Self Psychology Page 

What Does Heinz Kohut
Mean by the „Self"?
Associations on a Theme

Rudolf Süsske

German Version

[ Self Psychology Bulletin Board ]

This paper may also be downloaded.

The present version of this paper is based on a lecture presented under the same title at a meeting of the "Society for the Philosophy and Sciences of the Psyche" (Berlin, May 8, 1997) and incorporates audience questions and suggestions, particularly in the footnotes. A first version was presented in June 1993 at a continuing education seminar of the Department of Psychotherapy and Psychosomatic Medicine (Chair: E. Schiffer) of the Christian Hospital e.V. in Quakenbrück.

The following remarks, with all their digressions, address a single question: what does Heinz Kohut mean by the "self"? We are not primarily concerned here with Kohut’s contribution to research on narcissism (Kohut 1971), nor with his specific therapeutic concepts (selfobject transference and the like), but rather with his psychoanalytic and anthropological theory of the self, which contains the conceptual foundations of both his therapeutic practice and his intellectual excursions into sociology, art, and history.

"Conceptual foundations" suggests Kohut’s adversary: the metapsychology of Sigmund Freud. Freud’s own relationship to his metapsychology was at once passionate and detached. With respect to the problem of how the ego can control the drives, he wrote in 1937: "’We must call the Witch to our aid after all.'1 That is to say, the witch of metapsychology. Without metapsychological speculation and theorizing—I might almost have said fantasizing -, one gets no further" (Freud, Suppl. 365f).

The keystone here, and the stumbling block, was drive theory. The ego—consciousness—is not master in its own house. Its reactions are in perpetual conflict between instinctual wishes and the natural or cultural adversities of external reality. Both are subject to an epistemological qualification which Freud addresses in The Interpretation of Dreams:

"The unconscious is the true psychical reality: in its innermost nature it is as much unknown to us as the reality of the external world, and it is as incompletely presented by the data of consciousness as is the external world by the communications of our sense organs" (1900, 580).

This is also the source of the concept of the object and the correlated (cognitive) subject, containing structures which first constitute an object as an object. But what is important now is no longer the structures of reason or the intellect, but rather those of the instinctual drives. The source of the drives (which is inaccessible to psychological knowledge) lies in the biological bedrock; its goal is the release of tension. Objects exist chiefly for the purpose of gratification; that is, they serve to remove tension. Their obstinacy and recalcitrance, as well as the inner logic of drive development, not only banish people from the oceanic feeling of their self-sufficient primary narcissism, thus leading to the formation of ego structures responsible for cognition and action; they also challenge its moral standards. The destiny of the drives is to evolve the psychic structure of id, ego, and superego on the rocky path leading from autoeroticism to the Oedipus conflict. But life in the reality of nature and culture (the principle of reality) remains exposed to the depredations of a force striving anarchically toward pleasure. Psychoanalysis, as is well known, seeks to constrain this force: "Where id was, there ego shall be."

"Guilty" and "Tragic": Man’s Double Nature

Kohut does not reject this model of the psyche’s structure. "Man lives inside the pleasure principle; he attempts to satisfy his pleasure searching drives for, to relieve tensions which arise in his erogenic zones. The fact that man is often unable to reach his goals in this domain—not because of pressure from his environment, but primarily due to inner conflicts—leads me," Kohut writes, "to term him Guilty Man when regarded from this perspective" (Kohut 1977, 120).

This is the sphere of classical psychoanalysis, the structural neuroses, and their therapy. But man does not live only in a world of neurotic conflicts, or—in the best case—of drives sublimated into creativity. A primitive core of spontaneity exists in him, a "life beyond the pleasure principle," which strives for self-expression (cf. Kohut 1975, 269ff; 1977, 120f).

Kohut speaks of "the realization through action of the (life) plan laid down in [man’s] nuclear self." Kohut writes, "Here the undeniable fact that man fails more often than he succeeds leads me to give this aspect of man the negative designation Tragic Man, instead of ‘expressive’ or ‘creative man’" (Kohut 1977, 120f).

The individual whose self-expression is blocked experiences himself as prevented from building a tension arc from basic strivings (ambitions) to basic ideals (values, goals) using inborn talents and acquired abilities.

"This tension arc is the dynamic essence of the complete, non-defective self; it is the conceptualization of the structure whose establishment makes possible a creative-productive, fulfilling life" (Kohut 1984, 21).

"Narcissism" for the most part still stands for "self" in Kohut’s earlier writings (Kohut 1966 ff.), but two main tendencies inherent in man are already beginning to become distinct—his striving toward pleasure and his striving to give expression to the structure of his self, tendencies "which can either work together harmoniously or conflict with one another" (Kohut 1975, 273). There is no quantitative equilibrium between them, but the possibility exists that "the weaker ... sector may be able to play a supplementary role" (ibid.), just as there may also be lifelong "trouble-free cooperation" between them. There is no path leading from narcissism to object-love—selfobjects are not replaced by love objects— the two interact.

Kohut writes, "I have repeatedly stressed ... that object-love ..., like any other intense experience, strengthens the self. Furthermore, it is well known that a strong self enables us to experience love and desire more intensely" (Kohut 1984, 86).

So far we have only hinted at an answer to our question, "What does Kohut mean by the ‘self.’" Let us take a closer look.

 The Structure of the "Self"

Kohut is indefatigable in stressing the role of empathy, that is, sympathetic understanding of the introspection of the Other. Though he never says exactly what it is. Kohut intends empathy not only as a therapeutic agent, but also as an instrument of theoretical knowledge2; in many places in Kohut’s writings it even becomes a model of social conduct.

If we ask empathically what the self is, we get the following answers:

- "our sense of being an independent center of initiative and perception,"

- of being "integrated with our most central ambitions and ideals,"

- "and with our experience that our body and mind form a unit in space and a continuum in time" (Kohut 1977, 155).

Here we have, in embryo, the essential features of the self—more precisely of the bipolar self—to which we shall now turn.

If we leave the origins of the self unelucidated for the moment and, in the manner of most psychoanalytic theory, start from a hypothetical symbiotic phase, the course of events is as follows:

"The equilibrium of the complete security of the child is disturbed by the inevitable limits of maternal care, but the child replaces the previous perfection (a) by constructing a grandiose and exhibitionistic image of the self, the grandiose self, and (b) by assigning the previous perfection to a venerated, omnipotent (transitional) selfobject, the idealized parent imago" (Kohut 1971, 43).

Kohut uses the linguistic monstrosity self-selfobject relationship to designate this state of affairs. Selfobjects are thus objects3 or functions of objects which arouse the self-feeling, maintain it, or influence it positively.

Here only two clues—offered by Kohut—to the origin of the self, a matter which we will discuss in more depth:

(1) The child’s environment (selfobjects) reacts to its utterances as if it already had a self, or a self-experience of its own initiative, integrity, and continuity. That is, if empathic care by the mother is to be successful, she not only takes notice of some of the infant’s requirements and achievements (e.g., the sucking reflex, thrashing, etc.), but speaks to the child as an integral whole (see above).

(2) Psychoanalytical research has pushed back the acquisition of a rudimentary self to an earlier and earlier point. For Kohut, the child’s earliest expressions of narcissistic rage and anxiety are indications of a rudimentary self. An active striving toward total mastery and possession of the dimly perceived exterior of the selfobject manifests itself (earlier than had been assumed)—or, complementary to this striving, anxiety about the disintegration and fragmentation4 of the self (phenomena widely seen in the therapy of narcissistic personality disorders, among others).

Kohut makes repeated reference to the mother’s empathic care for the child’s entire self, by which he means not so much praise and satisfaction of needs as a style, an atmosphere of acceptance. The embryonic grandiose self is reflected "in the radiance of the mother’s eyes"; its own grandiosity shares in the omnipotence of its selfobject. The two are fused, yet take the form of a differentiated whole.5

Early on (in the first through third years of life), the parents will supportively mirror the strivings of the child’s grandiose self and not threaten their idealization as parents. Provided what Kohut calls the nuclear self has established itself, a stronger movement toward disillusionment begins. Margaret S. Mahler’s developmental model of separation and individuation (normalized to self psychology) is applicable here.6  Kohut calls the incorporation of the self-selfobject experience into the internal structure of the self transmuting internalization. What does this mean more precisely?

"The selfobject which possesses a mature psychological organization able to realistically evaluate the child’s needs and how to meet them will be absorbed by the child into its own psychological organization and rectify the child’s homeostatic disequilibrium through behavior" (Kohut 1977, 84).

The first point of this response is important: "absorption into its own psychological organization" is a more technical way of saying empathy: sympathetic understanding (Einfühlung) of the introspection of the Other. Rage and anxiety, for example, do not derive from the primordial drives which the mother must tame through love or strength (neutralized aggression); rather, they are "the experience of disintegration [of a] formerly complete and complex psychological unit of self-evident approval." The mother senses the child’s anxiety and initially shares it with him, then "picks up the child, talks to it while holding and carrying it, and thus creates conditions that the child phase-appropriately experiences as a fusion with the omnipotent selfobject" (ibid., 84f).

But fusion for the sake of transmuting internalization; relationship is transformed (participant) into structure:7   The selfobject (the mother) responds to the progression of anxiety—disintegration anxiety—experienced in the child’s self with slight anxiety, that is, with an affect signal that leads to calm; the affect abates—i.e., the anxiety fades. What applies in the early area of pre-linguistic affect is equally valid later (e.g., in therapy): I perceive the anxiety, I share it in a self-controlled, moderate fashion, and show how it can be made more intelligible and how it might go away. (Kohut’s two-phase transference—understanding and interpreting—and Winnicott’s holding function are both based on this phenomenon.)

A lack of fusion or harmful fusion, i.e., a lack of empathy or inadequate reactions, leads to a weakening of the ability to curb affect or to the acquisition of faulty, defective structures. In our example, if the selfobject responds with a spreading of affect instead of an affect signal. The disintegration anxiety of the self intensifies in the fusion with the mother’s panic reaction. Disregard or ambivalence of reaction patterns of selfobjects not only leads to disintegrating spreading of affect, but can also deprive the child of the security of the self’s reality and the appropriateness of his affect: "Is that really my feeling?"

Kohut’s ontogenetic timeline is as follows:

- Signs of success-oriented strivings and idealized goals become evident in infancy.

- Most nuclear grandiosity becomes consolidated as core strivings (ambitions) at the age of 2-4 years.

- The acquisition of nuclear strivings toward goals—matters of greater or lesser importance or transpersonal ideals which form the meaning of life—lasts until age 6.

To summarize: the nuclear self consists of strivings (ambitions)—which spring from the grandiose self—and transmute ideals, goals, and values—from the idealized parent imago. A tension arc exists between these poles which mobilizes talents and acquired abilities. More precisely: Kohut speaks of a tension arc as the "abiding flow of actual psychological activity ... between the two poles of the self" and of a tension gradient, an "action-producing state which arises between man’s strivings and his ideals" (Kohut 1977, 157).8

Man is established in his life plan with the formation of a specific bipolar nuclear self,9 which can succeed or fail. We now see Tragic Man more clearly.

Kohut’s description makes use of highly questionable metaphors from physics, such as electrical tension gradient (Kohut 1977, 156f). Somatically-related events are swiftly given meaningless causal interpretations.10

Merleau-Ponty, on the other hand, is at pains to describe operative intentionality of an embodied incarnated consciousness.

"Conscious life [is] supported by an ‘arch of intention,’ which lays out for us our past, our future, our human environment, our physical situation ... our moral situation, or rather brings it about that we are situated in all these relationships. It is this arch of intention which makes up the unity of consciousness, the unity of the intelligence, and the unity of the sensorium and motor activity. It is what loses its ‘tone’ in illness" (Merleau-Ponty 1945, 164f).

The psychoanalytic reification of (briefly put) situated, embodied, emotional, and cognitive achievements in concrete terms such as self, object, and so on makes a dialogue with these phenomenological concepts extremely difficult.11  We are entering rough waters where even early Merleau-Ponty is questionable at points.

Let us return to Kohut. The bipolar self owes its existence (constitution) to structure-forming self-selfobject relationships; recall transmuting internalization. This never comes to an end. Relationships of this kind never become superfluous, remain necessary throughout life.

"A person experiences himself as a cohesive, harmonious unit in space and time which is connected with its past and directed toward a creative and productive future, [but] only if he has the experience at every stage of his life that certain representatives of his human environment react enthusiastically to him, are available as sources of idealized strength and calm, present in nursing, but essentially always able to undertand his inner life more or less correctly so that their reactions and his needs are in tune and he is permitted to comprehend their inner life if he requires such support" (Kohut 1984, 84).

By this yardstick, fragmented selves are almost all Kohut can see in the ‘modern world’; he almost laments the disintegration of traditional ideals and values. He speaks frequently of the lack of great artistic or historical figures with the function of an idealized parent imago. Beyond the ordinary case (Grundregel)—i.e., outside the analytical situation—his alternative interpretative horizon is quite limited. Compare Russel Jacoby:

"The frigidity of warm, subjective love persists in its refusal to perceive the social mechanism which produces the frigidity. Stubborn adherence to subjective love drives it to its opposite, to the justification of a loveless world" (Jacoby 1975, 148).

This form of radical social criticism seems quite alien to Kohut. He leaps from dual relationships (mother/child) over the great figures and into "cosmic regions," speaks of "communing in a timeless transpersonal existence" (Kohut 1966, 162); in this way he extricates himself from complications reminiscent of his "leap" from physics into psychism.

After so much self and (tragic) self-realization, where is Guilty Man? Or is the recognition of drive-determined object relations an empty promise? One of Kohut’s critics once made this accusation in the form of a typical joke:

"A Hollywood film magnate is talking non-stop about himself to a young actress. In a rare moment of self-awareness he interrupts himself and says, ‘Enough about me. Let’s talk about you. What do you think of me?" (cited in Eagle 1988, 79).

The "Self" and the Drives

Kohut’s definition of the relationship between drive psychology and self psychology turns out not to be completely consistent. He speaks of complementarity,12 meaning that a psychic phenomenon can be expressed in terms of either, then gives clear primacy to self psychology, which incorporates object relations.13

His model of the bipolar self seems analogous to that of the instinctual drives. If we recall the introductory remarks above, drives are determined by (psychologically inaccessible) source, motivation, object, and goal (= release of tension). The most variable aspect—the object with its oppositional or gratificational character—impinges on hardly any important intraorganismic event. Otherwise Freud’s destiny of the drives would have no tangible meaning.

A source—strivings—and a goal—ideals and values—are also contained in the self, but they derive rather from the respondent self-structures of the parents as from nature (the somatic). Just as ego structures and achievements are based on inhibited gratification, strivings bring out talents and abilities, polarized toward the realization of ideals. Self-realization is not the release of tension, nor does is it fulfilled in actual relationships with others (Intersubjectivity).14

Drive theory vs. self theory: What could this mean exactly? According to classical drive theory, "oral-clinging behavior" is characterized as "the manifestation of

(1) a drive fixation on oral fixation points and

(2) a corresponding developmental arrest of the ego, in consequence of infantile gratifications to which the analysand’s pleasure-oriented immature ego has become addicted" (Kohut 1977, 74).

But this is not the focus of psychopathology, neither developmentally nor dynamically/structurally. As a consequence of the disrupted empathic reactions of the parents, the child’s self was not securely established. That is, in the case of a weak self threatened by fragmentation, the attempt to make certain of oneself in one’s own livingness leads to the "stimulation of erogenic zones, [which] then, secondarily, brings about the oral (and anal) drive orientation of the ego’s enslavement to the drive aims correlated to the stimulated body zones" (ibid., 75).

As with the character pathology of "anal stinginess," the mother, in accepting, rejecting, or ignoring the "fecal gift," is not reacting to the drive, but "to a self that, in giving and offering, seeks confirmation iby the mirroring selfobject" (ibid., 76).

Destructiveness (rage) is also not based on a primary drive, but represents a disintegration product of self-experience. According to Kohut (ibid., 88f), if a patient reacts to an interpretation with rage, that does not show a relaxation of resistance in which a neutralization is reversed and transmuted back into aggressive energy. It would be more in keeping with experience to describe it as representing a developmentally important traumatic situation which is recapitulated in the therapeutic context: a faulty, unempathic reaction of the selfobject refers to a state of affairs in which the child demands total control over reactions and complete empathy.

To summarize in general terms: "The establishment of fixations and of the correlated activities of the ego occurs in consequence of the feebleness of the self. The unresponded-to self had not been able to transform its archaic grandiosity and its archaic wish to merge with an omnipotent selfobject into reliable self-esteem, realistic ambitions, attainable ideals. The abnormality of the drives and the ego are the symptomatic consequences of this central defect of the self" (ibid., 81).

"Symptomatic consequences" means the formulation of a compromise between the demands of the drives and the achievements of the defenses. In self psychology, the latter corresponds roughly to the formation of defensive structures (see, e.g., ibid., 20). They "paper over" the primary defect. Compensatory structures, also established in childhood, compensate for the weakness of the self pole or sector by strengthening the other pole or sector; the self-perception of a weakened self in the realm of ambition and exhibitionism generally encourages the intense pursuit of ideals. We cannot discuss this point further here (cf. Kohut 1977, 160ff). By forming compensatory structures, the individual gets a second chance; according to Kohut, these too have repercussions in therapy.

Drive aims, defense mechanisms, and conflicts represent subordinate areas of the self15 whose disequilibrium can also affect feelings of self-esteem that do not have primary defects. Here we may speak of secondary pathology of the self. But because we do not wish to focus on the classification and therapy of personality disorders, let us turn to another point: If drive-determined symptomatic and character pathologies are disintegration products of a defective self, it might be asked when, where, and how anything determined by the drives becomes "integrated" into the self at all. This will be discussed as part of our reflections on the origins of the self.

From the Virtual Self to the Cohesive Self

Kohut assumes "that the newborn infant cannot have any reflexive awareness of himself, that he is not capable of experiencing himself, if ever so dimly, as a unit, cohesive in space and enduring in time, which is a center of initiative and the recipient of impressions" (Kohut 1977, 95).

And yet his environment reacts to the emotions of the newborn child ab initio as if it were a whole person, it "anticipates the later self-awareness of the child" (ibid.). Kohut speaks here of a virtual self, metaphorically described as "corresponding in reverse to that geometric point in infinity where two parallel lines intersect" (ibid., 96).

In Analysis of the Self (1971), Kohut still agrees with traditional psychoanalytic theory and assumes that the self is formed "through the fusion of its parts," in a sequence opposite to that of the fragmentation that may later occur under adverse circumstances. The child’s experience of himself as a somatic and psychological unit (see above) "takes hold gradually through the fusion of experiences of individual (at first still unassociated) body parts and isolated somatic and psychological functions" (Kohut 1975, 260).

Before a cohesive self forms, "stages of the fragmented self," isolated somatic and psychological fragments or nuclei—self-nuclei—must exist (cf. Kohut 1971, 48). In his "Remarks on the Formation of the Self" (1975), Kohut casts doubt on this thesis: there is no evidence, he writes, for a process of fusion of self nuclei (cf. Kohut 1975, 262f). Here we find the important claim that the development of the child’s self-experience is independent of that of his drive-determined areas. But this self-experience does not only arise independently, it grows in importance as "the experience of the body parts and individual functions first assumes a coordinate position, then a superordinate one. ... The parts do not make up the self, they become incorporated into it" (ibid.).

It seems significant that a Jungian (M. Jacoby 1985) lays particular stress on this importation of holism into psychoanalysis; others of Kohut’s hypotheses also find echoes in holistic psychology.

In seeking the developmental roots of the self, Kohut "obtained the impression that during early psychic development a process takes place in which some archaic mental contents that had been experienced as belonging to the self become obliterated or are assigned to the area of the nonself while others are retained within the self or are added to it" (Kohut 1977, 154).

This passage again refers to the significance of selfobjects, which treat the child’s emotions and the child as an entire self according to their abilities, their own structure, and the image of the virtual self which they contain. The process of differentiation and expulsion takes place in the self-selfobject relationship.

But the passage contains a difficulty. If we speak of self-experience or self-development and the destiny of drive-impulses as separate or posit their integration, the question no longer arises where disintegrated or not yet integrated material is, if the self is in fact to form the "center of a psychic universe" (ibid., 12) and not "part of a psychic apparatus." The two aspects cannot be as arbitrarily divided as Kohut believes. One would in fact have to speak of a self in the self. In our view, however, this would again lead to metaphoric and reifying talk of the self as an object with spatial extension and physical substance.

To return to the main theme: the self is one thing from its beginning;16 it is not the sum of its parts, still less fragments (which would assume the existence of a former whole). There are no separate, then patched-together parts, but rather (to use a formulation related to the work of Merleau-Ponty) increasing inner organization (via differentiation and expulsion) of an amorphous and opaque original unity. Thus the entire body (Leib) at the beginning of life has appellative character—it is all expression: crying, thrashing, looking are all barely separate embodiments of one intention which becomes differentiated later, permitting separation into gesture, mimicry, speech, and acts of movement executed with causal and semantic purpose.

Thus speech does not develop by stringing together meaningful sounds which are then syntactically and semantically joined to form a language. The child babbles, it produces all the sounds of spoken language (modulo its anatomical equipment and maturity), and, via differentiation and expulsion, retains those which are relevant in its language. However, what is expelled—nonsense—can later find its way into neologisms or other linguistic games.

For Kohut, the entire self precedes its elements (for a qualification, cf. 1975, 281). But with respect to the "parts of the body," the "somatic-psychological functions," and their occupation by the libido for "erogenic zones," he make the quite classical assumption that isolated parts are integrated into a whole.

Body, Self, and Beginninglessness

The virtual self (in the imagination) of the Other is not the beginning of the child’s own self-experience; there is no absolute beginning to reflexive experience, but beginninglessness. For consciousness, every beginning is a assumed one, the achievement of a consciousness with a past. We fall into an infinite regress or into the abyss of German Idealism.

Kohut and others notice rarely notice this philosophical problem because the perspective changes in ontogenetic thinking. When we participate empathically, as Kohut asks, in self-experience recalling scenes of childhood or infancy, such as disintegration anxiety, then—in the murk of the earliest pre-linguistic impressions—we change point of view and understand the putative pre-experience in the categories of a functioning organism. "I hold, indeed, that the states existing before the apparatus of the central nervous system has sufficiently matured and before the secondary processes have yet been established must be described in terms of tensions—of tensions increased, of tensions decreased—and not in terms of verbalizable fantasies" (Kohut 1977, 96).

Kohut rightly rejects Melanie Klein’s assumption that fantasies capable of verbalization exist in early childhood. Research on "the competent infant" is emphatic confirmation of Kohut’s position (cf. Dornes 1993). Self-experience of a vague spatio-temporal existence (experiencing); or else an active organism (living). Tertium non datur? I hazard two pieces of evidence for a possible answer:

This problem—from which the great battle over the origins of "self-feeling" or self-experience, and ultimately of self-consciousness, derives—seems to me to be one of the underdetermination of the body and of what Merleau-Ponty calls "intercorporeality." Embodied subjectivity avoids the subject/object split and protects itself by formulating intentional (self-)achievements from reifying concepts.17  The reflexive "I am" is a loan from the Other to the active, intentional, embodied achievements of the child. But the child also possesses a center of urges and perceptions. Only this center, this self, is not one of experience which is reflexive and thus capable of being remembered, but rather one of embodied ability, of operative (fungierende) intentionality (Merleau-Ponty).

One consequence of these reflections would require an investigation of its own—in this respect it remains intentionally null (leerintentional). Just one suggestive concrete example: "If I playfully take the finger of a fifteen-month-old child between my teeth and give it a little bite, the child opens its mouth. And yet it has hardly ever seen its own face in the mirror and its teeth do not resemble mine. But its own mouth and its teeth are directly perceived by the child—felt from inside—as tools for biting, and my jaw is directly perceived by the child—seen from outside—as equipped for the same purpose. ‘Biting’ has in fact an intersubjective meaning for the child" (Merleau-Ponty 1945, 403; cf. 405).18

But "intercorporeality" also takes place in the "mirroring" cited above, which would certainly not have to be restricted to the "radiance of the mother’s eyes." Along with empathy toward the child’s disintegration anxiety, mirroring and holding (Winnicott) and "amodal affect attunement" (Stern) are also meant.

But the embodied union of experience is not an adequate characterization of the context in which the self is born. "Even before language, the baby’s body appears as a text of pleasure and pain ... by which its parents are possessed, because they read their own destiny in every sound, every uproar, every smile, and every tear. In this way, every generation of parents is willingly subjugated by the majestic pantomime of their children" (O'Neill 1986, 254).

Brazelton and Cramer (1990) develop the concept of an imaginary interaction, which includes the directly observable interaction, the subjectiv expectations and the unconscious fantasies of the participants. "The expectations that parents cherish for their future child even before it is conceived are rooted in their own ... history. ... This context (fantasized relationships) is what the child enters when it comes into the world; its spontaneous behavior ‘awakens’ these fantasies and gives rise to interpretations and attributions. The child’s perception of itself is conditioned by this interaction—its self-image develops mirrored in the conscious and unconscious fantasies of the parents about themselves and about the child" (Hamburger 1995, 66). Here the fantasies which make up the child’s self are the parents’; the fantasies make themselves felt in the parents’ ways of behaving with the child.19  That is, they form a structure which is hardly founded in natural conditions,20 but also not adopted in such a way that it is accessible to consciousness, even potentially. The child’s virtual self is laid down in the Other, in his or her plans, wishes, and fantasies—entrusted to it, so to speak. This structure cannot be avoided, but must be called to account: all projective plans inevitably "misjudge" the child’s self.21

Consciousness, self-experience, always comes too late for its own beginning. This primordial lateness cannot be mastered constructively, but there is a reprise, so to speak a heteronomous spontaneity. The time before consciousness awakens—earliest childhood—is not accessible to memory; it remains a dream, a realm of imagination (but not one of willful fiction) in which the—sometimes traumatic—life history has found expression in a style, in the individual limitations of the embodied achievement. Thus this dream too has its (not insignificant) day’s residues.


End Notes

1 Goethe, Faust, Part I, Scene 6  [ return ]

2 Kohut’s ontogenetic hypotheses arose from introspective, empathic therapeutic experiences, not from the results of infant research. The "bipolarity" of the self is based on the discovery of mirror transference and idealizing transference (cf., e.g., Kohut 1971). Wolf (1988)—in specifying and differentiating transference modalities—seems to give bipolarity a background role.  [ return ]

3 "Strictly speaking ... selfobjects are neither self nor object; they are the subjective aspect of a function which is fulfilled by a relationship" (Wolf 1985, 271). Kohut’s self psychology does not want to be an object relations psychology or interpersonal psychoanalysis in the sense of Sullivan or Stolorow et al. (1987). However, misunderstandings continually arise (e.g., confusion of person and selfobject). But the dialogue structure of the psychoanalytic setting cannot avoid the problems of intersubjectivity. Perhaps the constant equivocations should be avoided by distinguishing the Other’s specific intentional relationship (Vermeinensweisen) in each case instead of speaking of reified "objects," "self," and "relationships." However, this lies beyond our scope.  [ return ]

4 This term does not necessarily refer to psychotic phenomena, but to quite mundane, normal psychological experiences. "If, for example, I meet a friend on the street who goes by as if I don’t exist, it doesn’t seem to matter to me that the friend was deep in thought. . . . For an instant I am less myself than before." If someone surprises me by saying I don’t look well, I feel "as if a piece of my self image, which requires but now lacks the confirming resonance of my environment, had been taken away" (Wolf 1988, 29f).  [ return ]

5 Contemporary infant research cast great doubt on the metaphor of "fusion" or symbiosis and a substantial lack of separation between self and object. However, this (older hypothesis) bore on the anthropological premise that there is an original blissful "perfection" which must be disrupted and replaced by an illusion of grandiosity and omnipotence (see above and Kohut 1971, 43). The constitution of world and the self supposedly owe their structure to the necessary transitoriness of an original edenic state. This idea continues in the psychoanalytic contributions to "natural religion."

The fact that the infant has an intentional relationship to the world from the beginning, has curiosity and a desire for effectiveness (a better term for omnipotence), speaks more for its "openness to the world" than for what Mahler calls a "bursting of the autistic shell."  [ return ]

6 Self psychologists now seek their own modification of developmental psychology which can also accommodate forms of pathology (cf., e.g., Wolf 1988).  [ return ]

7 This manner of converting interactive experiences into embodied and psychic structure has a cognitive counterpart in Piaget’s early theory.  [ return ]

8 The entire hypothesis (repeatedly cited) of a tension arc from (a) strivings and ambitions; (b) goals and values; and (c) talents and abilities seems dubious and needs certainly some explication.

With respect to (a): What seems to be intended is something like "the need for confirmation" or "striving for power and success" (Wolf 1988, 224), or striving for self-recognition, self-acceptance, ambition, attractiveness, and the feeling that one’s existence has value. Passivity (becoming admired and recognized) and activity (becoming the center of ambition and success-oriented initiative) are a continuum.

With respect to (b): If (a) represents, roughly, the pole of potentialities relating to experience and action, then ideals and values have a selective character, guiding experience and action. They are the symbolic representations of functions of earlier selfobjects which were incorporated into the self in the process of idealization/identification. The consequence, seen as a critical thesis about ideals and values, recalls the later Nietzsche: "The standpoint of ‘value’ is the standpoint of conditions of preservation and enhancement for complex forms of relative life-duration [here, the "cohesive self"] within the flux of becoming" (The Will to Power, aphorism 715).  [ return ]

9 We cannot go into the "typology of the self" which could be interpolated at this point, e.g., the cohesive, fragmented, empty, overloaded, or overstimulated self. For an overview, see Wolf (1988), 225f.  [ return ]

10 Metaphor of a "bipolar self" with spatial and material structure has weaknesses similar to those of Freud’s structural model. Stolorow et al. (1987) criticize, for example, the concept of a tension arc as "a retreat into mechanistic thinking ... Tension arcs, like drives, are not accessible to empathy. Empathy and introspection can view strivings and ideals as systems of affective meanings which are motivated according to their nature" (page 36).  [ return ]

11 The crux of the reference to Merleau-Ponty is the concept of intentionality, in which the relation (être au monde) logically precedes the relata. Its own definition depends on an act of abstract analysis whose conceptual instantiation must be continually backed up by subjective acts.  [ return ]

12 The complementarity thesis can be basically rejected (e.g., Stolorow et al.1987). In clinical practice the decision often seems to me to be a practical one, dependent less on (meta)psychological background assumptions than on "private theories on the psychoanalytical craft" (cf., e.g., Streeck 1995).  [ return ]

13 Following Kohut, some self psychologists (and other psychoanalysts) have thoroughly distanced themselves even from modifications of the ideas of classical drive theory (cf. note 11). A psychology of affect and its regulation (e.g., Krause 1983) or of "model scenes" (Lichtenberg 1989) seems to emulate Kohut. Interestingly, these models are similarly conceived in terms of the biology of behavior and dress up the old problem of the relationship between energy / causality and meaning (Ricoeur).  [ return ]

14 See note 4.  [ return ]

15 The neglect of sexuality in self psychology was and continues to be an important bone of contention. Wolf’s (1992, 113f) summary shows how sexual experience is interpreted:

"1. Subjective experience of sexual drive is understood by self psychology as a realm of phenomena which express or protect the constitution and cohesiveness of the self. 2. Sensual pleasure can increase to the point of sexual excitement and then achieve a state of temporary structural regression on an orgasmic level in which the components of the self lose their cohesive integration. 3. A strong self emerges from this ... regression with a new configuration of its components which gives it stronger cohesiveness. 4. A vulnerable self can protect itself from the feared regression by means of sexual inhibitions and compulsive configurations. 5. A self with a defective structure which is not able to inhibit and control orgasmic regression fills the structural defect with sexualized images of the necessary selfobject configuration, which is associated with lascivious selfobject experiences in fantasy and behavior." Wolf proposes "to make parallel distinctions with respect to aggression, i.e., normal self-assertive aggression, inhibited self-assertive aggression, and narcissistic rage" (ibid.)  [ return ]

16 If this thesis is correct—and there are good reasons for it outside self psychology—clinical experiences of radical fragmentation or fantasies of a "dismembered body" cannot be interpreted as regression in the sense that they demonstrate a lived/experienced stage of childhood. There would be no memory of it.   [ return ]

17 Cf. notes 4 and 12.  [ return ]

18 Chalking this up simply to the infant’s competence seems as underdetermined as understanding interaction in terms of the biology of behavior. Merleau-Ponty’s manner of speaking (though he did hold a chair in child psychology) seems too speculative to most "baby-watchers." Stern is perhaps an exception; for discussion, cf., e.g., Lichtenberg (1983).  [ return ]

19 Brazelton and Cramer (1990, 162-190) cite three typical interaction models in which the unconscious (and preconscious) fantasies of the parents assert themselves: (a) "The infant as a ghost from the past" (e.g., as a "reincarnation of an ancestor," as a mother figure, or as a judge figure; (b) the "restaging" of old relationship forms; (c) "the child as a part of its parents (e.g., the "villain," the "disappointing baby." Merleau-Ponty has similar observations in his Sorbonne lectures from 1949-52!  [ return ]

20 Here we encounter a formulation of the virtual self which is supplements the concepts of Kohut and Wolf but which also demonstrates the difficulty of making a strict separation between introspective-empathic knowledge and that of social psychology.  [ return ]

21 Here we leave ontological-scientific discourse behind. But the above formulation is an invitation to the thoughts of Lévinas: ideas and projective plans shatter before the Other’s "face".  [ return ]



Brazelton, T.B., / B.G. Cramer (1990). Die frühe Bindung. Stuttgart, 1994.

Dornes, M. (1993). Der kompetente Säugling. Frankfurt.

Eagle, M.N. (1988). Neuere Entwicklungen der Psychoanalyse. Munich.

Freud, S. (1900). Die Traumdeutung, St.A. Bd.1, Frankfurt, 1972.

Freud, S. (1937). Die endliche und die unendliche Analyse, St.A.Erg.,1975.

Hamburger, A. (1995). Entwicklung der Sprache. Stuttgart-Berlin-Köln.

Jacoby, M. (1985). Individuation und Narzißmus. Munich.

Jacoby, R. (1971). Soziale Amnesie. Frankfurt.

Kohut, H. (1966). Formen und Umformungen des Narzißmus. In Kohut (1985).

Kohut, H. (1971). Narzißmus. Frankfurt, 1973.

Kohut, H. (1975). Bemerkungen zur Bildung des Selbst. In Kohut (1985).

Kohut, H. (1977). Die Heilung des Selbst. Frankfurt, 1979.

Kohut, H. (1984). Wie heilt die Psychoanalyse. Frankfurt, 1987.

Kohut, H. (1985). Die Zukunft der Psychoanalyse. Frankfurt.

Krause, R. (1983). Zur Genese der Affekte, Psyche 11/83.

Lichtenberg, J.D. (1989). Modellszenen, Affekte und das Unbewußte. In E. Wolf et al. (eds.), Selbstpsychologie. Frankfurt, 1989.

Lichtenberg, J.D. (1983). Psychoanalyse und Säuglingsforschung. Berlin-Heidelberg-New York, 1991.

Merleau-Ponty, M. (1945). Phänomenologie der Wahrnehmung. Berlin, 1966.

Nietzsche, F. Der Wille zur Macht. Schriften aus dem Nachlaß. Stuttgart.

O'Neill, J. (1986). Der Spiegelleib. In A. Metraux, B. Waldenfels (eds.), Leibhaftige Vernunft. Spuren von Merleau-Pontys Denken. Munich, 1986.

Psychoanal. Seminar Zürich: Die neuen Narzißmustheorien. Frankfurt, 1981.

Stolorow, R.D., B. Bandchaft, G.E. Atwood (1987). Psychoanalytische Behandlung. Ein intersubjektiver Ansatz. Frankfurt, 1996.

Streeck, U. (1995). Private Theorien zum psychoanalytischen Handwerk. In W. Tress and C. Sies (eds.), Subjektivität in der Psychoanalyse. Göttingen-Zürich, 1995.

Wolf, E. (1985). The search of confirmation: Technical aspect of mirroring. Cited in H.A. Bacal / K.M. Newman (1994). Objektbeziehungstheorien: Brücken zur Selbstpsychologie. Stuttgart-Bad-Cannstatt.

Wolf, E. (1988). Theorie und Praxis der psychoanalytischen Selbstpsychologie. Frankfurt, 1996.

Wolf, E. (1992). Narzißtische Lüsternheit und andere Schicksale der Sexualität. In Schöttler/Kutter: Sexualität und Aggression aus der Sicht der Selbstpsychologie. Frankfurt, 1992.



Translated by Joel Rotenberg, he is a translator living in New York.

(Last edition and responsibility for errors and misunderstandings by the author)

The author, Dipl.-Psych. Rudolf Süsske (*1953), studied Psychology, Philosophy and Social Sciences in RUB Bochum and FU Berlin. Since 1990 he has been employed as a psychotherapist and supervisor (Balint-Groups with nurses) in the Department of Psychotherapy and Psychosomatic Medicine (Chair: E. Schiffer) of the Christian Hospital e.V. in Quakenbrück (Germany). Nonclinical interests in hermeneutics, phenomenology (Merleau-Ponty, Levinas etc.) and—at the present—"temporality / remembering in psychotherapy".

The author would appreciate any comments or criticisms about this paper—please e-mail at or visit his web site at

© The Author

[ German Version ]

[ Self Psychology Bulletin Board ]

Published by 3b
© 1998-2000
All Rights Reserved