is happening here is that the patient is expressing a fear that the therapist will get the message that the patient is unaffected by the termination of therapy. Perhaps, he will be hurt and/or not understand his own importance to the patient.

However, perhaps she is also expressing a fear that the leave-taking is trivial to the therapist. Maybe, she is expressing a dim awareness of the possibility that the therapist will defensively trivialize the relationship in his own mind if he hears her minimizing its importance to her. Further, maybe anticipating that the relationship is less important to the listener (Ed or me) than it is to her, she defensively trivializes it in her own mind and then fears the interlocutor may hear her trivializing it and may become hurt, etc. and so forth. So, in making one aspect of all this explicit, namely her fear that the relationship is trivial to me, I am perhaps dealing in many other resonances, but primarily I am reassuring her that it is not trivial to me. Was I aware of all of this at the time? Yes and no. It was a spontaneous decision on my part, not calculated but not uncalculated. But, could it be said I was choosing a meaning? Clearly, I was. Was I also choosing all the corollary meanings even though I was not explicitly aware of them? I am inclined to say that I was. But how could this be? How can one choose what one does not know? As therapists, I would submit that we do just this; that, paradoxically, we are responsible for what we do not know and, therefore, we choose what in commonsense terms we cannot choose, because it is our job to know that meanings are complexly numerous if not endless. I might add here that her drawn out “yeahh” followed by a smile and the comment “isn't that absurd?” and the more familiar “so, I guess it's time to leave, huh?” all could be taken as saying “OK, I got what I wanted (needed). I mean something to you, I know it and own it and that means I'm cured”. Note that the patient here is implicitly playing by the rules of psychoanalytic attachment theory, even more so than the therapist who keeps avoiding special rules, reasons, and psychoanalytic theories of mental health by insisting “It's up to you”.

We might ask if there are any other rules or assumptions, outside of specifically psychodynamic formulations, that either or both players are following in this exchange that would enable one to generate any of the various nuances of meaning just mentioned. I believe one such is instantiated here, and it may well be a common “rule” of social interaction in an egalitarian culture such as ours. It runs somewhat as follows: A believes that B's evaluation of A's significance to B will have some strong relation to A's

evaluation of B's significance to A. This is the sort of cultural given that I referred to previously as seemingly too trivial to bear mention, but does exert an important pull on our social activity. If we become aware of such a rule, we will thereby gain another small measure of awareness and consequent range of function as therapists. This rule is operative in the above segment, the patient acknowledges my interpretation as accurate “I know that's absurd”, and in her smile indicates she has received my implicit communication that the relationship is not trivial to me. Her next statement is interesting in the light of this last exchange. I can discern at least two messages in it. In saying, “So, I guess…” she is making an implicit connection between the leave taking issue and the question of our significance to each other (most saliently at this point, perhaps, her significance to me). The connective “so” is the tip- off here. But what is the nature of the connection between the question of termination and our significance to each other?

One subtextual reading is that the patient realizes that with our mutual admission that we value each other, a key dynamic issue of hers is thereby resolved. She can accept that she is a somebody. But, this is to attribute a psychodynamic meaning to her that she may or may not have. Another way to read it is to hear her as saying that now that this issue of our mutual significance is cleared up, she wants to know from me if she is truly “done”.

The first reading puts us on equal footing more or less. At least, it does not call attention to any role asymmetry. The second reading operates within a framework of client and expert. She is asking for my professional judgment as to whether she is ready to leave. My response, “It's up to you. It's for you to decide” was clearly to this second subtext. One might say that my response was counter-transferential in what it did and did not respond to. Uncomfortable for both ideological and characterological reasons with the role of the expert judge of another's mental health, I made a somewhat moralistic disclaimer. Perhaps “bad faith” would be a better term than moralistic because it contained an instruction to her not to see me as an “expert”, but paradoxically, this instruction was delivered from an expert's point of view. That is, I was in effect saying that it was my professional judgment that it would be best for her not to look to me as the expert judge of her degree of mental health. But, perhaps, apart from the specific content around the question of expertise, I was also counter-transferentially unwilling to acknowledge my importance to the patient and hers to me as central to the therapy. Here, in choosing one meaning, I was not aware of the meaning(s) I was not choosing to respond to. But, despite this lack of awareness, I would maintain that I was making a choice.5

With her next statement “But, it's also…there are so many things to work on still…” it seems that she is trying to convince me or herself or both of the supposed necessity of continuing, which she has already pre-judged. In particular, her statement that she has a long way to go seems to invite comment, seems to be invoking a “soft” conversational turn-taking rule, one perhaps especially germane to the therapy situation. That is, if A (the patient) says something and is convinced that B (the therapist) has an opinion about it, A's long pause means that B is now obliged or challenged to confirm or disconfirm A's statement. But, at this point, I play naive by refusing to confirm or disconfirm her statement “I have a long way to go”. Instead, I ask her about her internal state or what she's thinking, i.e., “What's up?”. This is a violation of the above-said soft rule of normal

conversation. The reasons for doing this are a matter of therapeutic principle or ideology. I do not want to take responsibility for her decision to stay or leave by positing some absolute standard of mental health (How far she has to go). In several sessions immediately preceding this, she had raised the question of termination, and each time I had remarked that it seemed to me that she had done a lot of work, but she could stay or leave as she wished. On several occasions when she had brought up outstanding problems, I confirmed for her that these were things she was still working on, and that we all have things that we are still working on.

On the evidence of the present session, it is hard to determine just what she heard in all this. Insofar as I agreed with her that there were things she was still working on, she might have heard me as suggesting that she ought to stay. Insofar as I am saying that we all are still working on things (presumably including me, that expert and paragon of mental health), she might have heard me giving her diploma. All of these meanings are close to the surface, as it were, of our exchanges. They can be easily inferred from the form of words that we have been using. A more theoretically derived dynamic understanding of her motivation is that as she becomes able to see more about herself and becomes more active in the work, she gets afraid of her own progress and the eventual separation from me that this seems to imply. She then has two responses to this conflict. One is to seal over and say everything is hunky-dory or as hunky-dory as it is going to get. The other response is to re-pathologize herself and re-submerge herself in the relationship with me. When I pulled the analytic move by my silent refusal to respond to her statement/question/challenge about her having a long way to go, I believe she got a little angry and compressed these two responses of hers. She said in reply to my question about what she was thinking that she was “tired of going over the same stuff”.

This seemingly simple statement was actually quite complex. The “same stuff” that she is referring to is indefinite in its reference. It could mean the topics of the last few sessions, or everything we have been talking about from the beginning. In calling it “the same”, the patient seems to be denying that the way she has been talking about it with me is deepening her understanding or helping her to change. Thus, “the same” has the force of an accusation or criticism of me. “You're not doing your job” she seems to be saying “You're not helping me go deeper.” It also seems to be a self-attack, a claim that she is still as screwed-up as ever and not getting anywhere. This take on what she is saying here derives not so much from dynamic metapsychology (although theory makes its contribution) as it does from a speech pragmatic reading of likely emotional responses to my refusal of her elicitation, that is the invitation to comment on her previous statement of how far she has to go. Among the many things at issue here is what Labov and Fanshel (1977), in their book on therapeutic discourse, call role strain and challenges to role competence. My silence is an instruction, to wit: You're supposed to figure this out for yourself. Her challenge is: Oh, yeah? You're not helping me like you're supposed to.

But none of this exchange is stark or openly angry. It is mitigated by tone. She has a tone of self-investigation, self-report, simple statement of fact. My “Whyzat” response is slightly humorous, seemingly friendly and curious. We are both playing down the conflict, preserving our own and each other's “face.” However, when she said it feels like she has not made much progress (seemingly taking responsibility for this herself but leaving the possible meaning that I have been negligent or incompetent), I became uncomfortable. I did not want it to be agreed upon between us that the therapy was running down and ending in a failure or only a very limited success. For one thing, I did

not think this was so; for another, my therapeutic self-esteem was offended; and I also had the feeling that she was fishing for me to say the opposite. And I did that with my “Weeeelll…I don't know…” statement, which, by the way, has a little trick at the end when I say “I don't know if you can really say that”. This is a pseudo-dispassionate disavowal of my self-interest in her not thinking this (which, interestingly, I don't really think she thinks). By saying “I don't know if you can say that” instead of saying “I don't know if I would say that” I am construing her statement as not being all that I hear it to be, in particular an attempt to get me personally involved in the question of whether or not she has made progress.

Finally, one last feature of the text calls for some comment. At the end when she says “OK” and “See you next week”, I was experiencing some tension. I was acutely aware, whether erroneously or not, that what I said in response to these remarks could be taken as tacit approval of her desire to stay and/or my wanting her to decide to stay in therapy. Thus, I took great care to restrict the scope of my remarks to next week only (“See you then”), and was especially careful to make my “OK”, echoing her “OK”, completely neutral in tone, almost satirically so, such that it could not be taken to mean “OK, I'm glad you're staying” or “OK, I agree with your decision to stay”. It is, of course, not clear that her “OK” was saying this. (In fact, we ended the therapy by mutual agreement two months later at the time of my and her summer vacation,) Rather, her “OK” may have been a response to my speech about no shoulds, something to the effect of “OK, I'll stop trying to palm this decision off on you” or perhaps “OK, alright already, stop preaching” or maybe “OK, no hard feelings”. Here the “OK” has an ambiguous reference. However, it seemed to me from the whole tenor of (the emotional/pragmatic/semantic drift of) the session, that she could just as easily be saying “Ok, you've convinced me to stay”, which is an action and motive that I wanted to disavow. Certainly, her statement about her chronic self-doubt, and need for and despair of her mother's approval seemed to be saying that although she had made progress, there was still significant work to be done. However, I was leery of prejudicing the case about how she could best make further progress. To that end, I strove mightily, one might say, a little forcedly or clumsily, in my speech about graduation, and in my terse, one might say, tense, final utterances, to present a neutral attitude about her staying or terminating, which, by the way, to the best of my self-knowledge, I actually felt. So, why all the striving to demonstrate neutrality?

I think, as I started to indicate above, the striving or protesting too much was an attempt to counteract lines of semantic force which seemed to be set up by our whole exchange to that point, to wit, that we agree that she should stay. Here is the power of language and the discourse taking a hold of both participants and potentially building a false consensus, a case of intertextuality run amok, outside of the participants' control or intention. The remark about missing her was also an attempt to correct something that seemed like it might be suggested by the whole preceding exchange. Namely, that I am the doctor without memory or desire dispassionately revealing to her a pre-existing truth that she is not in control of. This is a position which I find personally dishonest and dangerous.

Having presented a fairly detailed analysis, although not nearly as detailed as it could have been, of aspects of linguistic style in the therapy situation and their relation to the formulation of dynamic hypotheses, I want to summarize some of the linguistic/dialogic principles at play. First, linguistic dialogical/discoursal cues are our early warning system for knowing when something of psychodynamic and relational significance is going on. For example, conversational openings as revealed through the rules of sequencing and

turn-taking bear a functional relationship of unspecified complexity to therapeutic openings, i.e., opportunities for intervention of various kinds. Simply to codify a set of criteria as to what constitutes an opening would be a whole research project, let alone the attempt to characterize the relationship of conversational to therapeutic openings.

Second, if we want to hear all the possibilities of what our patients and we ourselves are saying and doing we must pay attention to the force(s) of utterances, that is, what kind of speech act a given utterance is, be it a statement, question, command, plea, entreaty, warning, promise, etc., plus the huge category of ambiguous or, as the linguists say, moodless utterances. There is a problem of over-simplification in this because it may be the case that the utterance is not the proper unit of analysis for the purpose of understanding therapeutic dialogue. This is because one utterance may have many forces, or different segments of an utterance may have different forces. And besides the question of illocutionary force, i.e., what kind of speech act is involved, there is the question of perlocutionary force, that is, the (consciously or unconsciously) intended and/or probable effect of the act on the listener which may not follow the divisions of utterance, but may rather be divided up along segments of utterances, distributed across utterances or be complexly embedded or nested in utterances. We are often, I would say, necessarily, doing many things at once and/or cumulatively over time. The above sample emerges out of the background of a long therapy that dealt with this patient's attachment to a disapproving parent, her consequent tendency to get involved in relationships (principally with male lovers) in which she was the subordinate party and another consequent tendency of persistent self-doubt about her life choices and competence. From my side, in the segment in which she doubts her progress in therapy, I can see retrospectively that I was responding counter-transferentially to being set up as the disapproving, judging parent (lover?). This response goes beyond the immediate dialogic, linguistic aspects of the exchange. In the whole emotional context of the relationship the words function not merely as bearers of quasi-propositional meaning but as emotional gestures, as actions.

It is important to remind ourselves that we can never have complete control of the consequences or construals of our actions, linguistic and para-linguistic. Because of the many meanings available to participants in an exchange, and the many different social and linguistic rules that come into play, there is a resultant opacity to intention, a kind of dialogic “practico-inerte” to use Sartre's (1963) phrase. Whether or not this is more prevalent in the therapeutic encounter than in other situations, it behooves us as therapists to have a healthy respect for the extra-intentional (which, incidentally, I do not mean as coterminous with the “unconscious”).

If we become aware of the contingency of meaning in the practico-inerte of interaction, that is to say, if we are cognizant of meaning's resistance to our control, then the activity of the therapist becomes the continuous attempt to enter the stream of meanings and the continuous effort to nudge the flow in one direction or another without necessarily having a grand plan or master interpretation. Respect for the “thingness” of meaning can help us see through the naiveté of the view of therapy as the linear application of higher order metapsychological theory (and technical postulates derived therefrom) to clinical “material.” To see the limits to what we can control and definitively understand is to revision our concept of what therapeutic skill is all about.