Postmodern Epistemology: The Problem of Validation And the Retreat from Therapeutics in Psychoanalysis
Michael J. Bader, D.M.H.
Psychoanalytic Dialogues, 8(1): 1-32, 1998

Social-constructivist and other postmodern currents within contemporary psychoanalys put a great premium on an epistemological critique of positivism and the authority of the analyst.

This focus on the essential ambiguity and constructed nature of experience implicitly tilts the analyst's interest away from a disciplined attempt to develop systems of validation that rely on observable patient responses as confirmatory data with which to judge the analyst's interventions and understandings. As a result, even while defending themselves against charges of relativism and solipsism, many postmodern writers still tend to idealize uncertainty and implicitly discourage the clinician from seeking greater accuracy and clarity about the patient. Epistemic doubt and an ethic that celebrates surprise, although useful as a corrective to tendencies toward rigidity and arrogance in technique, unnecessarily clouds the ubiquitous existence and possibility of accurate intersubjective understanding. This bias can be seen as historically linked to similar processes in academia in which progressive intellectuals abandoned a social change agenda in response to a growing political conservatism and cultural cynicism and, instead, became increasingly involved in "theory for its own sake." Similarly, the epistemological position of the postmodernists sponsors a bias against therapeutic activism and inadvertently rationalizes a growing pessimism in our own clinical practice.

Theories, like analysts, reveal as much by the questions they do not ask as by the ones they do. Nowhere in psychoanalysis is this point more emphasized than among the social constructivists and related postmodern theorists. As seen through this model, the analyst's interpretation of a clinical moment reflects a complex choice, riddled with subjective and theoretical bias. At every moment the analyst's choices reveal some things while leaving others in darkness, encourage the patient to deal with some issues and not others, construct one ending and not another. Analytic reality is thus always constituted as much by the roads not taken as by the roads that are.

On the broader level of cultural critique, the postmodern project takes a similar position, deconstructing our dominant ideals and values-Progress, Democracy, Equality, Freedom, and Gender-and showing us how these grand "narratives" (Lyotard, 1983) exclude and suppress the voices, experiences, and interests of the Other, be that Other a minority group, women, gays, and so on. These omissions operate behind our backs. We live our lives as if our ideal's and values were universal truths instead of constructions within a particular social and historical space or, as Heidegger put it, a particular "cultural clearing" (as cited in Cushman, 1995). The abstract ideal of equality in Western society, for instance, takes on a different meaning if examined in the context of the gross inequality within which it arose and was shaped and the ways in which the virtue of equality excluded people of color and women, was based in part on the need to justify a free market system, or assumed a particular system of gender relations. In other words, an important postmodern precept is that behind every narrative, including narratives of progress, lies a counternarrative of those left behind or of alternative paths not taken.

Viewed from within this sensibility, classical analytic technique itself-clearly an advance over its hypnotic precursors is profoundly marked by what it omits and takes for granted. Its emphasis, for instance, on the interpretation of transference distortion can be seen as based on the presumption of a neutral analyst able to bracket her or his own personal subjectivity. The possibility that the analyst's unconscious is crucially constitutive of the patient's experience is overlooked; it is not the classical theorist's main interest. For contemporary theorists, however, this ideal of the analyst as "the one who knows" what is really going on inside the patient is the most pressing issue to deconstruct, based as it is on a defensive denial of the omnipresent play of the analyst's psychology, which continually shapes what the analyst sees, says, and does. Thus, the relative absence in classical theory of a focus on the analyst's subjectivity invites contemporary critics to put the notion of neutrality in a new light, expose its limitations, and better understand certain clinical phenomena such as the patient's conflicts over accurately perceiving the psychological issues of the analyst (Hoffman, 1983; Greenberg, 1991). By problematizing what is taken for granted and not questioned in classical theory-the analyst's privileged access to objectivity-contemporary postmodern analysts deepen our understanding of what we do question, namely, the meaning of the patient's reactions to the analyst. The result is a more complex and compelling view of the analytic interaction.

In recent years, the postmodern sensibility has become increasingly influential in psychoanalysis. I argue that, in particular, its emphasis on epistemology, its focus on the constructed nature of analytic knowledge and experience, its critique of the positivist tradition of classical psychoanalysis, and its eagerness to remind us of the centrality of ambiguity, doubt, surprise, contingency, and heterogeneity in the analytic situation have increasingly shaped analytic discourse. As the pendulum begins to swing toward this new paradigm, however, it should lead us to ask if anything has been left behind. In so asking, we need to do to the contemporary goose what has been done to the classical gander and consider what is omitted from the current postmodern discourse, the questions that are not pursued vigorously, the "choices" made to not study certain problems in analytic technique, and the values that are de-emphasized. In this context, one might ask if the emerging constructivist ideal of the analytic relationship makes some interests seem more legitimate than others, some voices worth listening to more than others, some goals better pursued by the analyst than others.

Moreover, if the constructivist turn in analysis captures something of the spirit of our age, it makes sense to ask if this emerging perspective reflects something of the problems of our age as well. It has been argued (Mitchell, 1993) that the contemporary emphasis on the hermeneutic and intersubjective construction of reality mirrors a crisis of confidence in social authority, including the authority of science, technological progress, the professional "expert," and the classical psychoanalyst. In this historical reading, the postmodern sensibility dovetails with an emerging social consciousness that questions tradition, is skeptical about universal truths, and challenges traditional hierarchical role-relationships. However, our modern age also contains a great deal of social pathology, including intense feelings of anxiety, cynicism, and political hopelessness. Might we not ask if echoes of these social currents can also be seen in the postmodern approach (Leary, 1994)? (footnote 1)

I will critically examine one of postmodernism's main "centers of gravity" (Elliot and Spezzano, 1996): epistemology. I will argue that the postmodern emphasis on epistemology, on the inherent uncertainty and ambiguity of the analyst's clinical understandings, tends to shift our attention away from recognizing and expanding the frequent and significant moments of certainty and accurate understanding in the analytic relationship. Donnel Stern (1991) has argued that modern hermeneutics proceeds on the assumption that "misunderstanding is the natural state of affairs" (p. 56). In other words, the postmodern critique of how and what the analyst knows tends to draw attention more to how this knowledge is biased, contingent, and constructed than to how the analyst can use clinical data to systematically validate his or her knowledge and generate more accurate versions of this knowledge. In arguing that a positivist type of "accuracy" is impossible or inappropriate to pursue-reality is constructed and not discovered-these theorists tend to veer away from attempting to either define or operationalize their own systems of validation, systems that must, of necessity, inform their clinical work. The commonplace clinical questions that we are all called on to answer-are we on the "right" clinical track, are we "in tune" with the patient, is the analysis "moving," is the patient "getting somewhere," is the patient "getting better"-require at the very least a kind of informal system of validation, of rough criteria for answering yes or no. Yet, even as analysts, regardless of orientation, always focus on these issues in practice, the postmodern emphasis on uncertainty in theory does not lend itself to making this a primary object of study. Instead, the possibility of validation is acknowledged only to be ignored. To the extent that it is addressed, the focus tends to be on the vicissitudes of the analyst's subjectivity, the analyst's countertransference, rather than on developing a disciplined way to use patient variables, particularly the observable behavioral, affective, and verbal responses of the patient, as sources of confirmatory evidence.

In my view, this bias or omission leads to the frequent sense in postmodern writings of a lag between theory and technique, a gap between epistemology and clinical practice. I believe, further, that this tendency to distrust or eschew interest in issues of validation, of how a therapist should judge if she or he is more or less doing the right thing, sponsors a bias against a disciplined focus on therapeutic aims and, therefore, diminishes the clinical relevance of modern epistemological debates. To the extent that there is such a subtle "tilt" in postmodem theory away from generating useful clinical principles that might increase our efficacy, this tendency to neglect therapeutic aims reflects a more general bias within psychoanalysis that emphasizes process over outcome.

In the last analysis, this bias away from an interest in validation and therapeutics, the fact that these issues do not occupy a central place in most postmodern theorizing (with important exceptions that I discuss), is part of a broader cultural retreat from an idealism and optimism about progressive political and social change. It dovetails with the similar tendency in academia to turn to a highly abstract method and philosophy that "deconstructs" ideals that claim universality rather than critically confront in the public arena the injustices of prevailing systems of power. It may be legitimate to question ideals that have failed to live up to their promise. The problem is that social change movements are also energized by ideals, by an optimistic vision of how a society could and should be structured. I argue that as movements for social change collapsed in the 1960s and 1970s, as the society became more conservative, there was a widespread disaffection with politics, an increased cynicism about change, and a movement away from an interest in forming communities and toward personal fulfillment. Progressive intellectuals, excited at first with the critical possibilities of certain postmodem intellectual currents, shifted their energy to the pursuit of academic advancement and security based on turning out more theory with less public accessibility or political relevance. Postmodern criticism in intellectual and academic circles increasingly shifted its discourse to a more relativist, apolitical, and abstruse level. No longer able or interested in changing the world, these social critics, intellectuals, and academics turned toward an increasingly rarefied and abstract postmodern discourse unhinged from its full potential for political practice.

The turn toward postmodernism in psychoanalysis tends to both reflect and potentially advance a similar tilt away from practice-in this case, from developing principles to guide therapeutic practice. In its preoccupation with the inherent ambiguity of reality, as well as its increasingly philosophical approach to discussions of this reality, the postmodern sensibility can become unhinged from a systematic attempt to change that reality.

The Issue of Validation: The Postmodern Narrative

In an attempt to anticipate the objection that I am critiquing a "straw" constructivist, it might be useful to draw a clearer and more nuanced portrait of the position about which I am raising questions. In response to various assumptions in the classical paradigm, the postmodem perspective assumes that psychoanalytic experience isn't "knowable" in the same sense that one can "know" the weight of a table. The analyst's mind isn't a reliable instrument in the same way a scale is. Psychoanalytic "truth" is saturated with ambiguity, indeterminacy, and contingency, and the resulting truth claims are necessarily "constructions" and interpretations. In asserting this, the postmodemists are claiming that constructivism is fundamental to all experience and all knowing. It infuses ontology and epistemology, within and outside the analytic dyad.

Experience, according to this position, is "constructed," in part, by virtue of its being context dependent. The ways that we feel about ourselves and others are fundamentally influenced by our real and imagined audiences. If I appear to be irritated, you may not necessarily be able to infer that this is a core trait of mine. It might be related to something about you. So, as an analyst, what I understand about you as a patient might be a contingent response to me to begin with. And I do not, as an analyst, have a privileged grasp of that which I am communicating and to which you might be responding (Hoffman, 1983, 1991a). Not only is the analyst not omniscient about her or his own unconscious and how it is being communicated, she or he can never be completely aware of what she or he is communicating in the moment because self-awareness is always retrospective (Renik, 1993b). The analyst is always expressing her or his complex subjectivity and can only "catch" it-if at all, and never completely-after it is expressed. Thus, the analyst's inferences about the stimuli to and context of the patient's experience is always ambiguous, partial, and contingent (Hoffman, 1994). Clinical "facts" are constructed both prospectively and retrospectively. Consider Ho-man (1992):

"But in the social constructivist model there is another source of uncertainty that derives from questions such as, on the retrospective side, what is not yet understood about the meaning of what I have said or done? And on the prospective side, what qualities of relating are available to the patient and to me at this moment? Can I relate to the patient now in a way that is authentically expressive at the same time that it promotes new understanding or the realization of new potentials in the patient's experience [pp. 293-94]

The analyst's theory and psyche are always influencing what the analyst sees and what meanings the analyst detects, constructs, and reconstructs in and for the patient. From the very first moment of the first meeting, the analyst begins to loosely formulate hypotheses about the patient and the analytic interaction. These rudimentary or "low level" theories inevitably bear the stamp of the analyst's theoretical preferences and psychological dispositions. Schafer (1992) described in detail how the foundational theoretical assumptions of the analyst begin to organize the material, and Hoffman (1983, 1992) and others (Levenson, 1983) have focused on the personal biases and personal blind spots and limitations of the analyst. In fact, it is argued that the analyst's unconscious is also already a psychoanalytic unconscious. As Spezzano (1993a) put it:

It is not simply that an idea occurs to the analyst and she then works it over through her conscious awareness of psychoanalytic theory-but also that her immediate experiencing of the patient's communications is performed by an unconscious that has been reshaped by her tripartite (personal analysis, case supervision, and didactic) training [p. 212].

Neither party in the interaction has a privileged and unobstructed view of the other's psyche. There is no unmediated "truth" about the patient. The analyst is always making choices, irreducibly subjective choices, via his or her interpretations and understandings. Among many possibilities, he or she chooses one, and this choice inevitably changes the outcome. There is no "right" choice; each choice that the analyst makes constructs a different ending and a different reality.

The Problem of Validation (footnote 2)

But if the question "How do I know I'm right?" is epistemologically fruitless, how does a constructivist sensibility help an analyst with her or his central clinical concerns: deciding if her or his choices and constructions are attuned to the patient's experience, provide what the patient needs, and address the most pressing problems with which the patient might be unconsciously struggling? In the analyst's mind, in the background, the analyst may well believe that the moment is theoretically "shot through with uncertainty," but the postmodernist never quite makes it clear how this sensibility helps the analyst figure out what to do. And act the analyst must. As Hoffman (1996) has argued, the analyst and patient exist in real time; decisions and choices have to be made and the consequences dealt with. Therefore, the analyst has no choice but to generate and test hypotheses and to develop and use criterion for deciding if the intervention was on the mark. The constructivist might believe, in principle, that the "mark" is not ultimately verifiable in the sense that positivist science would have it. It might be a hermeneutic mark, an intersubjective mark, a co constructed mark, or a mark in analytic space. Nevertheless, it is a psychically real experience for which the analyst is reaching and for which effort he or she needs markers and confirmatory clinical guideposts.

Postmodern theorists have gone out of their way to attempt to address this problem. The issue is whether they have succeeded. For instance, Mitchell (1993) argues that postmodernism is not pure relativism. Using the metaphor of art, he pointed out that there is a clear difference between good and bad representational painting because "the subject matter has a claim on the painting (p. 65). Similarly, the patient's psychological reality, however it is more or less shaped in an intersubjective context, has to have a claim on the analyst's construction of it. Other writers in this tradition are even more emphatic in their rejection of the relativist label or the related charge of solipsism. Hoffman (1991b), for instance, argued against just this implication contained in a critique by Benjamin (1991), asserting:

There is nothing in this position that implies that, as Benjamin puts it, "we should give up reaching for the subjective experience of the other as an outside being." What it does imply is a certain element of doubt as to what constitutes "reaching for" and what constitutes "grasping" at any given moment, doubt that follows directly from the fact that, notwithstanding the asymmetry of the psychoanalytic arrangement, the analyst is always involved as a subject. Moreover, an element of doubt is not the same as total blindness or confusion. It doesn't, however, preclude our being able to "assume that our participation will accentuate the contrast between past expectations and a new shared reality [pp. 542-543].

And, again, in a discussion of the analyst's authority, Hoffman (1996) weighs in against the misrepresentation of constructivism as an "anything goes" philosophy and argues that experience is ambiguous but not amorphous. Experience, he reminds us, has

properties that are amenable to a variety of interpretations, maybe even infinite interpretations, especially if we take into account all the nuances that language and tone make possible. But infinite does not mean unlimited in the sense that anything goes. There are infinite numerical values between numbers 5 and 6. But that range excludes all other numerical values [p. 111].

Clearly, then, some descriptions are better than others. Reality-in this case the patient's psychic reality-somehow has a claim on the analyst's formulations. But the question then arises: how does the analyst decide that the relevant clinical universe is between 5 and 6 and not, say, between 6 and 7? As we "reach for the subjective experience of the other," how do we know if we've touched it? If we are not "totally blind or confused," how do we maximize the degree to which we see clearly?

We are back to the original issue: What questions does postmodern theory ask and which ones does it not? In my reading, postmodem answers to these questions of validation are unsatisfying. Sometimes, for instance, general hermeneutic criteria are invoked. Such criteria include "usefulness" (whether the interpretation generates "useful" conversation or narratives), coherence (whether an interpretation, through redescribing the patient's past and present, connects and makes narratively sensible previously unrelated feelings, thoughts, and memories), and even aesthetic satisfaction (Geha, 1993; Elder, 1994). In general, however, these criteria are rarely operationalized, tending instead to remain rather abstract and philosophical. That is, the authors rarely give examples of how hermeneutic or other nonpositivist validation measures might be used concretely to justify an interpretation, offer corrective feedback to a therapist attempting to tune in to a patient, choose the most salient issue to address, or otherwise guide the therapist's session-to-session or even moment-to-moment attempts to "grasp" the subjective experience of the patient in an attempt to alleviate the latter's suffering.

It is not as if the possibility of doing so is denied. The problem is, rather, that it remains hypothetical. For instance, Stem (1991), speaking approvingly of Gadamer's concept of the hermeneutic circle, even begins to suggest a framework for a rigorous clinical methodology:

All comprehension is a process of projecting partial understandings into fully rounded ones, and then modifying these projections on the basis of what we actually come into contact with in conversation with the other person. In other words, when we understand, we have been able to treat our projections like hypotheses ... [pp. 60-61].

However, the appearance of an interest in hypothesis testing is only that-an appearance. We see little further mention of what it means to "actually come into contact with" something that modifies our hypothesis. I assume that sometimes this "something" is, in part, the patient's observable and inferred responses to our projections. But by not addressing this issue, Stem leaves us with the impression that observable behavior of and in the patient-changes in affect, new memories, symptom resolution, subjective reports of well-being-might or might not be a central validation criteria for making our "partial" understandings a bit less partial, our hypotheses a bit more confident. The problem, again, lies more in what is not said than in what is said.

Instead, Stern (1991) moves in a direction increasingly seen among postmodern thinkers seeking to ground analytic technique in an evidentiary realm unique to a psychoanalytic hermeneutics. The focus on the patient as the key arbiter of our technique tends to yield to a focus on the analyst and her or his subjectivity as of primary importance in the task of improving technique. Stem spoke of the necessity of the analyst's "commitment to his preconception, which means the willingness to give oneself over to a way of seeing" (p. 69). Stern argues that

Gadamer's view does nothing to weaken the case for analytic discipline and education, because to adopt this perspective means placing great stress on the necessity for the analyst to question himself about whatever he takes for granted, to find a way into "seeing what is questionable." And seeing what is questionable requires a consistency of self-reflection that would be next to impossible without careful supervision and a training analysis that touches the analyst in the way he hopes his patients will be touched. To understand someone else requires innocence and openness, which are not only gifts, but accomplishments of education and experience [pp. 76-77].

The analyst's self-reflection and personal experience is increasingly the privileged road to validation as well as understanding. Hoffman (1992) makes this relationship between the analyst's experience and validation explicit in his discussion of a case of Frederickson's (1990) in which the analyst's enactment is justified on the basis of its salutary outcome. Hoffman argues:

It's misleading, however, to judge an action in an absolute way on the basis of what happens after it. The fact is that at the moment of action there is always more than one kind of handwriting on the wall. At that moment, the emotional authenticity of the analyst has to count for something in its own right [p. 296, italics added].

Spezzano (1993b) also expresses this tendency to both acknowledge the importance of validation and shift our attention away from the patient's behavior and toward the analyst's subjectivity. Arguing against Sass's charge of relativism, for instance, Spezzano (1993) says: "My point is precisely that the analytic process and the rational critical discourse that surrounds it are a scientifically self-correcting enterprise, fully capable of generating a solid core of knowledge" (p. 270). However, when one looks to Spezzano's descriptions of how he actually formulates what he does "on the ground" of clinical technique, we see a tendency (in theory) to favor the analyst's subjectivity over observable patient-specific criteria as sources of validation. For instance, in the course of illustrating the technical consequences of the contemporary paradigm with which he is affiliated, Spezzano (1995) tells us:

Rather than always waiting to deliver an interpretive "zinger" to the patient, I, more often than I once did, tell patients about impressions and thoughts I have. I admit that I do not know what to make of them yet and ask them to see if they can use it. Often patients associate to my associations, but even when they do not, I think it is useful to the goal of the patient identifying with the analyzing function of the clinician for me to struggle out loud to make sense of my associations rather than leave the impression of my mind as a flow of sharp insights [p. 43].

In the tradition of analysts who emphasize the patient's "use" of the analyst as a containing (Bion, 1967; Ogden, 1986) or transformational (Bollas, 1987) object, Spezzano reminds us that the medium is the message here. For some or even many patients, one can readily see how the analyst's playful and free-associative responsiveness can communicate salutary meanings and a model of a healthier form of relatedness. For other patients, however, one can imagine that it would not be useful at all. For instance, a patient who had been subtly traumatized by a parent who was playful and affectively spontaneous in a self aggrandizing, flaky, irresponsible, or narcissistic way might experience a therapist who, as like Spezzano suggests above, "struggled out loud to make sense of [his] associations" as pathogenic and continue to do so even if it could be confidently interpreted to him or her that this was a transference distortion. For this patient, a style that is more careful, restrained, deliberate, and even studied might be more conducive to accomplishing the analyst's aim. In my view, it seems useful to strive for a theory of technique in which, to the extent that it is authentically possible (Bader, 1995), we attempt to contribute to the creation of an ambience that is selectively responsive to a patient's needs. And, therefore, in the pursuit of this aim, we would be advised to attempt to define patient-specific criteria for guiding our style as well as interpretive content.

Spezzano might well agree with me; he certainly attempts to be as patient-specific as possible in his work. The question is whether a theory of technique informed by a constructivist epistemology helps us do this in the most efficient and therapeutically productive way possible. I suggest that this epistemology tends to be left at the door of the consulting room, appears only indirectly in the form of the analyst's increased modesty and willingness to be spontaneous, or else functions as an active hindrance to developing the clinical confidence necessary to maximally move the treatment forward.

Spezzano (1995) attempts to operationalize these issues, although I believe the problem of validation remains. He states that the aim of interpretation is

to enlarge the analytic space within which affect is contained and thought about and within which unconscious object relational paradigms become conscious. One interprets when interpreting is likely to have this impact and one refrains from interpreting when interpreting is likely to have the opposite impact [p. 27].

The question is to what extent do the patient's responses influence one's judgment about when to interpret? In my reading, Spezzano's answers do not focus sufficiently on the patient.

Spezzano's aim of "enlarging the analytic space" could be conceptualized, for instance, as having certain "markers," behavioral, verbal, or affective changes that are associated with this aim, markers that could be used as outcome criteria to let Spezzano and us know we are empathically responsive to what the patient needs. Although he no doubt uses such markers in practice, Spezzano does not tend to go in this direction in theory. Like Bion, Bollas, and Ogden, he tends to look often, at least in his theory construction, to the analyst's subjectivity for sources of resistance to an accurate understanding and response to the patient's emerging affects and pathogenic beliefs. Obviously, this is an extremely useful source of information and, to be fair, Spezzano is merely emphasizing, for the sake of contrast, one potential interactive style that a contemporary analyst might display. However, given our collective interest in making our technique patient-specific, in the context of this paradigm the focus on the analyst's psyche as the source of both distortion and validation tends to subtly supplant a rigorous focus on the patient as providing the relevant confirmatory evidence. In this sense, his theory does not do justice to the sense one gets of his highly attuned, flexible, and experience-near clinical practice. Again, I think this is another example of the disjuncture between the current interest in epistemology and clinical practice.

Clinical practice and, in particular, our theories about how people change in analysis have changed over time. As relational theories of technique and cure have come to prominence, the postmodern critique of the alleged objectivity of interpretation might appear to be less urgent and my concerns, therefore, less relevant. However, the constructivist challenge to the analyst's epistemological authority remains, as do the problems inherent in this challenge. For instance, Hoffman (1996) and Mitchell (1993) insist that even though our theory of therapeutic action has changed, our new emphasis on curing through "influence" must still essentially be a constructivist project, that the analyst is also creating what he or she is attuning to, choosing what potentials in the patient to affirm, and constructing some endings for a patient and not others. Relational factors, in other words, occupy the same epistemological position as interpretation. The problem, however, with now making the choice and effects of the analyst's actions and corrective influence also seem intrinsically uncertain is the same as making the choice and effects of interpretation uncertain. In other words, if this argument is meant to mean more than simply "we can never be 100% sure," it leaves open the problem of how we can be more or less sure of what we're doing, only now the problem is transposed onto the dimension of analytic activity, influence, corrective emotional experiences, and the provision of new developmental experiences. My argument is that although Hoffman, Mitchell, and others who believe in the centrality of influence are always presumably attempting to solve this problem "on the ground" by using the patient's responses as evidence, they feel philosophically restrained from making this process too objective, too subject to systematization, too "knowable." They talk like constructivists but act like scientists. Their constructivism hinders their pursuing the question of how the analyst does indeed always influence the patient, how this influence can be assessed, and how it can be more strategically and "accurately" applied.

Confidently assessing the effects of influence, like interpretation, is what gives the postmodemist pause. Because data are ambiguous until interpretively coconstructed, we cannot rely on them for validation. In summarizing what he calls the "limited constructivist" position of analysts like Schafer, Hoffman (1992), although critical of the limits of this position, clearly sympathizes with its distrust of the "facts":

Within this framework, interpretations are suggestions.... This is not to say that one cannot speak of one interpretation fitting the patient's experience more than another. But there is more leeway for a range of interpretations that are persuasive, and it is understood that, inescapably, there is some influence coming from the side of the analyst in deciding what line of interpretation to pursue. The "data," that is, the patient's associations and other aspects of the patient's behavior, cannot decide the issue by themselves [p. 290].

And, yet, I continue to ask, shouldn't these species of "data"-the patient's observable responses-have an important, if not definitive, impact on deciding the issue? If so, how exactly should we use the data to confirm or disconfirm our hypotheses? For instance, in the so-called limited constructivist position that Hoffman describes, if one interpretation "fits the patient's experience more than another," shouldn't we tend to accord it a greater truth value? Or, from another perspective, if one ambience, style, or experience is more corrective than another, as evidenced by the patient's becoming more free, more insightful, or otherwise better, shouldn't we be able to conclude that we have a more accurate and useful understanding of the patient's conflicts than before? Just because "the patient's associations and other aspects of the patient's behavior" are not the sole and absolute authority for judging the validity of an intervention, isn't it at least a powerful first step and certainly preferable to data contradicting an analyst's hypothesis? One might argue that the value of thinking about truth already supposes that there is such a thing and that our task is to get to "it" (see Dunn, 1995). However, even if the operational goals of your interventions were things like fitting the patient's experience, allowing the patient to find her voice, generating interesting conversation, developing an analytic process, telling a useful story, or relieving symptoms, I still suggest that it is useful to generate criteria for knowing you're reaching that goal more or less, with the aim being to reach it more. I realize that few would disagree with this, but I am arguing that it tends to be deemphasized in the postmodern critique.

In fact, this literature can lend itself to the implication that it is arguing that if we can't be 100% right, it is hardly worth trying to be 70% right, or even 20% right. We repeatedly read arguments like the following of Renik's (1993a):

Everything an analyst does in the analytic situation is based upon his or her personal psychology. This limitation cannot be reduced, let alone done away with; we have only the choice of admitting it or denying it. I think we tend to give lip service to the important truth that an analyst cannot, ultimately, know a patient's point of view; an analyst can only know his or her own point of view [p. 561 ].

And even though, ironically, Renik is one of the few postmodern analysts who believes that technique can and should be grounded in a process of hypothesis testing based on patient responses, his epistemological position here, at the least, makes that procedure suspect and, at the most, directs our attention away from the patient and onto the analyst.

The issue is that the postmodern critique helps us see some things more clearly while leaving other concerns out of focus. For instance, on the side of increased clarity, it seems to me that by focusing on the indeterminate and co-constructed nature of clinical "facts," our modesty about what we know and our freedom to more flexibly and personally respond to patients has been enhanced (Rabin, 1995; Renik, 1995b). As both a cause and consequence of our increased modesty, the analyst is more likely to validate the patient's perspective on his or her own psychology, the transference, and the analyst. And by debunking the impossible ideals contained in the old warhorses of abstinence and neutrality, the postmodem theorists have given us a greater freedom to loosen up and be more spontaneously "ourselves." We can more freely commit ourselves to our point of view because we understand that it is only a point of view. These are important advances, as far as they go. The danger is that because this critique is a corrective to skewed traditional images of analytic authority and one person psychology, it can mislead us to see in it a useful description of optimal analytic technique. As Mayer (1996) warns us, "The difficulty with postmodern criticism of theory is that it so easily becomes an attempt at a replacement theory, rather than a critique" (p. 245).

Hoffman (1996) argues that we cannot ever extract our personalities and our influence from the process of affirming the patient's experience and that we should not set up impossible ideals like this to strive for-and ultimately fail-to reach. He uses the analogy that aspiring to walk on water will inevitably interfere with leaming to swim. In my view, however, he misrepresents the alternative. The alternative is not a demoralizing quest for perfection but a process of bringing everything that we know and feel about the patient to make inferences about what she or he is up to; communicate that understanding in words and, if necessary (and it inevitably is), actions; and use patient-specific and observable criteria for telling us if we are on the right track. That "track" involves, at times, the patient acquiring insight, at other times mitigating his or her symptoms, or otherwise moving forward in the therapeutic direction that we have inferred the patient wants to go. Having something impossible to shoot for-perfect understanding-is not self-defeating at all. It is more like comparing your golf swing with that of a well-known professional: identifying differences, attempting corrections, noting results, comparing again. You may never be Jack Nicklaus, but you will probably lower your handicap.

The Possibility of Accurate Understanding

As I have been attempting to argue, I think that the applicability of the postmodern critique is limited by virtue of its inattention to the fact and possibility of reasonably accurate intersubjective understanding in the clinical process. Such understanding, of course, is an aspect of all relationships. There are many moments when one can say, with relative certainty, how one's words or behavior will affect another person. There are people with whom I am intimate whose psychological vulnerabilities, dispositions, and reflexes seem very clear to me. In certain moments with these people, I can predict with a very high degree of accuracy if something I say or do will hurt, please, give a feeling of pride, comfort, or elicit an angry response. The argument that subjectivity is inherently indeterminate, contingent, or otherwise epistemologically different from, say, a physical reality-although obviously true on its face-is not grounded enough in the phenomenology of everyday life.

Instances of our striving successfully for understanding are ubiquitous. Consider for figure out what their children are feeling or needing. The baby cries. Is she or he wet or hungry? The parent feeds the baby. The baby refuses to eat and continues to cry. The parent changes the diaper. The baby stops crying. The parent has tested a hypothesis and used behavioral criteria to decide that the hypothesis was correct. Based on repeated accurate intersubjective a moment how parents occurrences of this "experiment in nature," the parent learns to differentiate the "I'm wet" cry from the "I'm hungry" cry. The next time the baby is wet, the parent will be more likely to correctly respond immediately. By correcdy, I mean in a way that relieves the baby's distress. The postmodern argument that defining truth by reference to what "works" is profoundly different from defining truth by reference to what is "there" might be of some philosophical interest but of little use in everyday life or everyday clinical practice. In the current example, for instance, the aim is to help the baby, and the only means to this end lies in accurate attunement, aided immeasurably by this kind of common, low-level, informal, experimental design.

My purpose in using such a mundane, everyday example of the ubiquitous exercise of inference and validation is not because the postmodernists would deny that these interactions occur but that their epistemological interests in ambiguity leads them to remain somewhat abstract and general in relation to the challenge of clarification and validation. And, yet, as the evolutionary psychologists have told us, human beings are preadapted to use accurate perception to understand, manipulate, predict, and even control the human reality on which they are dependent for survival (Stavin and Kriegman, 1992). In other words, it is highly adaptive to be able to correctly "read" the intentions and feelings of the Other. As Kriegman (1996) puts it, "if we define science in a very simple way-as experimentation and observation designed to achieve attitudes and understandings about the world that lead to accurate prediction and control of events in the worldthen humans can be seen to be natural scientists" (p. 23).

So, too, in clinical work, the therapist can usefully act like a scientist while still doing justice to the constructivist critique of a rigid empiricism. For instance, I conceptualize my own approach as involving generating hypotheses about my patients. Based on a hypothesis derived from my picture of the patient and our relationship, his or her past, and my own psychological experience, I formulate an intervention that I implicitly predict will have a certain effect. The intervention may be verbal or may involve an attitude or even action. The effect might be a small, intraanalytic response such as insight, anxiety reduction, subjective agreement, greater affective freedom, or movement forward toward a formerly forbidden goal. If it does have the intended effect, my hypothesis is strengthened. If not, then the hypothesis might Michael J. Bader need revision. By strengthened, I mean that it is now a bit more likely that my hypothesis has described a process, structure, fantasy, or even lawful relationship involving the patient's psychology and experience of the world. It does not mean I have definitely discovered "it," the "thing itself," the "essence" of the patient's psychological Being. But it may well mean that I have come a little bit closer to "it." What I have discovered is asymptotic to the truth. Operating on the basis that this is "true" enhances my current and future clinical efficacy. Arguments that there is a significant and clinically relevant difference between "operating on the basis of a presumed truth and believing that one can more or less, at crucial moments, "know" the truth become specious, in my view, unless it can be powerfully demonstrated that this epistemological difference in the mind of the analyst generates important clinical differences.

For instance, a female patient who had been in therapy with me for four months told me that scheduling changes at work made continuing with me impossible for the time being. She told me she had benefited from our work and felt that it would be better if she tried to fly by herself for a while. She had a history of relationships with possessive and controlling men. Although I knew little about her mother, her father fit the possessive/controlling male mold as well. I was, therefore, a bit more cautious than usual in taking any immediate stand on the issue of her interrupting her therapy. I tried to explore all sides of the issue. The patient began to get worse. She became more rigid, less psychologically minded, and more detached from her self and from me. She could not reflect on this shift. Based on a few clues that she had given me about her mother-depictions of a woman who seemed rather detached and narcissistic-and in response to my observation that she was shutting down, apparently in response to my not taking a stand, I developed the hypothesis that my apparent neutrality in first exploring her wish to stop without taking a stand on it was experienced as a rejection and reevoked the feelings of hurt she had endured at the hands of a disinterested mother. I, therefore, explained this to her and decided to state clearly that in my opinion, she was not at all ready to leave, that I thought she should stay and continue our good work, and that I would go out of my way, if necessary, to see her at her convenience. She responded by beginning to tear up and soon acknowledged that she was quite relieved, that she had not really wanted to quit but felt that there would be no way I could or would want to accommodate her. She had memories about her mother in which the patient felt rejected by the mother who always seemed preoccupied with her own troubles and seemed hardly to know the daughter was there, much less in need.

I took this as confirmation that my hypothesis about the meaning of our interactions around her wish to stop was probably accurate. This hypothesis of mine, now strengthened, was that the patient was struggling at that moment mainly with rejection, not with autonomy as I first thought, and thus my initial "neutrality" was perceived as a subtle rejection. When I changed my hypothesis, I was able to attune myself better to what the patient was feeling and needing at the moment. The patient then felt safer and was able to begin to face some of the rejection and sadness in her life. And this hypothesis now became a powerful tool for dealing with the patient in the future. My picture of the patient's mind and the prominence of rejection in that mind was strengthened, and this increased clarity was available to be used in the future with therapeutic effect. Although I might have been wrong when I proceeded as if rejection were the main issue, I felt legitimately more confident that I was right.

Weiss (1993) has studied the analytic process extensively and claims to have demonstrated across a broad range of studies that hypothesis testing and validation can be reliably employed in a manner that is highly patient-specific and that does not irrationally privilege the analyst's rationality. For Weiss, the patient comes into analysis motivated to overcome his or her conflicts and works to do so in a planful, albeit often unconscious, way. The therapist's task is to discern both the nature of the patient, conflict and his or her mode of unconscious mastery and to facilitate it. Weiss believes that his theory of pathogenesis enables him to make predictions about the consequences of the analyst's attitudinal and interpretive activity. His model of psychopathology generates criteria that indicate whether the patient's plan is proceeding successfully or being derailed. He argues that his research has demonstrated that the patient's response to interventions that facilitate the latter's attempts at mastery is often immediate and discernible. The patient's anxiety will decrease, repressions of various kinds will start to lift, new material will emerge, symptoms will begin to recede, certain developmental tasks will be taken on, and so on. (for various studies of the operational use of process and outcome criteria, see Weiss and Sampson, 1986). In Weiss's theory, then, the patient's subjectivity, although unknowable in a complete sense, has a directionality, a course, a planfulness inherent in it that the analyst can sometimes infer with considerable accuracy and therapeutic benefit (for a similar discussion of planfulness, but from an evolutionary perspective, see Kriegman and Slavin, 1989). Because the analyst has certain observable criteria to use as a barometer of whether the analyst's interventions are furthering the patient's growth or hindering it, the analyst's psychology-although inevitably a source of bias and information-is not felt to be a primary barrier to or necessarily confounding of knowledge. In Weiss's theory, the analyst's psychological responses are valuable sources of information and clearly contribute to both the patient's therapeutic and pathogenic experiences of the relationship, but because there is a discernible "track" that the patient is trying to be on, the analyst's task is only to do whatever is necessary to help the patient do that. For Weiss, the analyst can pass or fail a "test" in words, actions, or attitudes, and the analyst's countertransference is never a privileged means to this end.

Finally, one can believe in the ubiquity of mutual influence and the analyst's ever-present subjectivity, which are central postmodern tenets, and still focus on defining operational criteria for judging the validity of one's interpretations. Renik (1993x), for instance, has written extensively on the need to deconstruct the analyst's privileged authority to say what is going on in the patient, the transference, the interaction, and the analyst. He has elaborated on the concept of the analyst's "irreducible subjectivity" in determining what he or she does and has inveighed against the analyst's covertly inviting the patient to idealize him or her as inherently wise. Yet Renik (1995x) also argues forcefully:

When an analyst identifies a resistance, he or she forms a hypothesis about an analysand's psychology.... I look at interpretation of resistance as the technique an analyst uses to test hypotheses about motivations interfering with an analysand's self-awareness.... The analysand's responses to an interpretation, the further material the analysand brings forth, are data that either confirm or disconfirm the hypothesis.... The process of hypothesis testing, via interpretation of resistance, even if complex and roundabout, is an empirical one. Analysand and analyst both make observations and inferences based upon them [pp. 87-88]

For Renik, the postmodern epistemology enters through his contention that when the analyst offers his or her understanding to the patient, he or she should know and even convey, explicitly or in his or her manner, that this understanding is simply an opinion,, an opinion based on mutually observable "facts of observation," and that it does not in any way invoke a special authority to read the patient's mind. Having said this, Renik is unique among the postmodernists in his emphasis on the centrality of the scientific method of hypothesis testing in furthering his analytic and therapeutic aim-namely, to produce a therapeutic effect by increasing the patient's self-awareness. His epistemology makes him far more modest than Weiss about what the analyst can really know about the workings of the patient's mind, but his way of getting at or reaching what the analyst can know is similar. (footnote 3)

The postmodern revolution in epistemology has decentered us and taken away our comforting beliefs in objectivity and science. As we lose our balance we tend to accuse these contemporary critics of abandoning us to chaos and demand that they give us something to hold on to. Yet, it might be argued that it is unfair to expect something of the postmodernists that we, as a field, have not fully worked out. The person who yells "fire" is not obliged also to tell us how to get out of danger or build a more fireproof building. We should not get mad at the messenger. It is the message that disturbs us, the message that we are more confused, less confident, less authoritative, less rational, less perceptive, less insightful, and less smart than we think we are.

The problem, as I see it, is that this new critique of a certain kind of scientific rationality tends to claim more for itself that it can deliver. It does more than yell "fire." It suggests a new way of listening, suggests a new way of knowing, and hints at a different view of the mind, with different assumptions about human nature. Although its basic epistemological position can subsume a wide variety of theories of change (e.g., Hoffman accepts the inevitable centrality of the analyst's unanalyzed influence, whereas Renik does not), the postmodern sensibility still aims to be a practical guide to clinical technique. Thus, when the postmodernist says, "Look, I know that everything isn't relative and that some interventions comprehend and touch the patient's separate experience more than others," we are led to believe that this theorist has a theory of validation that we can use. We are entitled to ask for it. Unfortunately, such a theory is too often missing.

The Possibility of Accurate Understanding

As I have been attempting to argue, I think that the applicability of the postmodern critique is limited by virtue of its inattention to the fact and possibility of reasonably accurate intersubjective understanding in the clinical process. Such understanding, of course, is an aspect of all relationships. There are many moments when one can say, with relative certainty, how one's words or behavior will affect another person. There are people with whom I am intimate whose psychological vulnerabilities, dispositions, and reflexes seem very clear to me. In certain moments with these people, I can predict with a very high degree of accuracy if something I say or do will hurt, please, give a feeling of pride, comfort, or elicit an angry response. The argument that subjectivity is inherently indeterminate, contingent, or otherwise epistemologically different from, say, a physical reality-although obviously true on its face-is not grounded enough in the phenomenology of everyday life.

Instances of our striving successfully for understanding are ubiquitous. Consider for figure out what their children are feeling or needing. The baby cries. Is she or he wet or hungry? The parent feeds the baby. The baby refuses to eat and continues to cry. The parent changes the diaper. The baby stops crying. The parent has tested a hypothesis and used behavioral criteria to decide that the hypothesis was correct. Based on repeated accurate intersubjective a moment how parents occurrences of this "experiment in nature," the parent learns to differentiate the "I'm wet" cry from the "I'm hungry" cry. The next time the baby is wet, the parent will be more likely to correctly respond immediately. By correcdy, I mean in a way that relieves the baby's distress. The postmodern argument that defining truth by reference to what "works" is profoundly different from defining truth by reference to what is "there" might be of some philosophical interest but of little use in everyday life or everyday clinical practice. In the current example, for instance, the aim is to help the baby, and the only means to this end lies in accurate attunement, aided immeasurably by this kind of common, low-level, informal, experimental design.

My purpose in using such a mundane, everyday example of the ubiquitous exercise of inference and validation is not because the postmodernists would deny that these interactions occur but that their epistemological interests in ambiguity leads them to remain somewhat abstract and general in relation to the challenge of clarification and validation. And, yet, as the evolutionary psychologists have told us, human beings are preadapted to use accurate perception to understand, manipulate, predict, and even control the human reality on which they are dependent for survival (Stavin and Kriegman, 1992). In other words, it is highly adaptive to be able to correctly "read" the intentions and feelings of the Other. As Kriegman (1996) puts it, "if we define science in a very simple way-as experimentation and observation designed to achieve attitudes and understandings about the world that lead to accurate prediction and control of events in the worldthen humans can be seen to be natural scientists" (p. 23).

So, too, in clinical work, the therapist can usefully act like a scientist while still doing justice to the constructivist critique of a rigid empiricism. For instance, I conceptualize my own approach as involving generating hypotheses about my patients. Based on a hypothesis derived from my picture of the patient and our relationship, his or her past, and my own psychological experience, I formulate an intervention that I implicitly predict will have a certain effect. The intervention may be verbal or may involve an attitude or even action. The effect might be a small, intraanalytic response such as insight, anxiety reduction, subjective agreement, greater affective freedom, or movement forward toward a formerly forbidden goal. If it does have the intended effect, my hypothesis is strengthened. If not, then the hypothesis might Michael J. Bader need revision. By strengthened, I mean that it is now a bit more likely that my hypothesis has described a process, structure, fantasy, or even lawful relationship involving the patient's psychology and experience of the world. It does not mean I have definitely discovered "it," the "thing itself," the "essence" of the patient's psychological Being. But it may well mean that I have come a little bit closer to "it." What I have discovered is asymptotic to the truth. Operating on the basis that this is "true" enhances my current and future clinical efficacy. Arguments that there is a significant and clinically relevant difference between "operating on the basis of a presumed truth and believing that one can more or less, at crucial moments, "know" the truth become specious, in my view, unless it can be powerfully demonstrated that this epistemological difference in the mind of the analyst generates important clinical differences.

For instance, a female patient who had been in therapy with me for four months told me that scheduling changes at work made continuing with me impossible for the time being. She told me she had benefited from our work and felt that it would be better if she tried to fly by herself for a while. She had a history of relationships with possessive and controlling men. Although I knew little about her mother, her father fit the possessive/controlling male mold as well. I was, therefore, a bit more cautious than usual in taking any immediate stand on the issue of her interrupting her therapy. I tried to explore all sides of the issue. The patient began to get worse. She became more rigid, less psychologically minded, and more detached from her self and from me. She could not reflect on this shift. Based on a few clues that she had given me about her mother-depictions of a woman who seemed rather detached and narcissistic-and in response to my observation that she was shutting down, apparently in response to my not taking a stand, I developed the hypothesis that my apparent neutrality in first exploring her wish to stop without taking a stand on it was experienced as a rejection and reevoked the feelings of hurt she had endured at the hands of a disinterested mother. I, therefore, explained this to her and decided to state clearly that in my opinion, she was not at all ready to leave, that I thought she should stay and continue our good work, and that I would go out of my way, if necessary, to see her at her convenience. She responded by beginning to tear up and soon acknowledged that she was quite relieved, that she had not really wanted to quit but felt that there would be no way I could or would want to accommodate her. She had memories about her mother in which the patient felt rejected by the mother who always seemed preoccupied with her own troubles and seemed hardly to know the daughter was there, much less in need.

I took this as confirmation that my hypothesis about the meaning of our interactions around her wish to stop was probably accurate. This hypothesis of mine, now strengthened, was that the patient was struggling at that moment mainly with rejection, not with autonomy as I first thought, and thus my initial "neutrality" was perceived as a subtle rejection. When I changed my hypothesis, I was able to attune myself better to what the patient was feeling and needing at the moment. The patient then felt safer and was able to begin to face some of the rejection and sadness in her life. And this hypothesis now became a powerful tool for dealing with the patient in the future. My picture of the patient's mind and the prominence of rejection in that mind was strengthened, and this increased clarity was available to be used in the future with therapeutic effect. Although I might have been wrong when I proceeded as if rejection were the main issue, I felt legitimately more confident that I was right.

Weiss (1993) has studied the analytic process extensively and claims to have demonstrated across a broad range of studies that hypothesis testing and validation can be reliably employed in a manner that is highly patient-specific and that does not irrationally privilege the analyst's rationality. For Weiss, the patient comes into analysis motivated to overcome his or her conflicts and works to do so in a planful, albeit often unconscious, way. The therapist's task is to discern both the nature of the patient, conflict and his or her mode of unconscious mastery and to facilitate it. Weiss believes that his theory of pathogenesis enables him to make predictions about the consequences of the analyst's attitudinal and interpretive activity. His model of psychopathology generates criteria that indicate whether the patient's plan is proceeding successfully or being derailed. He argues that his research has demonstrated that the patient's response to interventions that facilitate the latter's attempts at mastery is often immediate and discernible. The patient's anxiety will decrease, repressions of various kinds will start to lift, new material will emerge, symptoms will begin to recede, certain developmental tasks will be taken on, and so on. (for various studies of the operational use of process and outcome criteria, see Weiss and Sampson, 1986). In Weiss's theory, then, the patient's subjectivity, although unknowable in a complete sense, has a directionality, a course, a planfulness inherent in it that the analyst can sometimes infer with considerable accuracy and therapeutic benefit (for a similar discussion of planfulness, but from an evolutionary perspective, see Kriegman and Slavin, 1989). Because the analyst has certain observable criteria to use as a barometer of whether the analyst's interventions are furthering the patient's growth or hindering it, the analyst's psychology-although inevitably a source of bias and information-is not felt to be a primary barrier to or necessarily confounding of knowledge. In Weiss's theory, the analyst's psychological responses are valuable sources of information and clearly contribute to both the patient's therapeutic and pathogenic experiences of the relationship, but because there is a discernible "track" that the patient is trying to be on, the analyst's task is only to do whatever is necessary to help the patient do that. For Weiss, the analyst can pass or fail a "test" in words, actions, or attitudes, and the analyst's countertransference is never a privileged means to this end.

Finally, one can believe in the ubiquity of mutual influence and the analyst's ever-present subjectivity, which are central postmodern tenets, and still focus on defining operational criteria for judging the validity of one's interpretations. Renik (1993x), for instance, has written extensively on the need to deconstruct the analyst's privileged authority to say what is going on in the patient, the transference, the interaction, and the analyst. He has elaborated on the concept of the analyst's "irreducible subjectivity" in determining what he or she does and has inveighed against the analyst's covertly inviting the patient to idealize him or her as inherently wise. Yet Renik (1995x) also argues forcefully:

When an analyst identifies a resistance, he or she forms a hypothesis about an analysand's psychology.... I look at interpretation of resistance as the technique an analyst uses to test hypotheses about motivations interfering with an analysand's self-awareness.... The analysand's responses to an interpretation, the further material the analysand brings forth, are data that either confirm or disconfirm the hypothesis.... The process of hypothesis testing, via interpretation of resistance, even if complex and roundabout, is an empirical one. Analysand and analyst both make observations and inferences based upon them [pp. 87-88]

For Renik, the postmodern epistemology enters through his contention that when the analyst offers his or her understanding to the patient, he or she should know and even convey, explicitly or in his or her manner, that this understanding is simply an opinion,, an opinion based on mutually observable "facts of observation," and that it does not in any way invoke a special authority to read the patient's mind. Having said this, Renik is unique among the postmodernists in his emphasis on the centrality of the scientific method of hypothesis testing in furthering his analytic and therapeutic aim-namely, to produce a therapeutic effect by increasing the patient's self-awareness. His epistemology makes him far more modest than Weiss about what the analyst can really know about the workings of the patient's mind, but his way of getting at or reaching what the analyst can know is similar. (footnote 3)

The postmodern revolution in epistemology has decentered us and taken away our comforting beliefs in objectivity and science. As we lose our balance we tend to accuse these contemporary critics of abandoning us to chaos and demand that they give us something to hold on to. Yet, it might be argued that it is unfair to expect something of the postmodernists that we, as a field, have not fully worked out. The person who yells "fire" is not obliged also to tell us how to get out of danger or build a more fireproof building. We should not get mad at the messenger. It is the message that disturbs us, the message that we are more confused, less confident, less authoritative, less rational, less perceptive, less insightful, and less smart than we think we are.

The problem, as I see it, is that this new critique of a certain kind of scientific rationality tends to claim more for itself that it can deliver. It does more than yell "fire." It suggests a new way of listening, suggests a new way of knowing, and hints at a different view of the mind, with different assumptions about human nature. Although its basic epistemological position can subsume a wide variety of theories of change (e.g., Hoffman accepts the inevitable centrality of the analyst's unanalyzed influence, whereas Renik does not), the postmodern sensibility still aims to be a practical guide to clinical technique. Thus, when the postmodernist says, "Look, I know that everything isn't relative and that some interventions comprehend and touch the patient's separate experience more than others," we are led to believe that this theorist has a theory of validation that we can use. We are entitled to ask for it. Unfortunately, such a theory is too often missing.

Postmodern Sensibilities and the Turn From Therapeutics

The postmodern sensibility is grounded in the notion of paradox. Paradox, however, can lead to progressive or regressive resolutions. For instance, the radical and progressive edge of contemporary constructivism lies in its sponsoring a new freedom of thought and action in analysts who too often have been pressured into a stiff abstinence by their psychoanalytic superegos. Because we are now not omniscient, do not have to deliver interpretive "zingers" to our patients, are inevitably biased and human, and do not have a corner on truth, we can relax, get more involved, and take our patients and their point of view more seriously. However, the frequent insistence on the ambiguity and near, infinite complexity of the intersubjective and interactive fields in analysis and on the limits of understanding can also sponsor an unproductive confusion and pessimism about our therapeutic task. On one hand, then, it has opened us up to the existence of multiple realities, multiple narratives, and multiple "truths" and, consequently, increased our modesty and flexibility. On the other hand, by drawing our attention away from a focus on the patient for validating our technique, either because of the presumed inherent subjectivity of interpretation or because the focus is often on the analyst's mind, the effect of this philosophical trend can potentially make us less confident that we can do our job-understanding and helping the patient change. There is a tendency, instead, to fetishize uncertainty, idealize ambiguity, and admire complexity. We are free to be expressive but reminded that we cannot judge the effects of that expressiveness accurately.

As a consequence of having to attend to multiple levels, feedback loops, and concentric circles of interpretive activity, we are led to be suspicious of using data such as the patient's therapeutic progress as a central marker of the accuracy of these interpretations. Unfortunately, in this respect, postmodern and mainstream analysts are fellow travelers. We have always struggled with a tension between our therapeutic aims and theory of technique (Bader, 1994). For the classical or mainstream analyst, therapeutic progress, although a collective aim and personal desire, tends to be viewed as an indirect outcome of analytic activity rather than its central operational goal. By indirect rather than direct goal, I am referring to the historical bias in psychoanalysis against therapeutic "zeal" and the tendency to feel, instead, that a meticulous focus on the intraanalytic resistances or the transference-countertransference field should ideally generate the best therapeutic outcome without our directly trying to do so. Emphasizing therapeutic aims has often been viewed with suspicion as countertransference-based ambition, contaminating the optimal analytic attitude which is to analyze "without memory and desire" (Bion, 1967, p. 272). Understanding, an analytic aim, has historically been counterposed to helping, a psychotherapeutic one. The establishment of an "analytic process" often tends to be elevated over therapeutic results as our operational goal. The traditional psychoanalytic distrust of using symptom relief as a guide to technique is part of a more general skepticism about relying on patient-specific outcome criteria for clinical confirmation. In other words, the classical analyst is instructed to view whatever the patient says in response to interpretations as always a less than definitive and usually unreliable guide to whether the analyst is "right." The ebb and flow of the patient's therapeutic progress is a particularly suspect subset of data when it comes to establishing validity in classical technique. This has served to weight our clinical theory in the direction of received authority. In other words, all we can confidently do is proceed "analytically"-good outcomes will likely follow without our "trying" to produce them.

Born of Freud's and our need to differentiate analysis from other therapies and modes of healing, (footnote 4) the historic mandate to eschew a direct interest in symptom relief while holding analysis out as the most radical and enduring treatment is problematic today. This "tilt" away from therapeutics is particularly maladaptive in today's climate of skepticism about the efficacy of psychoanalysis and the increasing demands for our field to demonstrate cost efficiency in the context of a modern era of managed care and biological psychiatry. Unfortunately, although modem analysts with a constructivist sensibility have certainly rejected appeals to analytic authority and are more prepared to flexibly attune themselves to the patient's subjectivity, their emphasis on epistemology inadvertently continues this antitherapeutic bias in psychoanalytic theorizing. The constructivist sensibility adds its own unique twist to our theoretical tradition of privileging process over outcome. The message might be read as implying the following: "Because of the intrinsic nature of the process, a process coconstructed by two interpreting subjects inherently limited in their ability to fully understand themselves or each other, we cannot make claims to definitively understanding, much less curing, patients. We should not fool ourselves into thinking that we can use observable patient-specific responses as reliable feedback to sharpen and improve our technique. This should not and cannot be our primary focus because subjective reality is inherently ambiguous and our ability to find reliable regularities, lawful relationships, and useful validation criteria is limited."

This is an appealing metamessage. It takes the analyst off the hook of trying to be right and offers her or him a comforting rationalization. But, in so doing, it participates in the antitherapeutic bias that has tended to mark psychoanalysis and that justifies its results, when they are poor, and its therapeutic passivity by references to the near-infinite complexity of the unconscious, the transference-countertransference matrix, and the psychoanalytic process in general. To be sure, the contemporary emphasis on psychoanalysis as a "dialogic community" or "conversation" can read at times like a positive message urging us to hold our heads up high in defense of our unique contribution to the human sciences and not to defer to outside discourses and "experts" such as neurobiologists, positivist scientists, empirical outcome studies, or infancy researchers (for an example of this kind of argument at its best, see Spezzano, 1993a). Unfortunately, the danger is that this can also sound as if one were telling the passengers on the Titanic not to worry about that iceberg because they are on the most beautiful ship of its class!

In my view, this aspect of the postmodern sensibility is a symptom of the pervasive underlying pessimism about our roles as change-agents, as healers, as engaged exclusively in an activity the only reasonable purpose of which is to help cure people of their suffering. This pessimism, which has percolated within psychoanalytic theory and practice since its inception, has gained a particular salience today in the context of the decline of institutional and ideological support for psychotherapy and the broader culture of conservatism and cynicism about our ability to radically alleviate human suffering in general. In response to this pessimism, a postmodern epistemology that reminds us of the impossibility of discovering essences is a comfort. In response to our frustration at poor therapeutic results, an attitude that underlines complexity and uncertainty is a tonic. In response to our declining status among mental health providers who promise a cheaper, more efficient, and thorough product, the complicated and abstruse flavor of postmodern language and writing is personally and professionally restorative. In this way, we accommodate to the prevailing ethos that is objectively undermining our position while reinstating ourselves as worthy in our own imaginations.

Therapeutic Pessimism and the Retreat From Social Activism

Psychoanalysis - its theory-building, deconstruction, and evolutiondoes not exist apart from the social context and intellectual Zeitgeist surrounding it. The argument advanced here that the emerging constructivist sensibility in psychoanalysis has both liberated us from an authoritarian technique and sponsored a retreat from therapeutics has also been made in response to the broader postmodern march through academia Uacoby, 1987; Lasch, 1995; Gitlin, 1996). This march, beginning in the 1960s and 1970s, can be seen as the intellectual expression of the social ferment of the times, a ferment in which various radical challenges to the status quo-the antiwar, student, civil rights, Black nationalist, and feminist movements-began to make inroads into changing traditional structures of power and consciousness. Postmodern perspectives in the academy began to proliferate, and in departments ranging from architecture to English, traditional canons and authorities were challenged. The postmodernisms pointed out how the traditional Enlightenment ideals of Reason, Progress, and Equality were riddled with hypocrisy and contradiction because they excluded the interests and voices of minorities and women. They challenged the glorification of objectivity and science as universal virtues and showed how these values covered over prejudice and supported the interests of ruling elites. Their aim was to deconstruct beliefs in universal truth found in various disciplines and proposed, instead, a more perspectivel, democratic, and relativist approach. As Lehman (1991) put it in his analysis of the historical appeal of deconstruction, "Deconstruction capitalizes on the crisis of authority and the crisis of faith; it proposes a radical skepticism that suits the temper of a generation that came of age amidst credibility gaps, hype campaigns, and spin doctors" (p. 70). The postmodern impulse in psychoanalysis clearly flows from the same intellectual wellsprings and had a similar effect in provoking a reconsideration among analysts of so-called universal truths about psychology and therapy and an appreciation of the subjective and biased interests that lie behind claims to objectivity.

But the world changed, and so did postmodernism. By 1975, progressive social change movements had been defeated or were politically enervated. Our society had begun to turn more conservative (Edsall and Edsall, 1991; Derber, 1992; Wallis, 1994; Lerner, 1996). Unhinged from any effective public political movement, postmodern academics turned their energies toward more conventional uses: publishing, getting tenure, professional status. This narrowing and professionalization of the academic lives of former radicals mirrored the retreat from practical political engagement in the culture as a whole. (footnote 5)

Postmodernism began to be so estranged from social practice that its critical edge became blunted, its discourse more abstract. Terry Eagleton wrote: Poststructuralism "was a product of the blend of euphoria and disillusionment, liberation and dissipation, carnival and catastrophe which was 1968. Unable to break the structure of state power, poststructuralism found it possible instead to subvert the structure of language. Nobody, at least, was likely to beat you over the head for doing so" [quoted in Lehmann, 1991, p. 73].

Born of a desire to challenge traditional authority, democratize education, and provide a justification for the various liberatory practices in the 1960s and 1970s, important elements of postmodern theory have become rarefied, specialized, and politically irrelevant. It is not that the liberal and radical intellectuals who went into academia are no longer personally liberal or even radical. It is more that in the context of an enervated and dissipated left or liberal movement, politically apathetic and economically worried student bodies, and a prevailing conservative cynicism about the ideals of the past, these former radicals have taken to developing theory for its own sake and not for the sake of affecting or inspiring real social change. Speaking of deconstruction, Eagleton described this move toward political impotence: "[Deconstruction] ... is mischievously radical in respect of everyone else's opinions, able to unmask the most solemn declarations as mere dishevelled plays of signs, while utterly conservative in every other way. Since it commits you to affirming nothing, it is as injurious as blank ammunition" (Lehmann, 1991, p. 74).

The postmodern emphasis in psychoanalysis on epistemology, initially offering a useful corrective to a rigid and extreme form of positivism in classical technique, now risks firing blanks in the project of generating useful principles of technique that help the clinician "on the ground." The refreshing and liberating ethos of modesty, spontaneity, and respect for the patient's subjectivity that infuses the constructivist sensibility makes the clinical encounter more human, but its emphasis on epistemology makes it difficult to concretely translate this spirit into achieving better results. Psychoanalysis needs to apply itself to the scientific and practical task of getting better results and of describing how we get there. Curing patients is our equivalent of the activist changing the world. The postmodern turn in analysis, like the postmodern trend in academia, began with a critical thrust but has retreated from the clinical trenches into a relatively academic discourse with questionable applicability. Instead of responding to the attacks on psychotherapy and psychoanalysis with a renewed attempt to show concretely how we practically help people and how our theory can generate principles of technique that can be refined to help people better, we are tending to go the way of the postmodern academic and lose ourselves in elegant but potentially solipsistic theory. Gitlin (1996), speaking of the postmodern retreat in academia from politics, suggested that "while the Right was occupying the heights of the political system, the assemblage of groups identified with the Left were marching on the English department" (p. 148). We as analysts have to be careful that we do not leave ourselves open to a similar criticism that is, that we are more interested in what we cannot know in our attempts to be helpful than in what we can know.

  • Selfishness and the Politics of Meaning, vol. 13, no. 3, 1998.
  • You've Lost That Lovin' Feeling: The Problem of Sexual Boredom, vol. 13, no. 2, 1998.
  • Post-Modernism and Psychoanalysis: Fiddling While Rome Burns, vol. 12, no. 2, 1997.
  • Why Can't the Left Take Yes for an Answer? vol. 11, no. 5, 1996.
  • The Psychodynamics of Cynicism, vol. 11, no. 3, 1996.
  • Bruce Springsteen, Tom Joad, and the Politics of Meaning, vol. 11, no. 2, 1996.
  • Can We Still be Selfish After the Revolution: Sexual Ruthlessness, vol. 10, no. 2, 1995.
  • Shame and the Resistance to Spirituality, vol. 9, no. 6, 1994.
  • Helping the Patient Get Better: Psychoanalysis, Optimism, and Social Change, vol. 9, no. 3, 1994.
  • Psychoanalytic Update: The Old Psychoanalysis Just Ain't What It Used to Be, vol. 8, no. 3, 1993.
  • Psychoanalytic Update: Social Constructivism, vol. 8, no. 1, 1993 (with Kim Chernin).
  • Is Psychiatry Going Out of Its Mind? vol. 4, no. 4, 1989.
  • The Nature of Mental Illness: A Response to Reginald Zelnik's 'On Schizophrenia, Reductionism, and Family Responsibility,' 1990.
  • Looking for Addictions in All the Wrong Places, vo. 3, no. 6, 1988.
  • The Wooly Mammoth Fights Back, vol. 2, no. 2, 1987.

REFERENCES

  • Bader, M. J. (1994), The tendency to neglect therapeutic aims in psychoanalysis. Psychoanal. Quart., 63:246-271.
  • (1995), Authenticity and the psychology of choice in the. analyst. Psychoanal. Quart., 64:282-305.
  • Benjamin, J. (1991), Commentary on Irwin Hoffman's discussion: "Toward a social constructivist view of the psychoanalytic situation." Psychoanal. Dial., 1:535-544.
  • Bion, W. R. (1967), Notes on memory and desire. Psychoanal. Forum, 2:272-273; 279-280.
  • Bollas, C. (1987), The Shadow of the Object: Psychoanalysis of the Unthought Known. New
    York: Columbia University Press.
  • Cushman, P. (1995), Constructing the Self, Constructing America. A Cultural History of Psychotherapy. New York: Addison-Wesley.
  • Derber, C. (1992), Money, Murder, and the American Dream: Wilding from Wall Street to Main Street. London: Faber & Faber.
  • Dunn, J. (1995), Intersubjectivity in psychoanalysis: a critical review. Internat. J. Psycho-Anal., 76:723-738.
  • Edsall, T. B. & M. D. (1991), Chain Reaction: The Impact of Race, Rights, and Taxes on American Politics. New York: Norton.
  • Elder, C. (1994), The Grammar of the Unconscious: The Conceptual Foundations of Psychoanalysis. University Park, PA: Pennsylvania State University Press.
  • Elliot, A. & Spezzano, C. (1996), Psychoanalysis at its limits: Navigating the postmodern turn. Psychoanal. Quart., 65:52--83.
  • Frederickson, J. (1990), Hate in the countertransference as an empathic position. Contemp. Psychoanal., 26:479-496.
  • Geha, R. (1993), "transferred fictions. Psychoanal. Dial., 3:209-244.
  • Gitlin, T. (1996), The Twilight of Common Dreams: Why America Is Wracked by Culture Wars. New York: Holt.
  • Greenberg, J. (1991), Countertransference and reality. Psychoanal. Dial., 1:52-73.
  • Hoffman 1. Z. (1983), The patient as interpreter of the analyst's experience. Contemp. Psychoanal., 19:389-422.
  • (1991a), Discussion: Toward a social -constructivist view of the psychoanalytic situation. (discussion of papers by I-. Aron, A. Modell, & J. Greenberg.) Psychoanal. Dial., 1:74-105.
  • (1991b), Reply to Benjamin. Psychoanal. Dial., 1:535-544.
  • (1992), Some practical implications of a social constructivist view of the psychoanalytic situation. Psychoanal. Dial., 2:287-304.
  • (1994), Dialectical thinking and therapeutic action in the psychoanalytic process. Psychoanal. Quart., 63:187-218.
  • (1996), The intimate and ironic authority of the psychoanalyst's presence. Psychoanal. Quart., 65.102-136.
  • Jacoby, R. (1987), The Last Intellectuals: American Culture in the Age of Academe. New York: Farrar, Straus & Giroux.
  • Kriegman, D. (1996), On the existential/subjectivism-scientific objectivism dialectic in self psychology: a view from evolutionary biology. In: Basic Ideas Reconsidered: Progress in Self Psychology, Volume 12, ed. A. Goldberg. Hillsdale, NJ: The Analytic Press.
  • & Slavin, M. (1989), The myth of the repetition compulsion and the negative therapeutic reaction: An evolutionary biological analysis. In: Dimensions of Self Experience: Progress in Self Psychology, Vol. 5, ed. A. Goldberg. I Hillsdale NJ: The Analytic Press.
  • Lasch, C. (1995), The Revolt of the Elites and the Betrayal of Democracy. New York: Norton.
  • Leary, K. (1994), Psychoanalytic "problems" and postmodern "solutions." Psychoanal. Quart., 63:433-46.5.
  • Lehman, D. (1991), Signs of the Times: Deconstruction and the Fall of Paul de Man. New York: Poseidon Press.
  • Lerner, M. (1996), "Ihe Politics of Meaning. New York: Addison-Wesley.
  • Levenson, E. A. (1983), The Ambiguity of Change: An Inquiry into the Nature of Psychoanalytic Reality. New York: Basic Books.
  • Lyotard, J. F. (1983), Answering the question. What is postmodemism? In: Postmodernism: A Reader, ed. T. Docherty. (1993), New York: Columbia University Press, pp. 38-46.
  • Mayer, E. (1996), Psychoanalytic stories about gender: Moving toward an integration of mind and body. Gender & Psychoanal., 1:239-247.
  • Mitchell, S. A. (1993), Hope and Dread in Psychoanalysis. New York: Basic Books. Ogden, T. H. (1986), The Primitive Edge of Experience. Northvale, NJ: Aronson Rabin, H. (1995), The liberating effect on the analyst of the paradigm shift in psychoanalysis. Psychoanal. Psycho'., 21:467-482.
  • Renik, O. (1993a), Analytic interaction: Conceptualizing technique in light of the analyst's irreducible subjectivity. Psychoanal. Quart., 62:553-571.
  • (1993b), Countertransference enactment and the psychoanalytic process. In: Psychic Structure and Psychic Change, ed. M. Horowitz, O. Kemberg, & E. Weinshel. Madison, CT International Universities Press.
  • (1995a), The analyst's expectations in classical technique: Reflections on the concept of resistance.], Amer. Psychoanal. Assn., 43:8.3-94.
  • (1995b). The ideal of the anonymous analyst. Psychoanal. Quart., 64:466-49.5. Schafer, R. (1992), Retelling a Life: Narration and Dialogue in Psychoanalysis. New York: Basic Books.
  • Slavin, M. O. & Kriegman, D. (1992), The Adaptive Design of the Human Psyche: Psychoanalysis, Evolutionary Biology, and the Therapeutic Process. New York: Guilford. Spezzano, C. (I 993a), Affect in Psychoanalysis. Hillsdale, NJ: The Analytic Press.
  • (1993b), Illusions of candor: Reply to Sass. Psychoanal. Dialog., 3:267-278. (1995), "Classical" versus "contemporary" theory: The differences that matter clinically. Contemp. Psychoanal. 31:20-46.
  • Stern, D. B. (1991), A philosophy for the embedded analyst. Contemp. Psychoanal., 27:51-80.
  • Wallis, J. (1994), The Soul of Politics. New York: New Press.
  • Weiss, J. (1993), How Psychotherapy Works: Process and Technique. New York: Guilford. & Sampson, H. (1986), The Psychoanalytic Process: New York: Guilford.

(1) One of the main problems in discussing postmodernism in psychoanalysis is that it does not exist as a monolithic entity. As Elliot and Spezzano (1996) rightly pointed out, various sensibilities can be called postmodern. No one theorist carries all of them. "There are often significant differences on important matters between theorists who typically get lumped together under the same postmodem banner. And yet, we are always grappling with the challenge of attempting to artificially define certain anchorpoints of a position to have something solid with which to argue or integrate. The task of defining a sensibility to assess its strengths and weaknesses while still doing justice to the range of opinions, variety of emphases,_ and idiosyncratic contributions of it's supporters is a difficult one. If collapsed into a unitary "it" or "they," the poly-vocal conversation that postmodernism brings to psychoanalysis is diminished.

On the other hand, if seen in too much complexity and with too much appreciation for its variety and particularity, the postmodern orientation can be like an infinitely moving target, saying everything in general but nothing in particular. It is said that the Eskimo has 15 words for snow but to the rest of us, it is still transparently meaningful to just say "it is snowing." In other words, when we speak of postmodern psychoanalysis, I believe that we all have the generalized sense that there is a multifaceted "something" there, a "something" different or new that we are all trying to understand and integrate, even if every time we try to define that something in the abstract ("this is the essence of postmodernism," "postmodernism says. . ." "so-and-so is the epitome of the postmodern theorist . . ."), there turns out to be four exceptions and a dozen nuanced variations that are important. With these problems in mind, I nevertheless argue that questions should be raised about biases in an approach called postmodern, subsuming under this rubric such traditions as social constructivisrn and hermeneutics.
return to article

(2) I am grateful and owe a great debt to Owen Renik, Hal Sampson, Kim Chernin, and Tom Rosbrow, valued colleagues with whom I have had many hours of discussion and debate about the issues I explore in this section.
return to article

(3) Renik, then, does have a theory of validation involving observable patient variables as key outcome criteria in his interpretive hypothesis-testing. He believes that his empiricism leads him to favor an ethic of self-disclosure in which the analyst attempts to demystify her authority and expertise with patients. Renik believes that this postmodern sensibility is ultimately justified by the clinical outcomes he generates. In city experience, however, such an ethic of self-disclosure is not patient-specific: enough to provide an adequate clinical methodology. I-or instance, in my experience, some patients seem to feel safer, lift repressions, and deepen their work in response to interpersonal influences based on idealizations whereas others seem to get worse or shut clown in response to even the kind of tactful self-disclosures that Renik recommends. 1 believe that although Renik's distrust of certain forms of analytic authority is part of an important contemporary critique, it also risks being too general and immune to the particularity of the clinical moment. Ultimately, because Renik views his ethical position as empirically warranted-whereas I do not-the issue can only be settled through further studies of clinical data and not by mere argument or assertion here.
return to article

(4) I have argued elsewhere (Bader, 1994) that the "tilt" away from therapeutic aims in mainstream psychoanalysis was due to a number of factors: 1) Freud's conflicting identifications as both scientist-researcher and physician-healer; 2) the need to establish territorial boundaries between psychoanalysis and other healing modalities by privileging insight over behavior change; 3) the ideological battles over what constitutes "true" analysis that raged within American psychoanalysis in the 1950s, particularly in the debates over Alexander's approach; and 4) the disappointment and even resignation among some analysts over the actual therapeutic results of analysis. These factors operate "behind our backs" and contribute to the bias, still commonly seen in our literature, conferences, and training institutes, against putting too much stock in the patient's symptomatic improvement.
return to article

(5) Russell Jacoby (1987) viewed this trend as part of a decline of the "public intellectual." By public intellectual, he is referring to independent cultural and political theorists (e.g., Edmund Wilson, laonel Trilling, C. Wright Mills, Irving Howe, Simone de Beauvoir, etc.) who aimed their writing at an educated lay audience and who saw themselves as contributing to public life, a life not hidden and constructed by the university but fully engaged with the events of the day.
return to article

return to the top of page