9
Interpretation and Subjectivity

A Phenomenology of Resistance

Interpreting the Absolute Unconscious: Science and Objectivity

If unconscious experience does not have a single, predetermined meaning, but remains to be interpreted in reflective awareness, the effect of clinical interpretation does not depend on objective accuracy and cannot be judged on that basis. This is a clinical stance about the nature of the unconscious and the interpretive qualities of reflective consciousness, but it is also a philosophical position on matters of mind and experience, and so it should come as no surprise that philosophers have given their explicit attention to it. In independent proposals, hermeneuticists Paul Ricoeur (1970, 1977), Karl-Otto Apel (see Warnke, 1987), and Jiirgen Habermas (1971) have taken the position that, though consciousness is phenomenological, so that we are correct to understand it to be constituted socially and linguistically, the unconscious is just as thoroughly nonsocial, nonlinguistic, and nonphenomenological. In fact, they argue, the quality of unconsciousness is actually the result of the complete removal of experience from a social context; such contents have been "delinguisticized" and now exist in an "absolute" realm, a realm with a real, objective existence outside the range of phenomenology and social construction. This absolute unconscious (Ricoeur's term) acts on consciousness with (invisible) objective force, a force no less real and concrete than those operative in chemistry and physics.

All three of these hermeneutic writers consider psychoanalysis and Freud to be synonymous, and so they do not question the concept of drive, a stance that perhaps makes it more natural for them than it is for psychoanalysts today to attribute objective force to the unconscious. Perhaps, however, we can do away with the notion of objective force and still maintain some of their argument. It is at least barely possible to conceive an objective existence for even the socially constructed unconscious posited in "relational-conflict theories" (Greenberg and Mitchell, 1983) ever since the work of Fairbairn. That is, perhaps a socially constructed unconscious can still be conceived to cause its effects directly, by means of such sub rosa phenomena as fantasies, feelings, and wishes, all of which we have to conceptualize, in this case, as having objective existence (cf. Schafer, 1976). To understand fantasies, feelings, and wishes as objective, of course, would contradict the thesis of this book. But even if we were willing to stretch that far to accommodate the views of Ricoeur and others, it would be impossible to align them to the phenomenology of clinical interpretation.

If unconscious meaning is an objective fact, and the clinical function of language is to label it, then the purpose of interpretation is the accurate matching of facts and labels. The only interpretations we can possibly make of objectively existing unconscious content must be—like their objects—objective, scientific, and nonphenomenological. Under these conditions, the analyst would be expected to explain the patient's conduct and experience on the basis of nonintrospectible, but theoretically conceivable, absolute unconscious phenomena. The analyst would then be expected to convey these explanations to the patient in the form of objectively accurate interpretations. We might imagine, as a very rough example, that the only way a certain patient will ever learn that his problems with women have to do with his mother, because he cannot observe his own unconscious, is for the analyst to make some kind of objective statement about the matter. Or, to be (a bit) more subtle, we might cite the analyst's objective interpretation of the patient's resistance to knowing the relevant pathogenic experience. Fourcher (1992) calls these interpretations of the absolute unconscious "shovels" to distinguish them from the "lens" interpretations that formulate or "focus" what is implicit. Approvingly citing Ricoeur, Fourcher goes on to say that objective interpretations themselves, once offered, may go on to become causative agents in the unfolding of events in the treatment.

Interpreting Unformulated Experience: Subjectivity and Phenomenology

What is the alternative?

One can usually sense whether the verbal meaning one selects or constructs is adequate, that is, whether it is at least a barely satisfying answer to certain vague expectations one has developed prior to the appearance of the interpretation. But one cannot say whether that verbal meaning is the single best answer to one's expectations. One often knows quite clearly, in other words, when an interpretation is inadequate, but it is not similarly possible to know when one has arrived at the understanding that best suits one's purpose.¹

This is an important difference between the formulation of unformulated experience and the traditional psychoanalytic understanding of interpretation. In most traditional views, the goodness of an interpretation is judged by a correspondence between the verbal description and the nonverbal, unconscious material it encompasses. Consider Freud's (1913) metaphor of the landscape outside the moving train window, a metaphor he devised to capture the nature of free association, and which Donald Spence (1982) offers, approvingly, as a description of what free association should be. The analysand in this model is not to select what to say, but simply to describe passively the contents passing through her mind, as a passenger on the train would describe the passing view. Language is used as a set of labels in this view, and accuracy is the criterion of truth.

Psychoanalysts and analysands do judge the goodness of their interpretations, of course. They do that continuously. But the accuracy of our portrayals of unconscious meaning is virtually irrelevant as a truth criterion. "Accuracy" is not really even a meaningful term in discussing the interpretation of unformulated experience, because the term cannot be defined by reference to an observable relation between itself and its object. We know the object (to repeat the essential point) only by means of our interpretation of it. That means clinical interpretation is not objective and scientific, as Ricoeur and his colleagues claim, but subjective and phenomenological. And it spells the end of correspondence theory in psychoanalysis. We can no longer hold that the nonverbal unconscious meaning is the "real" one that our words simply clothe or represent, or to which they correspond.

Correspondence Theorists in Psychoanalysis

Two notable contemporary articulators of psychoanalytic correspondence theory are Wilma Bucci (1985), who proposes a dual coding model for psychoanalytic research, basing her thinking in the work of Paivio (1986, 1991), and Donald Spence (1982, 1987, 1988a, 1990, 1993), whose approving citation of Freud's train window metaphor I have already mentioned. Spence, though he has done so much to bring hermeneutics into psychoanalysis, claims nevertheless that the ideal of clinical interpretation is still the accurate, passive verbal representation of the nonverbal, as if a noninterpretive registration of the view outside the train window were actually possible—as if the view itself were not already an interpretation. Despite his hermeneutics, then, Spence is an objectivist. He does accept that we have no choice but to couch our reflective thoughts in language, and he accepts that language can have constitutive properties; this much of his position is hermeneutic. But he laments these properties, claiming that because of them language almost inevitably distorts what it represents. Here, in the assumption that language is most properly used to represent accurately what is already there, he deserts hermeneutics and lapses into objectivism.²

In her influential 1985 article, Bucci makes no bones about rejecting hermeneutics, calling it "both an incorporation of and a backlash to verbal dominance theory and the associated behaviorist position" (p. 600). Bucci argues for a cognitive science model that will allow experimentation and quantifiable results on psychoanalytic propositions. Since I am basing my thinking not in science at all, but in a hermeneutic approach to language, it is odd that Bucci's position is in some ways closer to my own than Spence's.

In Bucci's view, experience is coded in two separate systems, one verbal and the other nonverbal and imagistic. For the most part, language is what makes experience conscious. To be mutative, language must connect with something in the patient's mind that is already there: a nonverbal representation. And yet the nonverbal representation, because it can only become conscious by means of language, cannot be directly accessed to check the degree of correspondence between itself and the language used to represent it. To this point, Bucci and I think along a similar path, except that I also make room for unconscious meaning implicit to language and she seems not to. It is when we consider the kind of connection that must be made between language and nonverbal representations that differences emerge.

Here Bucci's theory becomes a classical objectivist theory of correspondence, for Bucci explicitly eschews truth criteria of aesthetic appeal and pragmatic effectiveness, and insists instead that the connection between language and nonverbal representations must be evaluated only by an assessment of representational accuracy. She writes that "the issue of verification concerns the extent to which the language spoken by the patient and heard by the analyst, and the interpretations with which the analyst responds, correspond to the mental representations—memories of the past, images, fantasies, and beliefs—registered in the patient's mind" (p. 601). The means by which Bucci proposes to conduct this assessment of validity are somewhat novel. They have to be, because she grants that the nonverbal representations are not accessible. But, in the end, the view is traditional.

Bucci decries any way of thinking that takes psychoanalysis away from science. "If language determines thought, then we cannot look beyond language to any independent level of representation. Furthermore, if it is the case that a scientific theory can refer entirely and only to observables, then indeed we must despair of setting psychoanalytic theory in scientific form" (p. 600). Well, as I have txied to say, granting language constitutive properties does not have to amount to the claim that language determines thought, only that it cocreates it—that is, that language operates constructively, but within constraints. When we maintain experience in an unformulated state for defensive reasons, we are avoiding making an interpretation; we are not preventing an uncontrolled "appearance" of experience in consciousness. But with that caveat, Bucci's assessment seems correct to me. We do need to stop insisting on setting psychoanalytic theory in scientific form; but first, we need to stop seeing that alternative as the embrace of despair.

The Validity of the Verbal

In making sure to give nonverbal patterning in experience its due, it is possible to bend over backwards and go too far. We need to remember that verbally structured experience also has an integrity of its own. In any particular instance, what we come up with when we develop a convincing reflective understanding of ourselves is not a pale form of a more full-bodied and basic nonverbal truth, but the certain kind of truth that verbal language makes available to us, and that is available to us in no other way. To understand in any terms, verbal or nonverbal, takes interpretation; but to understand in a way we can reflect on explicitly takes the particular kind of interpretation offered by verbal language—not in spite of its constitutive properties, but because of them.

Thus there is no reason to believe that the reality analyst and patient so painstakingly construct, however specific it is to their work together, is either arbitrary or frivolous. Postmodern views, for obvious reasons, encourage the preservation of whatever uncertainty we can manage; but in directing our attention to what we don't know, they do not necessarily threaten the conviction with which we defend the thoughtfully considered clinical constructions we do know. Analysts in these postmodern days remain just as passionate as analysts have always been in searching for what is most true and useful, and they maintain the strongest kind of feelings that one thing is more true and useful than another. Giving choice and conviction their due in the process of thought, as postmodernism does, is not at all the same thing as saying that we can conclude anything we please and still claim to be carrying out our work responsibly. We still have to choose the point of view that works the best, that is most complete and satisfying in its account of the phenomena in question. And we do not have to accept that reality itself has no structure other than that which we impose upon it. There are many ways of understanding "what is," and some of them—the ones that are most appealing and convincing to me, as a matter of fact—argue that each human being and the world around him or her are a unity, that person and world both participate in the creation of experience (e.g., Feffer, 1982, 1988).

Interpretation from the Patient's Perspective

The phenomenological contradictions in the concept of the objective interpretation come from both the analyst's perspective and the patient's. Even if it were feasible for analysts to be objective in their private thoughts about the patient (and even that limited goal is not realistic), it is utterly impossible for the patient to experience the analyst's spoken interpretations in an objective mode. No patient accepts or rejects an interpretation because it is "objectively" true or false, but because it is or is not subjectively convincing. The successful interpretation touches the patient in a way that the patient can identify in her own experience. Therefore, even in a world in which the analyst's observations could be objective, "scientific" (third-person) accuracy would be a futile goal for interpretation. We use no shovels.

It certainly is a common enough occurrence for the analyst to think she is right and the patient to think the analyst is wrong (or vice versa). We might cite these disagreements as evidence that the patient's sense of conviction is not always the determining factor in how the analyst makes interventions, and we would be correct. At times, patients and analysts may even argue about whether a particular interpretation is accurate, or whether one is better than another (e.g., Hoffman, 1992b). The psychoanalyst, too, after all, develops subjective convictions about the best way to understand.

But if the disagreement provokes strong feelings on the part of either or both participants, we expect something to "give" before too long. Either the analyst finds a better way to make the point (or to help the patient see why she avoids it); or the analyst decides, on the basis of what the patient has to say, that she (the analyst) was wrong; or the patient, because of the depth of the collaboration, accepts that the analyst may be on to something and keeps chewing over the analyst's point, until finding a way to link it to her (the patient's) own experience and develop conviction about it. The analyst and patient may also decide to maintain a mutually respectful disagreement until (or unless) an agreement becomes possible.

But if the disagreement does not progress in any of these ways, and patient and analyst stay at loggerheads with one another for more than a brief period of time, it is not our inclination to say that one of the two is right and the other wrong. At least, even if this statement were true, it would not be the focus of our clinical interest. We would instead begin to think in terms of a transference-countertransference bind.

I intend this little digression to be an illustration of the fact that analysts do not—cannot—depend on being "right." Rather, they think about how to account for the patient's experience from within the patient's perspective. Because they know that only those observations that arouse the patient's recognition will be useful, analysts try to observe from within what they imagine the patient's frame of reference to be. And even when analysts do make observations from a different perspective (on the basis of their own theoretical leanings, experience, or emotional reactions), they ask themselves how such observations would be framed from within the patient's point of view, and even use their capacity to frame their observations in those terms as a test of the observations' accuracy and utility.

Objectivists might argue that third-person, scientific interpretive accuracy would be effective if it weren't for the patient's resistance, that it is only resistance that keeps the patient from accepting the analyst's objectively correct observation. With even greater justification, it might be claimed that resistance must be interpreted in the third person. Because the patient obviously cannot observe her own unconscious reasons for defending against a particular kind of experience, the analyst must suggest it by way of interpretation. Such an interpretation, if it really does have nothing to do with the patient's conscious experience, must be couched in the third person, and therefore qualifies as an objective observation by the analyst of an unconscious fact about the patient. The analyst's objective interpretation of resistance is perhaps the epitome of the shovel interpretation. It cannot possibly be couched in phenomenological terms—or so the argument goes. This is the key issue here, because unless we can understand some way that the patient develops a phenomenological appreciation of the analyst's interpretation of material that is (for the patient) unconscious, we have little choice, from a purely logical perspective, but to fall back on the idea of third-person, objective interpretation.

"Fit" and Recognition

The patient always has a sense of whether or not the analyst's interpretation "fits." Without the feeling of fit, the patient can develop no sense of conviction. And without conviction, the interpretation may as well have appeared in the patient's morning newspaper, for all the good it will do.

Usually, and always in the case of unconscious material, the judgment of fit the patient makes about what the analyst has to say is not as simple as the evaluation of whether or not the interpretation "matches" some preexisting content in the patient's mind. The judgment of fit in psychoanalysis is much more complicated than the judgments we make about the physical world. Because preexisting nonlinguistic representations must be interpreted in language before they can be reflected on, there is no such thing as pristine "preexisting content" that exists in the patient's mind, and against which the "accuracy" of the verbal interpretation can be tested. In a very real sense, the interpretations we make of nonlinguistic unconscious representations are those representations—because they are as close as we can get to thinking about them. In fact, we can say not only that evaluating fit is not a matter of judging brute accuracy, we can actually say that it cannot be. As Loewald (1960) taught years ago, interpretation has an organizing function. It is not a set of correct labels, but a redescription of the patient's experience at a level of differentation and articulation higher than the patient had heretofore reached. In this process of redescribing, an interpretation may bring together pieces of experience, or even of logic or emotional argumentation, that have never before appeared in the patient's mind in a single configuration. An interpretation is a new Gestalt.

What, then, accounts for the patient's feeling of fit? In certain instances that tend to be rare and dramatic, the perception of fit follows the patient's sudden feeling of always having known, in some vague way, what the analyst is trying to say. More commonly, the patient feels that what the analyst says has "pulled something together," and is correct (or better, useful) in that respect. (But note, although the patient judges the analyst to have made a good integration, this does not necessarily imply that the particular integration is the only one that could have been made.)

It is interesting and heuristically useful to realize that, at least in this account, the patient judges the fit of the analyst's interpretation in the same way we all judge whether or not we ourselves have said what we set out to say. When we speak, our words either satisfy us or they do not; they either give us the sense that something has been expressed, has been grasped in words, or they do not. We often say, about our own descriptions, "No, that's not exactly right. That's not exactly what I want to say." Or, in more fortunate circumstances (and these are deeply satisfying), we say "Yes, that's it exactly. That is exactly what I meant." When we are trying to capture a thought in speech or writing, most often we feel something like, "That's close. It's the best I can do for the time being, but there may be something more that I can't grasp yet."

We seem to know when we have fit our words to a preexisting meaning—or perhaps better, a preexisting absence. The fact that we are capable of sensing such a thing, and that patients invariably evaluate and use (or discard) their analysts' interpretations on such a basis, means that there can be no such thing as an effective shovel interpretation, because shovel interpretations are themselves understood to be the "causes" of patients' reactions to them. The judgments patients make about their analysts' interpretations are not whether they are scientifically accurate, but whether the interpretations speak to them. Has the analyst said something that the patient recognizes as something she herself would have said, if she could have? If an interpretation is useful, it can only be because the patient has grasped it phenomenologically—which is exactly how shovel interpretations cannot be grasped.

The Patient's Feeling of Safety

There is in psychoanalysis a long history of conceptualizing the interventions that "lead up" to interpretation (see, for instance, Bibring's classic 1954 contribution). One of the subliminal reasons this literature had to exist—one of the reasons psychoanalysis needed descriptions of interventions like Bibring's "suggestion," "abreaction," "manipulation," and "clarification"—was that there had to be some way to explain why the patient accepted interpretations of resistance. With preparation over time, the relevant material could be brought slowly into the patient's preconscious via the noninterpretive interventions until eventually, at just the right moment, the analyst could make the mutative interpretation, finally freeing the patient from unconscious tyranny.

But this literature really only moved the problem of how patients recognize the utility of an interpretation of unconscious processes farther back from the moment of the interpretation, much the way Mitchell (1984) shows that certain theorists have explained difficult clinical problems by assigning them ever earlier developmental origins. Even if we admit kinds of interventions other than interpretation itself, the analyst must still make an initial foray into unconscious material at some point in the treatment; and even if that foray is not formally categorized as an interpretation, the patient must have some kind of reason to consider it.

It was Freud (1913) who offered the beginning of an answer to the problem. He wrote that, although the positive transference should not be used for the purpose of suggestion in psychoanalysis, a treatment should only be called psychoanalysis "if the intensity of the transference has been utilized for the overcoming of resistances" (p. 143). This observation (which has not been elaborated as much as might be expected³) acknowledges that, if the patient is to recognize the truth of the analyst's interpretation of the unconscious, the patient must have some awfully good reason to care about the analyst's interpretation. The patient, that is, must harbor the strange and seemingly paradoxical willingness to work toward the acceptance of understanding she otherwise wants to avoid. What emotional or esthetic purchase is the patient supposed to get on an interpretation that is wholly unfamiliar to her conscious experience, but that the analyst claims is objectively true? What is the patient to do with such an interpretation? And if objective interpretation is not the royal road to unconscious material, what is?

In the case of an interpretation that is uncontroversial, one that integrates without challenging dynamic equilibrium, what I have already said about "fit" is not terribly difficult to accept. But where does what I have said leave us in the case of resistance? What can a phenomenological recognition of resistance possibly mean? Is the phrase not an oxymoron?

It is in most of life. But there is a crucial ingredient in the analytic situation that appears in no other relationship, and it is not mere suggestion. It is rather the collaboration between analyst and patient, initially undertaken as a result of a mutual decision to begin treatment, but eventually deepened and widened because of the patient's trust in the analyst—which is, in turn, born of the atmosphere of safety the analyst provides.⁴

"Trust" is a terribly overused word, one that has been applied in the field of psychotherapy with appalling naivete. In referring to the patient's trust in the analyst, I certainly do not mean to suggest what has so often been implied in nonpsychoanalytic accounts of therapeutic practice: that the therapist is somehow to persuade the patient, or provide some convincing demonstration, that she is worthy of trust, so that the patient will "open up." No transference worthy of the name could possibly be breached by this kind of frontal attack, which is more accurately described as a reaction formation against the therapist's aggression than a genuine attempt to establish trustworthiness. In this scenario, trust is equivalent to being accepting, uncritical, and nonretaliative. All of these characteristics, if they are not assumed in too thorough or hidebound a fashion, are certainly worthy. But none of them is what I am referring to in using the word "trust."

Acceptance that the analyst intends for its own sake is useless. Unconditional positive regard is ultimately ingenuine. Real acceptance does not preclude the analyst's rejecting internal reactions; it only precludes making the patient responsible for alleviating them. Therefore, acceptance that is a contribution to the patient's feeling of safety and the establishment and growth of a therapeutic collaboration can only be the acceptance inherent in the analyst's genuine curiosity—compassionate curiosity, to be sure, but curiosity nonetheless. To claim the presence of a collaboration, as a matter of fact, is just another way of saying that the patient has accepted the analyst's curiosity about her and has developed the capacity to be curious about herself, free (or relatively so) of reflexive self-hatred.

And what is the analyst's curiosity.? It is not merely the conscious application of inquisitiveness, the asking of questions. It is the more basic capacity to become aware of questions, of gaps in the material, to sense that there is something there to be curious about. Curiosity is the sensitivity to the possibility of a question. It is the means by which the analyst gains self-reflective access to experience, the means by which she disembeds herself from immediate circumstance and unconscious assumptions, especially those that structure the interpersonal field of the analytic situation. The analyst's accepting attitude toward her inevitable involvement in enactments with the patient, and the curiosity that this acceptance allows about, not only the patient's experience, but her own, is the only means at the analyst's disposal to provoke this crucial process of disembedding. And each time she is successful in provoking it (though she never knows quite how it has been accomplished), she is able to refrain from the enactment in question and instead interpret the patient's transference. Thus, transference interpretations are the only evidence the patient can really depend on that the analyst is not "drowning in the countertransference" (Racker, 1968), the only authentic indication that the analyst is able to know and tolerate the countertransference and use it to help the patient. Of all the analyst's interventions, therefore, transference interpretation (and the countertransference interpretation that so often precedes it, usually privately) plays the most significant role in establishing and maintaining the patient's feeling of safety (Hoffman, 1983).⁵

By repeatedly transforming the limits of her own understanding into thought, then, the analyst becomes able to offer the patient something more than interpretations of experience that are already known. As she disembeds herself, the analyst comes upon hitherto invisible possibilities for new experience. And thus, although spoken questions may be the most visible evidence of curiosity, they are the result of the phenomenon, not the phenomenon itself. I mean to emphasize the openness and unbidden perceptions from which inquisitiveness arises.

The Place of the Therapeutic Collaboration in the Patient's Judgment of "Fit"

Once a relatively thorough collaboration has developed (and it is always relative, of course, because even the best collaboration is inevitably shot through with holes that are unknown to both analyst and patient until the analysis itself reveals them), there is more experience available to the patient in this unique interpersonal setting than there is anywhere else in life. The patient is not only able to be more curious than usual; often she is eager to be, even about painful matters. When the patient feels that the interpersonal atmosphere is reliably safe, the rewards of knowing are often great enough to make the pain of knowing, and the uncertainty that precedes and often accompanies it, worth bearing. Merleau-Ponty (1962), for whom the analyst's interpretation is much more than the mere delivery of a packet of information, makes the point this way in a discussion of a patient with an hysterical aphonia.

We can agree, then, on the centrality of the analytic relationship in the effectiveness of an interpretation. But what exactly is it that the patient senses about unconscious content that allows her to determine the "fit" of the analyst's interpretation?

Prior to the interpretation, it often makes sense to say that even the collaborative patient has had no conscious sense of the meaning conveyed in the analyst's interpretation of unconscious material. Even so, in response to a successful intervention, the patient working in a relatively safe-feeling analytic atmosphere senses (i.e., is willing to sense, is interested in sensing) a "place," a kind of empty meaning-mold of the same shape as the interpretation, a "place" that, despite its not having been felt before by the patient, may nevertheless feel, in the very moment of its appearance, as if it has been waiting for the interpretation to fill it. This "place" is not just a manifestation of the patient's acceptance of the new meaning; it is the simultaneous recognition of this meaning and the absence that is now suddenly understood to have preceded it. The painter Ben Shahn (1957) refers to "the shape of content"; the collaborative analytic patient, in response to the analyst's successful interpretation of unconscious meaning, senses a preexisting "shape of absence." In rare and exhilarating instances, the interpretation fits the absence as a hand fits a glove, or as Cinderella's foot fit the slipper. Most of the time, though, while the fit may be recognizable, it is not necessarily perfect—nor need it be. It needs only to be good enough to move the analytic inquiry forward. There is almost always room for further shaping of the new meaning, to say nothing of the inevitable changes over time we expect to find in the object of the analyst's interpretation. I have already quoted Herbert Fingarette's (1963) maxim that insight is like seeing a well-formed ship in a cloud instead of a poorly formed rabbit (see chapter 2). We can now add that the cloud itself changes with the passage of time, and so the ship inevitably disappears; and even before it does, if one keeps looking, perhaps one sees a tree in the cloud that is even better formed than the ship.

The analyst helps along the process by which the patient recognizes the new meaning by couching interpretations in a way that allows the patient, either explicitly or implicitly, to reconstruct the means by which the interpretation came into the analyst's head. The analyst, in other words, often conveys interpretations by helping the patient to grasp the analyst's subjectivity. She does this not to maintain democracy, even if she believes a certain kind of democracy is important in analysis (and I do), but because she knows that if the patient sees how the analyst arrived at her conclusion, it will be easier for the patient to recognize the interpretation in her own frame of reference. The analyst behaves this way whether the interpretation in question is a rearrangement of matters of which the patient is entirely aware, or is an attempt to give voice to material that has been dissociated in the more traditional fashion.

The patient is always listening to the analyst in a context, and this context is the ground on which the interpretation falls; this "ground" determines the kind and range of significance the interpretation can have. We might call the ground of an interpretation its "context of significance," or its "context of relevance." Or, adapting a term used by Heidegger and Gadamer to a slightly different, more prosaic purpose, we might call this context the momentary "horizon" of the patient's experience. The limits of this therapeutic horizon determine what can appear in the patient's experience, what meanings experience can have for her. The horizon is closely related to the interpersonal field. As a matter of fact, we might call the therapeutic horizon a manifestation of the field. And the interpersonal field in the psychoanalytic situation contains, in addition to the transference-countertransference, the atmosphere of safety, in whatever degree it has been realized. Actually, in consonance with those writers, ranging in orientation all the way from Brenner (1980) to Hoffman (1983, 1991, 1992a, b, 1994, 1996) and Levenson (1972, 1983, 1991), for whom the therapeutic collaboration and the transference must be understood as parts of a single, unitary relationship between analyst and patient, and so cannot possibly be considered in isolation from one another, it is more truthful to say that the atmosphere of safety expresses the transference-countertransference, or is part of it.

In discussing the patient's willingness to stretch her dynamic limits in response to the atmosphere of safety, we should not overlook the fact that, when the patient feels safe and the transference is positive, she may very well want to please the analyst. It would be dangerous if the outcome that derived from this motive were simple suggestion—that is, the patient's uncritical acceptance of the analyst's interpretations. But we know that, most of the time, even patients who feel warmly toward their analysts are uninterested in parrotting them. And if the analyst also watches and listens for uncritical acceptance, it is unlikely to go unanalyzed.

The patient in a positive transference, however, even if she has no particular wish to parrot the analyst, does have a powerful and unobjectionable reason to give serious consideration to anything the analyst says. This openness to the analyst is unobjectionable because it is essentially just a honing of the patient's curiosity, a willingness to look for the subjective evidence to substantiate or disprove the analyst's observations. Even patients in a negative transference, if they have managed to develop the rudiments of collaboration before (or even during) the time the angry or disappointed feelings set in, can recognize "fit." The angry patient may not be willing to give the analyst the satisfaction of sharing this acknowledgment of fit, at least not until some later time, and there are times during which even the capacity to recognize fit is compromised by the negatively tinged interaction. Except in the case of intransigent negative transference, or negative therapeutic reaction, the patient's interest in recognizing fit never disappears altogether, once the patient has accepted psychoanalysis as a collaborative endeavor.

As an example, consider the patient who begins arriving late to sessions and insists that her behavior is due only to traffic, business meetings, or some other factor exterior to the treatment. The analyst who tries to interpret the first or second instance of such lateness, unless the patient has been compulsively punctual up to that point, is liable to be met with outright rejection from the patient. Even if the analyst waits until the behavior goes on long enough to establish a convincing pattern, she often runs into a brick wall if she simply asserts a dynamic reason for the lateness in the face of the patient's denial of such a reason. On the other hand, if the patient understands and accepts the general outlines of the analytic task, and the analyst presents the interpretation by laying out the agreement between the two of them to understand whatever they can about the patient, the patient is liable to be able to see, when it is pointed out, that her lateness deprives her of some of the opportunity to carry out the task she is otherwise committed to pursue, and that it is at least plausible that something more than traffic or business meetings keeps her from arriving on time. The important thing is that even if the analyst's internal observation of the patient is "objective" and made in the third person, the interpretation most certainly is not. The patient accepts what the analyst has to say, not because it is an objective fact, but because the interpretation makes subjective sense to her.

After this initial intervention (if it is successful), and perhaps after examining lingering feelings that the analysis of lateness is really only a disguised slap on the wrist from the analyst (a common enough feeling, and commonly enough correct), the patient is in a much better position to judge the fit of any particular interpretation of the lateness the analyst might offer. Of course, the patient is also in a better position to develop her own hypotheses about the matter, because the analyst's convincing curiosity, her devotion to the analytic task, has relieved the patient and led to an increment in the feeling of safety. That allows the patient greater access to the kind of experience that might make a defensive measure like lateness understandable. Finally, once the patient has been able to see that it is reasonable to hypothesize that her lateness is defensive, analyst and patient are in a position to address why the patient was so insistent that the lateness could only have been a matter of external circumstance. All these steps, and more, are taken without the analyst doing anything that qualifies as the uncovering or labelling of absolute meaning.

The Atmosphere of Safety in the Perception of Feelings of Tendency

The analyst gears interpretations to the patient's subjectivity because she knows, without having to be told, that that is how people understand anything. And knowing this, the atmosphere of safety becomes crucial. In thinking about the process of interpretation, we find that a well-established therapeutic collaboration, rooted in the atmosphere of safety, has the effect of helping the patient attend (or cease inattending) to certain meanings, or to the possibility of making those meanings, that would otherwise remain outside the patient's conscious horizon. The therapeutic relationship increases the possibilities for articulating meanings. What I am suggesting is that the relationship, in fortunate instances, makes possible a direct breach of the patient's defenses. Defense, in other words, is an interpersonal process, not only an intrapsychic one. No part of the inner world is isolated from the social realm. As Fairbairn (1954) writes in regard to hysterical patients, "the resistance . . . is not so much a resistance to the psychotherapeutic process as a resistance to the psychotherapist himself' (p. 105). This point is in direct disagreement with Apel, Habermas, and Ricoeur, who understand the repressed internal world to be isolated, removed from social contact.

"Fit" and Unformulated Experience

When the patient in a successful therapeutic collaboration begins to recognize the possibility of previously disavowed meaning, she is moving from the use of unformulated experience as familiar chaos (chapter 3) to its use as creative disorder (chapter 4). This means a move from the denial of curiosity (in certain circumstances, and about certain aspects of experience) toward its affirmation, and a move from the refusal to acknowledge (again, in particular circumstances and in regard to certain content) what William James (1890) referred to as "feelings of tendency" and "signs of direction" in thought toward an openness to these same parts of experience. As a matter of fact, what I have described in this chapter as the collaborative patient's capacity to judge the "fit" of the analyst's interpretation could just as well be presented as the patient's recognition, in response to the analyst's intervention, of the relevant feelings of tendency in her own experience. The patient in a well-established collaborative analytic relationship is beginning to think effectively, for the first time, about parts of life that had heretofore remained outside the range of reflection. It is this new curiosity and freedom, rather than any particular new interpretation, that is the truest reflection of the patient's therapeutic movement.

Interpretation from the Analyst's Perspective

Modern interpersonal and relational psychoanalysis is built on the premise that anything the analyst does with or about the patient, including the formulation of private and public observations, is a participation in the relationship (see Hirsch, 1996, for a review), and all participation is inevitably influenced by the unconscious factors that structure the relationship—unconscious factors that it is the task of the analysis to grasp. Edgar Levenson (1972, 1983, 1991), who was among the first to make this point and has always made it forcefully, purposely provoked his psychoanalytic readership by titling his first book The Fallacy of Understanding (1972). From this perspective, which is shared by all the many writers who qualify as what Irwin Hoffman (1983) calls "radical critics of the blank screen model," there can be no such thing as a simple, "objective" observation or interpretation. We must be too skeptical and curious to accept anything the analyst observes about the patient as a simple, objective response from an unbiased observer. This point encompasses even the content of the analyst's empathic responsiveness (Spence, 1984, 1988b; Friedman, 1985; Moses, 1988; Stern, 1988, 1994).

We know from everyday experience in our consulting rooms that the analyst's informed subjectivity will often be useful to the patient; but there is no getting around the fact that subjectivity is what it is. To make an interpretation, private or public, requires the analyst to give shape to unformulated experience, and that is a task every human being has no choice but to do in a subjective way. Coming full circle, that is another way to say why every clinical intervention must be a participation in the relationship with the patient. After all, an interpretation is merely a particular instance of the analyst's generic human capacity to give verbal shape to unformulated experience. How could the analyst formulate an interpretation without being influenced by the interpersonal field? Such a thing cannot happen between any two people, as we saw in chapter 8.

The analyst evaluates the interpretation in the same way the patient will when she hears it; the analyst, that is, judges the degree to which she has been able to say something satisfying. For analysts, words either satisfy or disappoint. Good interpretations give the analyst the feeling that the chosen words have expressed what they were intended to say; they fill that "shape of absence," just as the interpretation will do, if it is useful, for the patient. We take it for granted, though, that the analyst's own feeling about the interpretation is not enough to make it successful. It is offered with the hope that it will arouse the patient's answering sense of recognition and conviction.

Gadamer's Perspectivism and Psychoanalytic Interpretation

In concluding this chapter, I turn to the perspectivistic hermeneutics of Hans-Georg Gadamer (1975, 1976; see also chapter 11), which entails a conception of the kind of reality that unifies and underlies both my phenomenology of psychoanalytic interpretation and my rejection of the objective, unitary, nonsocial unconscious. Gadamer is simultaneously a perspectivist, a phenomenologist, and a realist. He acknowledges that reality exists, but he also insists that it is inevitably understood from a social, historical, linguistic, and (we can add, from our psychoanalytic vantage point) individual perspective. Any interpretation, psychoanalytic or otherwise, is a perspective.

We can understand one view of perspective (but not Gadamer's) by means of a visual metaphor. If you hold in your hand an object, and you turn it this way and that, each time you turn it you will view it from a different perspective, and each perspective offers something new about the object that cannot be appreciated from any other perspective. Reality, if we were to understand it on the basis of such a model, would be made up of any number of such perspectives, couched in all the various modes in which we apprehend the world. Taken together, all these perspectives would be a complete picture. But because each era and each cultural tradition encourages access to only a portion of the perspectives that are possible, and because each individual has access to only a portion of the perspectives available in her time and social groups, an individual's view of reality, including the reality of what she herself is, must necessarily be highly selective. In this view, which is more or less traditionally essentialist, the task of psychoanalysis is to help the analysand lift the personal and cultural veils and reveal as much of internal and external reality as possible. This model is useful as far as it goes. Schachtel (1959, 1970) used it to great effect in articulating his conception of the allocentric mode of perception, in which what it means to know the object fully is to perceive it from every possible perspective.

The problem with the approach is twofold. First, it reduces the individual's role in the construction of experience to the selection of an angle of view on a preformulated object, and that, as far as I am concerned, unreasonably cramps our conception of the constitutive involvement of the individual in the genesis of her own experience. The construction of experience is not, like Saki's short story "The Lady and the Tiger," simply a matter of which doors to open. We are more involved in how to formulate the experience we encounter than that. Gadamer's perspectivism is as much about how we formulate the object as it is about how we view an object that already exists. Second, this "angle-of-view" perspectivism leads to a constricted view of reality. Such a reality is unitary and unacceptably concrete. It hides nothing, it implies nothing. It contains no particularly vague or unformulated truths. If you take the right "photograph" of it, it reveals anything, because all its truth is right there to see. To expose it, you need only observe it from the right angle.

As Gadamer envisioned it, reality is quite different than that. It is an inexhaustible resource. Like great works of art, it is so rich and allows so many perspectives—interpretations—that it is even possible for two contradictory perspectives to be true, or, in the slightly more conservative expression Gadamer sometimes uses, contradictory interpretations can each "have their truth." Think of Tevye's quandary in Fiddler on the Roof. He has just agreed with one neighbor's account of what would be a fair solution to a disagreement when the opposing neighbor offers a reasonable but diametrically opposed solution. With a good deal of beard-scratching, Tevye also agrees with the second neighbor. When a third neighbor objects that they can't both be right, Tevye says cheerfully, for emphasis stabbing the air with his forefinger, "You're right, too!"

Tevye's quandary is one of the core problems of the psychoanalytic situation. In conservative psychoanalytic thought, reality is as simple as the object one turns in one's hand, and the interpretations of both the patient (the transferential grasp of the analyst) and the analyst (the interpretation of the patient's transference) therefore can be judged simply by reference to reality. Analyst and patient simply have to be sure they are talking about the same angle; they can also verify that by comparing their experience with objective reality. In this scenario, Tevye is wrong: only one of his neighbors can be right.

For a growing number of contemporary psychoanalysts, though, the patient's idiosyncratic or problematic perceptions of the analyst and of people in outside life are seldom simply wrong (e.g., Racker, 1968; Levenson, 1972, 1983, 1991; Hoffman, 1983, 1991, 1992a, b). They are instead perspectives or interpretations with a certain validity of their own, and it is this validity (among other things) that it is the analytic task to articulate. Like Tevye's neighbors, the patient and the analyst are both right.

Tevye's position is Gadamer's position, too. For Gadamer, because the object of understanding is known only in light of the interpretations made of it, there is no independent comparison possible between interpretation and reality. We see what interpretations teach us to see; here we have a deep respect for the individual's constructive activity. We also have the manifestation of Gadamer's respect for the unimaginable richness of reality: it is so complex that it can accommodate multiple, even conflicting interpretations. In discussing Gadamer's thesis of the mimetic function of art, Georgia Warnke (1987) says it this way:

The important thing here is that Turner is not teaching us to revise our perception of the sea in an arbitrary way. The painter, for Gadamer, is not merely introducing a visual convention that is some Derridean combination of the arbitrary and ultimately meaningless traditions available in her era. Gadamer's view of tradition is affirmative instead. On the basis of what is available in his time and place, Gadamer says, Turner reveals to us a bit of reality.

All understanding for Gadamer is dialogic. It occurs only in conversation; in fact, understanding is conversation. Now, conversation means something in particular to Gadamer: it means the mutual attempt of the two partners in the conversation, just as I have presented the collaboration of the two participants in the analytic situation, to see the perspective of the other.⁶ And as I have already mentioned in passing, for Gadamer any perspective occurs within a set of horizons, or a context. That context is the set of conditions within which the perspective has its truth. If you wish to understand a perspective that is not your own, you must provoke within yourself an experience Gadamer calls a "fusion of horizons." That is, like Hegel, Gadamer sees in disagreement not just an obstacle, but an opportunity. The "fusion" is that moment when, after however much conversation is necessary, and from within your own horizon, you grasp the horizon of the other. In a famous passage, Gadamer (1975) writes:

This agreement, however, should not necessarily be understood as the victory of one view over another, or even as the achievement of a synthesis, though it certainly may be either of these things as long as both parties in the conversation freely accept the solution. The essential aspect of agreement is the new capacity, given the context in which the interpretation is made, to grasp the truth in what the other has to say. If we add to this equation the element of the patient's feeling of safety, making it possible to understand how the patient might eventually find her way to a position within which she can grasp the truth in what the analyst says about aspects of her own (the patient's) unconscious experience, psychoanalysis is very much a dialogue in the spirit of Gadamer.