Excerpts from Must Read Books & Articles on Mental Health Topics
Books, Part XXXVIII

Relationality: From Attachment to Intersubjectivity
Stephen A. Mitchell (2000)
Chapter II- Drives & Objects

In a series of dense, tightly reasoned metapsychological essays over the course of three decades, from the 1950s through 1970s, Hans Loewald worked on and worked out certain central, difficult problems that were of great concern to him: the nature of mind, the relationship between actual events and psychical internality, and the revitalization and transformation of the past within current experience. It would not be too reductive to say that Loewald was working on problems of being and time, and putting it that way highlights the impact on Loewald's work of the three years he spent studying philosophy with Heidegger in Freiberg. But if the problems that gripped Loewald were Heideggerian, the conceptual world that he lived in and loved was Freudian.
     Loewald had a profound respect and deep passion for the psychological explorations Freud undertook and the conceptual tools Freud gave us for investigating, thinking about, and talking about unconscious processes. Yet Loewald clearly felt that Freud's theories were not tidy, final explanations, that he had opened up a largely uncharted realm that was left to us to explore. And there were fundamental problems in Freudian theory, problems of being and time, that Loewald was gripped by and felt required attention. Loewald believed that Freud's own grasp of these issues was lacking, inevitably so, because of Freud's place in time and the history of ideas. It was always important to Freud to reaffirm the scientific status of psychoanalysis and the objectivity of his discoveries, and to systematize whatever his flashes of intuition had illuminated by making his metapsychological systems consistent with the biology and physics of his time. So Freud's metapsychology lumbered along, like a wagon train behind a scout, colonizing, according to the fashions of the day, newly opened territory.
     But it was precisely in the assimilation of Freud's insights into the intellectual conventions of his day, and our day as well, that Loewald felt some of the most remarkable implications of Freud's discoveries remained undeveloped. Loewald reminds us repeatedly that in a genius as fecund as Freud's, there are always multiple meanings, contradictory positions, subtexts, unexplored paths. And the problems in which Loewald was interested often required a return to just those points in Freud's work where imaginative clinical leaps had been harnessed into familiar intellectual conventions. Loewald returns, again and again, to pick up the thread, to venture down the road not taken.
     Loewald felt that the dominant interpretations of Freud's work by the Freudian authorities of his (Loewald's) day were woefully inadequate. Loewald was working during the heyday of structural theory and ego psychology, and although it was not his nature to become involved in ideological disputes, it is clear that Loewald had serious problems with both of these approaches. In a review of Psychoanalytic Concepts and the Structural Theory by Arlow and Brenner, a book that was to become a pillar of the then contemporary Freudian structural theory, Loewald (1966) concluded: "Many unsolved and obscure questions of psychoanalytic theory look as though they were now taken care of and cleared up. This approach simplifies theory at too high a price, for the gain of a smooth, neat surface" (p. 58). And although Loewald sometimes identifies himself with "ego psychology" as the dominant ideology of his day, his ego psychology is clearly different from the mainstream ego psychology grounded in Hartmann's work. In fact, as we noted in the previous chapter, Hartmann often operates as the unspoken antagonist in many of Loewald's essays. And Loewald's final book, Sublimation, was in many respects an alternative approach to Hartmann's concept of "drive neutralization" as a bridge between primary drives and human intellectual and cultural achievements. Thus, Loewald's contributions are located within and saturated with Freud's thinking, but it is not the Freud of the Freudian authorities of Loewald's day. It is not even Freud as Freud understood himself.
     Reading Loewald is tricky. The language is Freud's, but the meanings have often been changed, slowly, from one paper to the next. And a consideration of Loewald's theory of object relations, the
subject of this chapter, requires us to plunge into a thicket of the most contested, rhetorically burdened terms of all: instincts, drives, objects, and object relations. Therefore, we need to stick close to Loewald's own language and careful arguments. I hope to show that in his struggle with basic problems of Freudian theory, Loewald presented us with a vision of mind and relationships with others that in many respects anticipated much of current psychoanalytic thinking and in some respects provides us with a visionary glimpse of human experience the implications of which we are still struggling to understand. Like Freud, Loewald can be read differently, from various angles; like Freud, he can be read in pedantic or imaginative ways. What I am most interested in, ultimately, is not the philosopher's Loewald but the clinician's Loewald, the illumination that Loewald's contributions provide contemporary analytic clinicians in our current struggles with the complexities of the analytic process.


There are two features of Loewald's commitment to Freud that are particularly striking. First, Loewald believed that the central feature of Freud's contribution was his theory of drive--his uncovering of the instinctual, primitive, "lower" sources of human motivation. Second, Loewald believed that there was something fundamentally wrong with the way drive was understood, both by Freud and by mainstream psychoanalysis.
     In a review of the Freud-Jung correspondence, Loewald (1974b) makes clear this dual concern in his appreciation of Jung's contributions on spirituality and self-transformation, which overlap, in many significant respects, Loewald's own interests, particularly his final work on sublimation. Yet, Jung's abandonment of Freud's unmasking of the seamier, primitive roots of human experience was a serious blunder.

Psychoanalysis, I believe, shares with modern existentialism the tenet that superpersonal and transcendental aspects of human existence and of unconscious and instinctual life [so much stressed by Jung] can be experienced and integrated convincingly--without escapist embellishments, otherworldly consolations and going off into the clouds--only in the concreteness of one's own personal life, including the ugliness, trivialities, and sham that go with it [p. 416].

Despite the interest he shared in "higher" transformations of the human spirit, it was essential to Loewald never to forget the "lower" sources of motivation that Freud's revolution had uncovered, his unmasking of our intricate hypocrisies, and his revelation of the body-based underbelly of all our activities, the body in its full corporeality, in its surfaces, its parts, its excretions. Loewald reads Jung as enjoining us, like the Bishop of Yeats's Crazy Jane poems, to "Love in a heavenly mansion, / Not in some foul sty." But Loewald responds to Jung like Crazy Jane responded to the Bishop:

'Fair and foul are near of kin,
And fair needs foul,' I cried.
... Love has pitched his mansion in
The place of excrement;
For nothing can be sole or whole
That has not been rent [p. 255].

Yet, despite the extraordinary power of Freud's vision, his blend of fair and foul, there was, for Loewald, something fundamentally off about the way Freud typically thought about drives. This had to do with beginnings, the locus and origination of experience.
     Freud said many, many different things about drives and instincts and early experience. But throughout, Freud felt that an appreciation of man's biological nature, our Darwinian bedrock, entailed granting the origination of experience to body-based instinctual impulses. The "source' of drives, as he put it, is in a body part. Drives emerge from the body and make a demand on the mind for work. The locus of activity begins within the individual and pushes outward toward the world. The id predates the ego, which grows like a membrane on its surface, to shelter the id from externality and to mediate its interface with the outside world. The center of the individual, despite our thorough socialization over the course of development, is in the id. "The core of our being, then, is formed by the obscure id, which has no direct communication with the external world and is accessible even to our own knowledge only through the medium of another agency" (Freud, 1940, p. 197). Life for Freud is generated through the clash between the id and the external world, and it is precisely because of that fundamental incompatibility that the id itself never directly meets the external, interpersonal world. As Jay Greenberg (1996) has put it, "Freud is expressing a vision in which human development and its psychopathology are rooted in the collision between organism (drive, thing) and environment (object, word). Personality is formed in this collision and is therefore always a tragic compromise" (p. 891). It is precisely the nature of the drives as presocial and asocial that links Freud to one line of his intellectual ancestors -- the romantics -- and operates, as Rapaport (1958, p. 727) pointed out, as a hedge against the pull toward adaptation and socialization.
     Despite its explanatory power and appeal, there was, for Loewald, something fundamentally wrong with this vision. Perhaps the central feature of Loewald's revisions of Freudian theory is his shifting the locus of experience, the point of origination, from the individual to the field within which the individual comes into consciousness, and this has been making its way into contemporary Freudian thought. In the beginning, Loewald says over and over, is not the impulse; in the beginning is the field in which all individuals are embedded. Experience does not proceed, as Freud believed, from inside outward, from the id's impulse, through the ego, into negotiation with the outside world. Experience initially moves from outside inward, from an increasingly differentiated unity of which the individual is a part to the development of the individual through an internalization of those external patterns.
     Loewald found the authority for this basic reconceptualization of the nature of drives in Freud's reference to the "oceanic feeling" in the opening of Civilization and Its Discontents (1930). Romain Rolland had challenged Freud's critique of religion as infantile by evoking the mystical experience of oneness. Freud responded by suggesting that life begins in an experience of boundarylessness: "Originally the ego includes everything," Freud states, "later it separates off an external world from itself" (p. 68). Perhaps this early condition is reexperienced in mystical states, Freud conceded, although he himself reported never having had such moments. Freud's depiction of the origins of the ego in this passage, so important to Loewald, is difficult to reconcile with what might be regarded as the major trend in Freud's portrayal of the ego in other works, in which the ego is an early developmental achievement. If the ego arises on the surface of the id to mediate between the id and the external world, it is hard to imagine how the ego in the beginning could also include the external world. This is one of those points at which the richness of Freud's imagination exceeded his efforts at theoretical integration.
     But the passage from Civilization and Its Discontents became the centerpiece of Loewald's revision of the concept of drive. If experience begins in a boundaryless unity, Loewald reasoned, mind, at its fundamental levels, cannot be composed of body-based impulses emerging from the individual and clashing with the external world. The very experience of being an individual mind and an individual body distinct from other minds and bodies-all this is a secondary development, a reorganization. And that changes the very meaning of the term "drive," leading Loewald (1972) down a road very different from those Freud traversed.

Instincts or instinctual drives ... arise within and develop from a psychic matrix or field constituted essentially by the motherchild unit ... not as biological forces, [but] as forces that ab initio manifest themselves within and between what gradually differentiates into individual and environment (or ego and objects, or self and object world ... ). Instincts remain relational phenomena, rather than being considered energies within a closed system, to be "discharged" somewhere [pp. 152-1531.

In Loewald's view, Freud began to shift from his earlier energicdischarge (Loewald sometimes refers to it facetiously as Freud's "fuel-injection") notion of drive to a relational notion of drive in 1920 with the introduction of the concept of Eros in "Beyond the Pleasure Principle." This shift, however, was never complete, and Freud's relational model of drives remained a secondary, largely undeveloped avenue. But it was central to Loewald's project and recently has been further developed in the contributions on Eros in the writings of Jonathan Lear (1990,1998).
     Consider some of the redefinitions of basic drive theory terms that derive from Loewald's reconceptualization of drives as relational rather than energy-discharge phenomena.
The terms "primary" and "secondary" process have "a deeper meaning," Loewald (1972) suggests, than in their original usage as modes of energic regulation.

Mental and memorial processes are primary if and insofar as they are unitary, single-minded ... undifferentiated and nondifferentiating, unhampered ... by laws of contradiction, causality, and by the differentiation of past, present, and future and of subject and object, i.e. by the differentiation of temporal and spatial relations [pp. 167-168].

Thus, the term "primary process," Loewald suggests, should be used in reference to the original state of the infant-mother field, in which there is no organization as such, in which all the usual distinctions that make possible our ordinary experience are missing. "The secondary process is secondary insofar as in it duality becomes established, insofar as it differentiates; among these differentiations is the distinction between the perceiver and the perceived" (p. 168).
     In secondary process, Loewald suggests, the relational field has become organized, and temporal and spatial distinctions have become established. Secondary process does not emerge spontaneously by itself; it is introduced to the child by caregivers. It should be noted that secondary process cannot ever exist independently of primary process; it presumes an underlying unity that it organizes and differentiates. "The secondary process," " Loewald (1977a) suggests, "consists not simply in splitting, dividing, discriminating,... but that in this same act the original wholeness is kept alive by an articulating integration that makes a textured totality out of a global one" (p. 196).
     In Freud's drive model, the term "cathexis" refers to an investment of energy. In Loewald's (1977a) relational drive model, cathexis refers to organizing activity. Different kinds of cathexes refer to different ways in which the relational field can be organized. Thus,

object-cathexis is not the investment of an object with some energy charge, but an organizing mental act (instinctual in
origin) that structures available material as an object, i.e. as an entity differentiated and relatively distant from the organizing agent. Such a cathexis creates -- and in subsequent, secondary cathecting activities re-creates and reorganizes -- the object qua object. It is objectifying cathexis [p.195].

     It is crucial to note here, and we soon take this up in greater depth, that Loewald is taking pains to suggest that objects, in a psychological sense, do not exist independently of the subject. Objects are created by being invested with significance through organizational (objectcathectic) activity out of the "primal density" or primary process.
Similarly, Loewald suggests,

narcissistic cathexis ... is not investment of a pre-existing ego or self with some energy charge, but a mental act (instinctual in origin) in which "available material" is not differentiated from the cathecting agent, not distanced in the cathecting act; the cathexis is identificatory, not objectifying [p.195].

In object cathexis, one is drawing a boundary around a piece of experience, differentiating something out, and saying, "This is you." In narcissistic or identificatory cathexis, one is drawing a boundary around a piece of experience, differentiating something out, and saying, "This is me."
     It should be apparent that through all these redefinition the very sense of "drives" has been radically transformed. The id, the repository of drives, is no longer understood as the initiator of motivation and, as Freud put it repeatedly, cut off from the external world; for Loewald, the id and the drives are relational patterns through which experience is organized. In one of my favorite passages from his classic paper on "On the Therapeutic Action of Psychoanalysis," Loewald (1960) suggests that far from being a hedge against social forces pulling for adaptation, the id is itself created through mutually adaptive interactive processes. Like the remains of a buried city, the fragments may be unrelated to current social reality. But those fragments were hardly pushed up from the center of the earth. Like primary process, they are residues of a once richly interactive social life. To use Freud's archeological simile, it is as though the functional relationship between the deeper strata of an excavation and their external environment were denied because these deeper strata are not in a functional relationship with the present day environment; as though it were maintained that the architectural structures of deeper, earlier strata are due to purely "internal" processes.... The id -- in the archeological analogy being comparable to a deeper, earlier stratum -- integrates with its correlative "early" external environment as much as the ego integrates with the ego's more "recent" external reality. The id deals with and is a creature of "adaptation" just as much as the ego -- but on a very different level of organization [p. 232].


What is an object? For Freud, objects are other persons, or body parts, or things that have been discovered to be useful in reducing the tension of drives. For Klein, objects are a kind of teleological image wired into drives themselves, like Jungian a priori archetypes toward which desire is inherently directed, which then become psychically intermingled with real others and parts of others in the external world. For Fairbairn, reversing Klein, objects begin as real others in the external world toward which "object-seeking" libido is directed, which may defensively and compensatorily become transformed into internal presences. In all these accounts, objects exist as either real external entities or prewired properties of experience.
      Loewald's understanding of "objects" is quite different from all these accounts. For Loewald, experience begins in an undifferentiated state; there are no objects, no drives, no self, no others, no now, no then, no external, no internal. Everything is experienced in terms of what Loewald calls a "primal density." All the distinctions and boundaries with which we are familiar are superimposed upon this primal density.
     One of the central psychoanalytic questions has always been "Why do we seek objects?" (see Greenberg, 1991). There have been many different answers: pleasure, safety, attachment, recognition, and so on. In Loewald's perspective, the question actually does not make sense. It presumes "we" and "objects" are separate phenomena. Yet, "In the formation of the ego," Loewald (1949) suggests, "the libido does not turn to objects that, so to speak, lie ready for it, waiting to be turned to. In the developmental process, reality, at first without boundaries against an ego, later in magical communication with it, becomes objective at last" (p. 19). For Loewald (1971a), we are our objects, and our objects are us. The distinction between drives and objects is a developmentally later, secondary process superimposition upon the primal density in which self and other are not yet sorted out.

Objects are not givens. On the contrary, a a highly complex course of psychic development is required for environmental and body-surface stimuli to become organized and experienced as external, in contrast to internal, and for such sources of stimulation, gratification, and frustration eventually to become objects [p. 127].

Loewald later concludes:

Objects are "originally connected" with instincts in such a way that the problem is not how they become connected in the course of time and development, or why they become con- . nected. Rather, seen from the standpoint of instinctual life, the problem is: how what later is distinguished as object from subject, becomes differentiated, in the course of mental development, from instincts [p.136].

This crucial point may seem abstract and difficult to grasp, but it is actually quite familiar to any interpersonally oriented psychoanalyst or couple or family therapist. Consider the way things work in couples or families. We are all fundamentally conflicted about all the major issues of life: seeking versus regulating pleasure; expressing versus restraining aggression; spending versus conserving money; rededicating versus reopening commitments, and so on. We all contain intense feelings on both sides of all these issues. But as couple and family relationships are often structured, one or the other side of these fundamental universal conflicts is assigned to different persons. One partner wants a commitment and the other evades it; one family member spends recklessly while another patrols expenditures; one child expresses the family outrage while the other is a model of virtue; one partner seems to want sex all the time while the other seems to avoid it. What goes on here? Each participant individually (and collaboratively with other participants) creates his own subjectivity and creates his objects by superimposing on the rich, affective, conflictual density of experience a simplifying scheme, through sorting out and assigning different qualities to different participants. It is just this sort of layering process, Loewald is suggesting, through which secondary process is generated out of primary process.
     Consider the implications of this novel theory of the origins of objects and object relations for Loewald's understanding of the crucial processes of internalization, projection, and identification. In virtually all other psychoanalytic theory, internalizations, projections, and identifications are discrete acts, generally understood as ego defenses. The paradigm for this understanding was Freud's (1917) formula that identification follows abandoned object cathexes -identification and internalization are defenses against the pain of loss.
     What is distinctive about Loewald's approach is that objects are not sought, or found, or taken inside, or expelled. In the beginning, there is no outside or inside. We are one with our objects; objects and self emerge out of the material of dense, affectively laden experience. Let us see if we can clarify some of the implications of this difference in theory through a thought experiment.
     Picture yourself waking up in a room adjacent to various other rooms with other people and things. You go into one of those and bring that person back into your room, thereby "internalizing" him, or bring some things that were in other rooms back into yours, thereby "internalizing" those properties that you now can claim as yours. This is the traditional model of objects and internalization.
     Now picture yourself in a large space with several other people and many things all jumbled together. No distinctions exist; everything is yours; there is only yours. Let's call this state of affairs "primary process." Later, you start creating boundaries and borders around some of the other people and things. Some you put in rooms of their own, separate from you. Others you keep in your space. Others are in in-between spaces with two-foot-high room dividers. Let's call this more complexly differentiated state "secondary process." Finally, let us assume that these room dividers are made of some sort of translucent material, so they are both there and not there; they can be felt in some sense modalities but also disappear in others. This is Loewald's model.
     For Loewald, internal objects and identifications begin as you; their sense of otherness is a product of secondary process differentiations and sorting. But primary process is operative not just in the earliest months of life; it is an ongoing organization of experience. At one point, surely evoking Heidegger, Loewald (1978a) suggests that the primary unity of experience might "best be called being" (p. 36). Primary process, in its undifferentiated unity, operates as a level of organization, simultaneous and parallel to the secondary process that dominates our consciousness with its differentiated objects. "It is this interplay between unconscious and consciousness," Loewald suggests, "between past and present, between the intense density of undifferentiated, inarticulate experience and the lucidity of conscious articulate experience, that gives meaning to our life" (pp. 49-50).
     In the traditional model, internal objects and identifications were created when you moved objects and things from other rooms into yours, and projections were created when you moved things from your room into others. In Loewald's model, there was no moving around, only different patterns of organization. Self and other are created by selectively drawing boundaries around some features of experience and excluding others. The experience of self is generated in the identificatory process of creating internality; the experience of otherness is generated in the projective process of creating externality. Of course, this analogy can take us only so far. The imaginary rooms are about the relationship between mine and not-mine. Psychoanalytic models of the psyche are about the relationship between me and not-me.
     Thus, for Loewald, the distinctions between self and other, internal and external, are psychological constructions. The interpersonal and the intersubjective are secondary constructions, developmental achievements, generated out of an underlying, undifferentiated psychic field, in which there are no persons and no subjects. And internal objects are also secondary constructions. The "other in oneself" as Loewald (1978a) puts it, is, "only the end product of a complex differentiating -- from another viewpoint, self-alienating -- process that takes its start in the primary unity of the infant-mother psychic matrix" (p. 14).
     Mind for Loewald (1965) is like a viscous psychic medium, within which relational configurations, interactions with others, are suspended and continually assimilated into the self and alienated from the self. "Internalization ... is conceived as the basic way of functioning of the psyche, not as one of its functions" (p. 71). Internalization (like sublimation in Loewald's later writings) is sharply contrasted with defensive processes. "In internalization, in contrast [to repression,] the ego opens itself up, loosens its current organization to allow for its own further growth" (p. 75).
     On higher levels of organization, objects have a vivid sense of otherness, of externality. On more primary levels of organization, the divisions are thin and permeable. Over time, externality may dissolve altogether as object libido becomes narcissistic libido and others become self. Significant interactions with early caretakers transcend the dividers between internal-external, now and then. Thus, for Loewald, the distinctions between internal and external, self and other are not objective features of the way the world is, but constructions on a gradient of possible constructions. Objects, or rather our interactive experiences with objects, take on varying degrees of internalization and externalization. Consider Loewald's (1972) description of early identifications, in which he is arguing against the traditional notion that internalizations are distinctive defensive responses to loss and deprivation.

Early identifications with the parents occur under circumstances that have nothing to do with deprivation or loss, but with a closeness amounting to lack of separateness, as though what is perceived or felt in this intimacy, by that very lack of distance, becomes an element in the child or helps to form his character -- as though the parent's trait is continued into the child, without his having to give up anything [pp. 162-1631.

Loewald envisions mind as a latticework of interactive identifications, simultaneously in different degrees of assimilation on different levels, with a sense of self on one side and a sense of externality on the other. This vision has close connections to contemporary relational notions of multiplicity of self-states and selforganizations (Mitchell, 1991; Davies, 1996, 1998b; Bromberg, 1998).
     Loewald's theory of object formation solves what for me was always one of the most interesting unsolved problems in psychoanalytic theorizing. Why are the residues of early object relations so persistent and resistant to change? It is just this feature of human psychology that makes our work so difficult, that necessitates such long stretches of time. Freud could describe it, but he couldn't really explain it. His metapsychological pleasure principle claims that we seek pleasure and avoid pain. Yet, the durability of early traumatic experiences and relationships is probably the most widespread psychological cause of human suffering. Polymorphously perverse libido, in all its plasticity, should be able to discard painful objects and find new ones. Yet the depth of our loyalty to painful early objects (which was the clinical basis for Fairbairn's redefinition of libido as not primarily pleasure-seeking but as object-seeking), which we encounter over and over in analysis and life in general, is staggering. Freud attributed this phenomenon to what he termed the "adhesiveness" of the libido, but threw up his hands at a compelling explanation by attributing it to a mysterious Death Instinct. And Meltzer (1975) used this same word, which I have always found so experientially vivid, in describing "adhesive identifications" in autistic states. What is so marvelous about Loewald's theory is that it dramatically reframes the whole problem. Primary identifications are so adhesive because there is a boundary between me and my objects only on a conscious, secondary process level of organization; on a primary process level, I am my objects, and my objects and I are always, necessarily, inseparable. They can never be expelled. This suggests that what can happen in psychoanalysis, what does happen, is not renunciation or exorcism of bad objects, but a transformation of them.


Self and objects are related to each other in Loewald's model of mind through interactions, and interactions are related to each other through time. To approach Loewald's understanding of time, which, like most other things, soon departs from conventional understandings, let us consider what he has to say about the complex, reciprocally generative relationship between perception and memory.
     What is the difference between perception and memory? Memories, Loewald (1978a) suggests, are experiences that have an "index of pastness" (p. 66), and perceptions are experiences that convey a quality of the present. But memories were once perceptions; they reproduce perceptions, which reverberate through them like echoes. And perceptions would not be possible without memories. Adult cataract patients whose vision is restored have no way to organize what they are looking at; the completely novel is indecipherable. We can only perceive something by re-cognizing it in terms of past perceptions, or memories.
     Thus, internal object relations, the internalized interactions with others that are the latticework of mind, are bound together in time. Time is the basic fabric of the psyche. And memory, Loewald (1972) suggests, is that psychic activity that traverses those temporal fibers, making links, continually creating channels through which "interactions with the world continue to reverberate" (p. 156) in a way that makes self-reflective, personal experience possible. Memory, Loewald suggests, is

the central, all-pervasive activity of the mind by which our world and our life gain breadth and depth and continuity in flux, and change in continuity, by which, in other words, our life and world acquire dimension and meaning, [making] memory virtually synonymous with mind itself [p. 149].

     Once again, these distinctions, which sound very abstract, . are extremely rich clinically. Consider the way in which we work with memories in the analytic situation. The patient may associate to an earlier time, his fifth birthday, let us say. That is clearly past, not
present, a memory, not a perception. But in what Loewald (1978a) calls "poignant remembering" (p. 66), with which we are most frequently involved in clinical psychoanalysis, we might ask the patient to see if he can get in touch with what that birthday was like, what it felt like. Our hope is that the earlier state of mind is not just being represented in an intellectual fashion, but to some extent being reexperienced.

As we become absorbed in such memories not only do we lose, as we say, the sense of time and space, but we tend to repeat, relive, internally and in our imagination, what we perhaps wanted only to recall as past events.... In all such experiences, while our rational processes may continue to operate and to articulate the material of experience, at the same time another level of our mind has been touched and activated [pp. 66-671.

      We have seen that Loewald believed that self-other and internalexternal are secondary constructions upon a parallel organization in which self-other, inside-outside are undifferentiated. Similarly, he also believed that our experience of time as duration -- past as distinct from present as distinct from future -- is a secondary construction upon a parallel organization in which these temporal categories do not exist. It is only as boundaries between self and other are constructed that past and present are also distinguished: "in a deeper sense, only by virtue of the differentiation of subject from object -- which is the primordial separation -- does memory arise" (Loewald, 1972, p. 160). Yet, just as with the boundary between self and other, the boundary between present and past exists only on a secondary, not a primary process level. Thus, the present and the past, perception and memory always retain their Siamese relationship (see also Grotstein, 1981) with each other, joined on one level and differentiated on another. Perception and memory, Loewald suggests, are bound together, necessarily, in a dialectic of reciprocal influence.
     In our customary way of thinking about these things, past, present, and future are discrete categories reflecting the objective passage of time as a succession of moments, one after the other. For Loewald, feelings of past, present and future are constructions that create a sense of before, now and after, and provide answers to the questions, What happened? What is happening? What will happen? Each of the concepts -- past, present, and future -- has no meaning in itself. Past in relation to what? Present in contrast to what? They imply each other, and create a subjective sense of connection, a narrative scaffolding for organizing experiences. According to Loewald (1971a), "We encounter time in psychic life primarily as a linking activity in which what we call past, present and future are woven into a nexus ... the nexus itself is not so much one of succession but of interaction" (p. 143). What we experience as our mind itself is the result of the continual regeneration of the links between past and present. Loewald (1972) states that "interactions with the world continue to reverberate, are reproduced, and thus lay the foundations for the development of an internal world, in the form of memorial processes" (p. 156). And, as a true constructivist, Loewald suggests that mind is continually, actively reconstructed through linking. "That this linking activity is automatic and unconscious in most of our daily life," Loewald (1971b) suggests, "obscures the fact that it is an activity" (p. 145).


Because Loewald writes about analytic process only in the most abstract terms, and his writing lacks virtually any clinical examples, it is impossible for analytic clinicians who have fallen under the spell of Loewald's extraordinarily rich vision to discern how Loewald himself actually worked. But I am not even sure this matters. As McLaughlin (1996) notes,

What has made Loewald's writings so beckoning to many of us is that he provided theoretical grounds to support a far broader range of technical activities than those sanctioned in the 1960s by the classical viewpoint in whose language he spoke. In his implicit clinical depiction of the working analyst, he shaped idealizing and appealing images about what a good analyst would wish to be. Yet these images are quite encompassing, making it easy for each of us to read into his depictions our own particularities, shaped according to our individual aspirations, and affirming of our preferences [pp. 901-902].

For our purposes here, I want to briefly address the implications of Loewald's revisions for the two most basic features of the analytic process: the analytic situation and the analytic relationship.
     Loewald views the human psyche in radically interactive terms. Our minds are open systems embedded in an interactive matrix with other minds, and our sense of self is a function of the internalization and continual reproduction and memorialization of those relationships. Loewald (1960) stresses repeatedly "the role that interaction with environment plays in the formation, development, and continued integrity of the psychic apparatus" (p. 221). In traditional terms, as Loewald describes them, the analytic situation is a medium within which the "closed system" of the patient's mind is revealed and interpreted by the analyst operating "as a reflecting mirror ... characterized by scrupulous neutrality" from a vantage point outside that system (p. 223). In Loewald's vision, the analytic situation is an open, interactive matrix, in which the analyst is "a co-actor on the analytic stage" (p. 223). And interaction is the key, previously unexplored feature that Loewald stresses over and over, with the implication that "a better understanding of the therapeutic action of psychoanalysis may lead to changes in technique" (p. 222).
     But how open is open when it comes to the analysand's engagement with the analyst as a real person in the analytic situation? And how open is the analyst as a psychic system encountering the patient? There are places where Loewald seems to suggest a radical, mutual openness and engagement indeed, suggestive of the kind of unique intersubjective mix Ogden writes about in connection with his term "the analytic third." " The analyst, Loewald (1977a) suggests,

is, on the now pertinent level of the patient's mental functioning, drawn into this undifferentiated force field ... he has to be in touch with that mental level in himself, a level on which for him too, the distance and separateness between himself and the patient are reduced or suspended. Ego boundaries, the whole complex individuating organization of self-object differentiation tend to dissolve [p. 379].

To return to our earlier analogy, we might say that the analyst joins the patient in the undivided room of primary process, or that the patient and analyst slowly come to find and differentiate themselves and each other collaboratively in an undivided room.
     Loewald (1974a) makes it clear that it is the lived reality of the transference-countertransference experience and its interpretive understanding for both participants that makes deep change possible. The analyst's role is not one of "detached spectatorship "; Loewald stresses the importance of "the analyst's capacity and skill of conveying to the patient how he, the analyst, uses his own emotional experience and resources for understanding the patient and for advancing the patient's access to his, the patient's inner resources" (p. 356). Thus memories become revived as perceptions, ghosts are raised, and new and different perceptions generate different, less dissociated memories, as ghosts are transformed into ancestors. "It is thus not only true that the present is influenced by the past, but also that the past -- as a living force within the patient -- is influenced by the present" (p. 360).
     Consistent with his understanding of the analytic process as interactive, Loewald recommends that the analyst engage the patient in lively terms, without the traditional restraints that come with aspiring to an impossible objectivity and neutrality. He frequently compares the analyst's role to that of the parent, not so much the mother of the infant as in the British school of object relations, but most often the parent of an adolescent. "So-called educational measures, and at times encouragement and reassurance, are used. If used judiciously they often make possible and enhance the more strictly psychoanalytic interventions in all phases of an analysis" (Loewald, 1977b, p. 375).
     Although Loewald warned against the dangers of infantilizing the patient through overuse of developmental metaphors, he has come under fire by contemporary relational critics for infantilizing the patient and also (to borrow a pun from Phillip Bromberg, 1998) "for adult-erating" the analyst (p. 144). In consistently portraying the analyst as, necessarily, a "mature" object, Loewald grants a privileged status to the analyst that rings false to many contemporary ears. The deconstruction, in recent decades, of the automatic assumption of the analyst's maturity, along with all other forms of authority, makes Loewald's attribution unpersuasive. Even a cursory glimpse inside the politics of analytic institutes and associations quickly reveals that psychoanalysts are not paragons of maturity.
     Loewald was writing at a time when analysts of all persuasions took for granted that the analyst's analysis made him or her more rational and mature, developmentally more advanced than patients. In this respect, I think that this criticism of Loewald and his contemporaries is well founded. But I think Loewald's critics are missing something and this concerns, as it usually does with Loewald, the particular meanings he attributes to words we tend to assume have commonly understood meanings.
     Yes, Loewald regards the analyst as more mature than the patient, but what does he mean, exactly, by "maturity"? Maturity for Loewald is not the customary advanced position along a linear developmental scale; for him, maturity is the capacity to navigate among and bridge different developmental and organizational levels. Consider this passage from Loewald's (1949) earliest psychoanalytic paper, which presages so much that was to follow.

It is not merely a question of survival of former stages of ego-reality integration, but that people shift considerably, from day to day, at different periods in their lives, in different moods and situations, from one such level to other levels. In fact, it would seem that the more alive people are (though not necessarily more stable), the broader their range of ego-reality levels is. Perhaps the so-called fully developed, mature ego is not one that has become fixated at the presumably highest or latest stage of development, having left the others behind it, but is an ego that integrates its reality in such a way that the earlier and deeper levels of ego-reality integration remain alive as dynamic sources of higher organization [p. 20].

Loewald (1977b) thus portrays the analyst not as solidly, consistently parental, but as straddling levels of organization .7 "The difference between the patient and the analyst is that the former is at the mercy of that primitive level (inundated by it or disavowing it), whereas the analyst is aware of but not given over to it" (p. 379). In the recent psychoanalytic literature, different authors strike different sorts of balances in depicting the contrast between the analyst's surrender to primary process and the analyst's maintenance of at least some foothold in secondary process. Loewald often seems to assume considerable control on the analyst's part over his own participation, although in his final work (1986) he seemed to be stressing increasingly the dangers of rationality as screening out unconscious communication and the sense that the analyst's control is, necessarily, episodically lost and regained.
     I find it compelling to make use of Loewald's contributions on the analytic situation by distinguishing analysand and analyst, not in terms of characterological levels of maturity, but in terms of the difference in their roles and the impact of those differences on the states of mind that become available to them. As Lear (1998) points out, for Loewald, "the psychoanalytic field is constituted by a differential in psychological organization: the analyst is more highly organized than the analysand" (p.133). But, consider Loewald's (1974a) very careful choice of words in the following passage; what he is conveying here is complex.

[The patient] knows that he has come to another adult for help, hoping or trusting that the analyst is more experienced, more knowledgeable, and more mature in regard to emotional life than he himself. As an adult the patient also knows that he is not altogether the child he makes himself out to be, or the child he lets take over in the regressive pull of the analytic situation.

The patient "hopes" the analyst is more experienced, knowledgeable, and mature in regard to emotional life. In contrast, the patient "makes himself out to be" a child in the "regressive pull" of the analytic situation. Loewald is tracing the ways in which the different roles shape the greatly overlapping experiences of patient and analyst in different ways. The analyst, in his assumption of professional responsibility, also hopes he is more mature and knowledgeable and tries to act that way, much as a parent might with a child with whom he is emotionally enmeshed. The patient does not have to be responsible and organized in this way; in fact, we consider the patient who is trying to be responsible in this way to be resisting the precious opportunity the analytic situation provides for a freedom from conventional accountability, a surrender (Ghent, 1992) to unintegration. Loewald regards as essential for the analytic process the analyst's capacity to bridge primary and secondary process, self and other, past and present, reality and fantasy.
     Thus, the relative maturity that Loewald assigns to the analyst vis-a-vis the patient is not so much concerned with the content of experience -- not primarily a difference in the organizational level on which the two operate -- but rather with a relatively greater facility for navigating shifts and convergences between different levels of experience. The patient, for extended stretches between interpretations, is encouraged to suspend that facility; it is a central feature of the analyst's professional responsibility to keep bridging concerns always in mind. These differences, and the dialectic between expressiveness and restraint they generate in the analytic process, are taken up in the final chapter.
     For Loewald, the analyst's own internal bridges among his own differentiated but linked levels of organization make possible the creation of an interpersonal analytic situation that becomes patterned, within the transference-countertransference, into revitalized islands of disparate dissociated experiences of the patient's. And these interpersonal patterns in turn become internalized for the patient (to some extent for the analyst as well) into a newly enriched, internal world. Consider one of my favorite of Loewald's (1977b) depictions of the analytic process. My hope is that this chapter may have contributed to making what might have seemed quite abstract, now richer in its primordial referents.

Psychoanalytic interpretations establish or make explicit bridges between two minds, and within the patient bridges between different areas and layers of the mind that lack or have lost connections with each other, that are not encompassed within an overall contextual organization of the personality. Interpretations establish or re-establish links between islands of unconscious mentation and between the unconscious and consciousness. They are translations that do not simply make the unconscious conscious or cause ego to be where id was; they link these different forms and contents of mental life, going back and forth between them.... What is therapeutic, I believe, is the mutual linking itself by which each of the linked elements gains or regains meaning or becomes richer in meaning--meaning being our word for the resultant of that reciprocal activity.

Loewald (1978a) revealed himself somewhat more in his Yale lectures for the general public than in his professional writings. There he said, and I would like to end this chapter on this note, that "one does not have to be a mystic to remain open to the mysteries of human life and human individuality".

A Meeting of Minds: Mutuality in Psychoanalysis
Lewis Aron (1996)

Heinrich Racker (1968) wrote what may be described as the "anthem" of contemporary relational psychoanalysis:

The first distortion of truth in "the myth of the analytic situation" is that analysis is an interaction between a sick person and a healthy one. The truth is that it is an interaction between two personalities.... each personality has its internal and external dependencies, anxieties, and pathological defenses; each is also a child with his internal parents; and each of these whole personalities--that of the analysand and that of the analyst--responds to every event of the analytic situation.
This is, for very good reason, an often cited quotation, for it sums up much of the essence of the principle of mutuality that is fundamental to the relational model in psychoanalysis. Racker drew our attention to a number of principles here, including that psychoanalysis is an interaction between two people, who each respond to every event in the analytic situation. That is to say, Racker viewed both members of the analytic couple as being participants in the interaction. Furthermore, Racker refers to the mutuality of neurotic interaction between patient and analyst in his insistence that neither is fully healthy, rational, or adult in his or her functioning.
     In a less well-known footnote to this famous passage Racker noted:
It is important to be aware of this "equality" because there is otherwise great danger that certain remnants of the "patriarchal order" will contaminate the analytic situation. The dearth of scientific study of countertransference is an expression of a "social inequality" in the analyst-analysand society and points to the need for "social reform"; ....
In Racker's prose we can hear the reverberations of the democratization of society and of the emergent feminist critique that has come to play an important role in the development of relational theories. Racker recognized and warned us of the dangers of polarizing the analyst and analysand with respect to their health and their capacities to judge reality. He recognized the issues of power, dominance, and authority that are at play between them. When Racker wrote of "equality" between patient and analyst, I believe that he used quotation marks because he did not really mean equality but, rather, the mutual containment of health and neurosis by both the analyst and the patient. Edgar Levenson, the best known contemporary interpersonal analyst, has stated that Racker's object-relational position and Sullivan's (1953) interpersonal dictum that "we are all more simply human than otherwise" (p. 32) "converge in a concept of psychoanalysis as a mutual, respectful exploration of a joint reality" (Levenson, 1987a, p. 214).
     Similarly, McLaughlin (1981), a well-respected and senior Freudian analyst, wrote that the traditional view of transference reinforces a model of psychoanalysis that differentiates sharply between a "patient with psychopathology and a physician with a cure" (p. 642). Psychic reality and objective reality become split; transference, psychic reality, and the infantile are attributed to the patient, and objectivity, external reality, maturity, and health are assigned to the analyst. One can see in this brief sampling of Freudian, interpersonal, and Kleinian thinking how a relational position that emphasizes mutuality can transcend narrow political and theoretical affiliations.
     Both the shift from drive theory to relational theory (Greenberg and Mitchell, 1983) and the movement from an epistemology of objectivism to constructivism (Hoffman, 1991) have taken place in the wider context of social, political, and economic changes in our larger society. PreWorld War I, Europe, where psychoanalysis was born, was a far more authoritarian time and place than are the societies in which psychoanalysis is now practiced. Social changes in the past century have increased the likelihood that such authority will be challenged and scrutinized. The notion of an objective, "neutral" observer has been called into question on epistemological grounds. Even in such a hard and exact science as physics, Heisenberg and Einstein demonstrated that the position of the observer and the acts of observation influence the nature of the data gathered.
     In the postmodern world, conceptions of truth are under attack everywhere. Poststructuralists, such as Derrida (1978, 1981), see meaning as multiple, unstable, and open to interpretation, relative to particular social, political, and historical contexts, and they thus move away from Grand Theory, which purports to assert universal Truth. Philosophers allied with postmodernism, such as Rorty (1979), criticize their predecessors as "essentialists" who assumed that there were innate essential meanings rather than historically contingent or local meanings. Postmodern psychologists, or social-constructionists, have begun to deconstruct such concepts as personal autonomy and the self
(Cushman, 1991; Gergen, 1991, 1994). Deconstructionists decenter the biases and assumptions of texts and move peripheral and marginalized perspectives to the center. Foucault (1984) reveals that multiple perspectives are erased from texts and suppressed and that those voices which are heard are connected to power and with strategies to maintain power. Feminists, particularly postmodern feminists, have alerted us to how fundamental conceptions such as gender, sexuality, and race are cultural and linguistic conventions that, when deconstructed, reveal hidden relations of authority and power. The presumed fixity of the existing social order is destabilized, and prominence is given to suppressed voices; in the mental health field, for instance, to the cries of women, children, minorities, gays, and patients.
     Where does all of this leave analysts in their attempts to be the relatively neutral interpreters of patients' psychic reality? The classical conception of the analytic process is based first and foremost on the historical understanding of transference as a distortion of reality. The analyst is in the position of having to interpret to the patient where, when, and how the patient has distorted or misperceived the analyst, who is, after all, a relatively blank screen. Analysts, protected from emotional overinvolvement by their relative silence, distance, anonymity, and containment, as well as by a presumed thorough training analysis and a capacity for ongoing meticulous self-analysis, are thought to be in a good position to sort out what in patients' perceptions is real and what is error or distortion. The analyst must be in a position to judge what is real and what is not; hence, the distinction between transference and reality, transference and alliance, transference and relationship. But when authority is questioned in all areas of life, when interpretation itself is seen as an operation of power, when truth is seen as relative and context bound, when the dominant perspective is decentered and marginalized voices are moved to the foreground, when meanings are seen as socially constructed, and universal laws are devalued, when the very notions of objectivity, reality, and truth are challenged, when every Grand Theory is attacked--where does that leave the psychoanalyst? How can analysts interpret the transference, perform that fundamental psychoanalytic act, when that act, as it has been traditionally conceptualized, implies that analysts have the authority to distinguish truth from distortion and to assign meaning as if they were not involved in a relationship of power over analysands in and through that very interpretive act?
     Our most cherished and fundamental beliefs about psychoanalytic technique are all being questioned. With the abandonment of drive theory, the dimension of frustration and gratification that for so long was the single best guide to the role of the object has been decentered.
Consequently, the rules of abstinence, anonymity, and neutrality have had to be forsaken. We do not take for granted the absolute truths of our metapsychologies; how could we when we have so many of them to choose from? So our interpretations must be based on something other than these foundations. It was inevitable that the postmodern condition would affect psychoanalysis, and it has under a variety of rubrics and schools including social-constructivism, hermeneuticism, narrative approaches, and what I have called relational-perspectivism.
     Relational-perspectivism eschews the role of the authoritative analyst, who knows truth and represents reality and therefore health, in favor of a view of the analyst as a coparticipant involved in a mutual if asymmetrical endeavor. The tendency to polarize into dichotomous categories such concepts as analyst-analysand, therapist-patient, observer-observed, health-neurosis, scientist data, experimentersubject, follows from an acceptance of the traditional dichotomy of subject-object that, in turn, has been intricately connected to the dichotomy male-female. Ideas of male-female opposition have been present in both Eastern and Western cultures throughout history. In the West women have been regarded as the repository of all that is not male; women are assigned the role of "other" to men. The objectivist and positivist experimental model of scientific research that we are so painstakingly taught in graduate school is based on these distinctions. In Francis Bacon's scientific model, nature was depicted as female, to be subdued and conquered by the penetrating male gaze (Keller, 1985). Throughout Western history and culture women have been associated with passion and emotion, and men to reason, technology, and civilization.
     In Freud's effort to constitute psychoanalysis as a positive science, he portrayed the analyst's functioning in the image of the phallus. The analyst was thought of as the fearless and adventurous male who seeks to uncover, expose, and penetrate the feminine "unconscious." The analyst needs to be sharp and insightful, brave and intrepid, fearless in "his pursuit" of the feminine unconscious. The analyst is to be "objective," and therefore "the subjective factor"--emotion, passion, subjectivity--has to be eliminated. With the shift in emphasis from a one-person to a two-person psychology, with the rise of a relational and intersubjective psychoanalytic theory, with the acknowledgment of the subjectivity of the analyst comes a shift in our view of psychoanalysis from a relatively detached to a personally and emotionally engaged activity. The subjectivity of the analyst, countertransference in its broadest sense, is not to be eliminated, but used, and the patient's experience of the analyst's subjectivity is to be articulated. The distinctions heretofore accepted between observer and observed, analyst and analysand, rational and irrational, male objectivity and female subjectivity, all collapse, and ambiguity, multiplicity, and paradox take over the center where clarity and identity have prevailed.
     Does contemporary psychoanalysis advocate that we abandon objectivity? Postmodern theories push in the direction of relativism, and for this and for other reasons postmodern thinking generates intense controversy. At stake are fundamental questions about how we know what we know and whether meaningful progress in knowledge can ever be made. The older, classical, positivist or objectivist model may have had the disadvantages of being somewhat authoritarian and even patriarchical. But at least in that model the analyst had to accept some responsibility! The analyst may have maintained the myth that "he" was healthier, wiser, more rational, and more mature than the patient, at least while he was being the analyst. But with all of this, "he" had to assume the ethical and moral responsibility to conduct the analysis appropriately, maintain appropriate boundaries, secure the analytic frame, interpret correctly and in a timely fashion, distinguish between what was real and what was distortion and what was progressive from what was regressive, and differentiate meaningful and authentic emotional expression from defense and resistance. Once we remove this striking asymmetry from the analysis, what is left to authorize the analyst to make these distinctions and thus guide the process as we would expect any expert to guide a professional undertaking? In short, with the postmodern collapse of the analyst's authority and power, what happens to professional expertise, professional responsibility, and professional ethics?
     I want to assert in the strongest possible terms that the abandonment of metapsychological truths and theoretical foundations does not necessitate the surrender of ethical standards, professional responsibility, or clinical judgment. Quite the contrary, in line with what I described in Chapter 1 as an affirmative postmodern sensibility, I believe that an acceptance of the relational-perspectivist approach that has guided my thinking throughout this book leads to the recognition that analysts must accept responsibility for the fact that it is their own personality, their own subjectivity, that underlies their values and beliefs, that infuses their theoretical convictions, and that forms the basis for their technical interventions and clinical judgements. There can be no technical choice or clinical decision that is not imbued with the analyst's subjectivity. I agree with and wish to emphasize Hoffman's (1995) forceful statement that, instead of regarding the countertransference as one factor among many that is to be considered by the analyst in making any intervention, we must recognize "the analyst's subjective, personal, countertransferential experience as the superordi-
nate context in which everything else, including theory, is embedded" (p. 108). Therefore, instead of disclaiming personal responsibility and attributing their understandings to an abstract metapsychology or universal theory, analysts must accept personal responsibility for their interpretive understandings and clinical interventions. Our understanding is always value laden, and, our values are always personal. There can be no neutral understanding or interpretation.
     Bernstein (1983) has described how, having been stripped of the possibility for secure knowledge, we have come to suffer from "Cartesian anxiety," the fear that we will be left with nothing but radical relativism. It is because of our Cartesian anxiety that we long for foundational knowledge. Bion (1990) wrote:
When approaching the unconscious--that is, what we do not know, not what we do know--we, patient and analyst alike, are certain to be disturbed. Anyone who is going to see a patient tomorrow should, at some point, experience fear. In every consulting room there ought to be two rather frightened people: the patient and the psychoanalyst. If they are not, one wonders why they are bothering to find out what everyone knows.
Being a psychoanalyst, like being an analysand, is no easy matter. It is, and as Bion says, it ought to be, frightening. As analysts, we long for indubitable, foundational knowledge. We want a solid and reliable theory to guide us and relieve our anxieties. But we, like our patients, must struggle without easy solutions. We must continue to make technical choices, to practice in one way rather than in another, to create certain ground rules for ourselves and our patients, to believe in some things and not in others. But we must accept that these choices reflect our own subjectivities; they are personal, and not only technical or theoretical choices. We must choose, but we cannot disclaim our choices as the inevitable outcomes of abstract and universal principles, Rather, we must accept our choices as based on our values, which, in turn, are reflections of who we are.
     We might question how we can speak to patients with any sense of conviction if we abandon our beliefs in fundamental truths. Hoffman (1992a, b), Mitchell (1993a), and Stem (1992) have suggested that, paradoxically, within a constructivist approach analysts are freer to speak their minds, because with the elimination of external standards of truth there is more room for spontaneous self-expression and personal conviction. Emphasizing how this approach leads to dialogue rather than to dogmatism, Spezzanno (1993) writes, "To say this is not to give up truth. It is to give up certainty about truth. By giving up certainty we accept endlessness as the most certain thing about our discussions" (p. 29). Since there are no foundations on which to rest, conver-
sation is endless. The analyst cannot end discussion simply by asserting authority; rather, all truths are partial, perspectival, and momentary and need to be questioned and further analyzed.
     Advocating that we acknowledge the profound mutuality in clinical psychoanalysis and a greater egalitarianism in our practice as well as in our epistemology does not entail an abandonment of professional expertise or discipline, nor does it require an abandonment of the essential asymmetry between patient and analyst. There is a critical difference between authority and authoritarianism, between conviction and expertise based on experience and training, on one hand, and inflexible certainty on the other. Those psychoanalytic authors (Hoffman, 1991, 1992a, b, c, 1993; Stem, 1992, Mitchell, 1993a) who have most persuasively advocated constructivist approaches have repeatedly pointed out that those approaches do not reduce the need for intellectual and professional discipline and a certain kind of analytic objectivity. But, here again, how can I speak of objectivity? From a perspectivist position, what is meant by analytic objectivity?
     In many respects, I have argued in this book that relational psychoanalysis adopts a dialectical approach, attempting to maintain tension between seemingly opposed principles, balancing the intrapsychic and the interpersonal, the intrasubjective and the intersubjective, the individual and the social, autonomy and mutuality. Rather than maintain the polarization of objectivism versus relativism, the philosophical view that has guided my thinking in this book may be referred to as "the dialectical sense of objectivity" (Megill, 1994, p. 7). Positivist principles of absolute objectivity exclude subjectivity, as Freud did, leading to an aperspectival objectivity or a "view from nowhere," a "God's eye view." Objectivity, in the modernist age, was viewed as an unproblematic quality of knowledge. Megill, surveying the current interdisciplinary debate about objectivity, views it as a contingent, varying, and deeply problematic product of cultural practice. He develops the notion of "dialectical objectivity," which "involves a positive attitude toward subjectivity. The defining feature of dialectical objectivity is the claim that subjectivity is indispensable to the constituting of objects" (p. 8). We do not gain anything if our critique of traditional thinking leads us to engage in a simple reversal of classical values. Traditionally, objectivity was prized and subjectivity was disdained. A simple reversal would disparage any attempt at objectivity on the grounds that it was impossible to achieve and would celebrate radical relativity and undisciplined subjectivity. The dialectical sense of objectivity, in contrast, recognizes that the extreme polarization of these concepts is itself an aspect of the problem. Dialectical objectivity is informed by subjectivity and includes within itself reflection on the subjective.
     To return to the famous statement with which I began this Coda, Racker (1968) went on to say:
The analyst's objectivity consists mainly in a certain attitude towards his own subjectivity and countertransference.... True objectivity is based upon a form of internal division that enables the analyst to make himself (his own countertransference and subjectivity) the object of his continuous observation and analysis. This position also enables him to be relatively "objective" towards the analysand.
Racker's description of relative objectivity, like the rest of his contributions, was well ahead of its time and anticipated the notion of a dialectical sense of objectivity as I have just described it, an objectivity dialectically achieved only by reflexive inclusion, rather than elimination, of one's own subjectivity.
     Traditionally, the rationale for analysts to have their own training analyses was so that they would gain "control" of their countertransferences, to know their own "personal equation" well enough so that it could be eliminated from consideration, to be analyzed well enough so as not to act out or participate interactionally with the patient. From the relational point of view being developed here, there is nothing in our training more important than the depth of our own analyses. Since, from the perspective being elaborated here, our subjectivity underlies everything that we do as analysts, and since, within the relational framework that I have described, access to the analyst's affective reactions is critical for his or her effective functioning (whether the analyst chooses to sustain and contain these affects or to express or act on them), the analyst's own analysis remains the most important element of our professional training.
     Our analytic expertise does not reside in our being "thoroughly analyzed" and therefore immune to neurosis, healthier than our patients and therefore better able to judge reality. Rather, our expertise resides in our acquiring and honing certain personal, interpersonal, and professional skills. These include the ability to reflect on our own participation in interpersonal relationships while recognizing just how limited that reflectiveness is at any moment; the ability to attend to our own affective experience and to reflect on this experience and symbolize it, along with the recognition that at any moment there is always a great deal that we are not attending to in our own experience; a facility in using an analytic theory or model with enough dexterity, or the capacity to move back and forth between a wide variety of analytic models (Mitchell, 1993a), so as to help patients construct pragmatically useful narratives of their own lives; a proficiency in deconstructing whatever storylines our patients present to us or that we have constructed with them so that we and they do not become rigidly fixated to any one narrative construction; the capacity to tolerate a certain level of anxiety and depression as our patients observe and scrutinize our participation with them and explicitly or implicitly comment to us on the nature of our participation with them, and the capacity to enjoy, celebrate, and take pride in the shared growth, intimacy, and mutual satisfactions inherent in the psychoanalytic process. This is a very short list of some of the numerous skills that are necessary for the clinical practice of psychoanalysis; none of them requires the adoption of fundamentalist notions of truth and reality, and all of them testify to the enormous degree of discipline required for the competent practice of psychoanalysis.
     Relational psychoanalysis suggests that it is not truth that patient and analyst pursue, so much as it is meaning that they attempt to construct (Mitchell, 1993a). Meaning is generated relationally and interpersonally, which is to say that meaning is negotiated and coconstructed. Analysts cannot simply construct psychoanalytic narratives or interpret patients' associations any way they please; constraints imposed by the structure of reality limit the possibilities of our plausible constructions. Analytic objectivity is dialectical and dialogical; we rely on our subjectivity, but our subjectivity is shaped and constrained by input from the object of our investigation, who is, after all, a separate subject. A dialectical sense of objectivity includes our subjectivity but does not ignore these constraints. Analytic objectivity is negotiated, relational, and intersubjective. Meaning is arrived at through a meeting of minds.
     I want to bring this book to a close by repeating Ferenczi's (1931) statement, for it is as true now as it was then: "Analytical technique has never been, nor is it now, something finally settled" (p. 235). There is no single correct way to do psychoanalysis, although there are many ways that do not work very well for far too many patients. Analysts should not be taught technical prescriptions, although they can be taught ways to think about the process of making informed clinical decisions; and they can be instructed in clinical theory, which will serve them well as a compass with which to navigate the deep and troubled waters of clinical psychoanalysis. Rather than viewing the psychoanalytic theory of technique as a manual dictating certain types of behavior, I prefer to view our theory of technique as a system of signposts that encourage reflection-in-action. Analysts have no choice but to experiment; to innovate; through trial and error to learn what seems to deepen the analytic work for themselves and for their patients; and to interpret and intervene, guided by their own moral and professional convictions and by the ethical standards of their professional community. This view of clinical practice leads to intervening not on the basis of a Grand Theory, a prescriptive theory of technique or a "technical rationality" (Schon, 1983), but on the basis of the analyst's own subjectivity (which includes his or her personal history and professional experience, a knowledge of theory, and clinical wisdom) as it has been shaped in dialogue with the subjectivity of the other, the individual patient. Since communal dialogue and endless conversation is the very basis of this meaning system, one caveat for analysts in their explorations is that they should do nothing with patients that they would not in principle be willing to discuss with colleagues, with the professional community, and with the public.
     Bion (1990) taught:
There is a thing known as "classical psycho-analysis": the analyst has an analytic situation in which he practices analysis; he has patients who are suitable patients and gives them suitable, certified-correct interpretations. I have never known that state. The analytic situation is the situation which the particular practitioner finds is adequate for himself

     Far from eliminating "the subjective factor," relational psychoanalysis asserts that analytic objectivity is constituted dialectically, rooted in the intersubjective relationship. Psychoanalytic findings may be said to be objective to the extent that they emerge intersubjectively, mutually constructed between analyst and analysand. Psychoanalytic theory, as well as our codes of ethics, may similarly be termed objective in the sense that psychoanalytic knowledge is communal or relational knowledge, socially derived, and based on the negotiated consensus of the psychoanalytic community of practitioners and theorists. That is, it is based on a meeting of minds.