The Theory of Narcissism
in the Work of Freud and Klein
HANNA SEGAL
&
DAVID BELL
FREUD’S THEORY OF NARCISSISM
Freud’s paper on narcissism marks a watershed in the development of his thought. By 1913 the theoretical model laid out in chapter 7 of The Interpretation of Dreams (1900) had been steadily developed and expanded. The paper “On Narcissism,” however, caused a “disagreeable jolt” and “some bewilderment” (Jones, 1958). This paper saw the first revision of Freud’s instinct theory and marked the beginning of a major return to theoretical questions that received their fullest exposition in the “Papers on Metapsychology” (1915). Reading the paper, one has a palpable sense of Freud’s uneasiness with it. He wrote to Abraham, “The narcissism paper was a difficult labour and bears all the marks of a corresponding deformation” (Jones, 1955).
In the first phase of the development of his theory Freud’s principal object was to trace the vicissitudes of the libido through an examination of the perversions. But by the beginning of the next phase, ushered in by the narcissism paper, he was becoming increasingly preoccupied with the functioning of the ego. Only four years before the narcissism paper Freud had coined the term “self-preservative instincts” (1910b).
Although much of Freud’s theoretical work is couched in instinctual terms, his writings convey a vivid awareness of an internal world. Until the theory of identification was established, however, there was no way of conceptualizing this. Reading Freud, one often gets the impression that his two sides —the literary and the scientific—live uneasily with each other. The theory of identification provided a means by which these two ways of conceptualizing could be linked. In 1916, only a year after writing “Mourning and Melancholia,” he wrote eloquently on the issue of character and the internal world in “Some Character Types Met with in Psycho-Analytic Work,” in which he discusses Shakespeare’s Richard III and Ibsen’s Rosmersholm. In 1914, however, he attempted to conceptualize some very important observations in connection with homosexuality and the psychoses without any explicit theory of internalization or identification.
The paper “On Narcissism” is, as Freud said, “deformed” by the lack of an adequate conceptual apparatus to contain the important observations he wished to consider. It is one of the marks of his genius that, even when there was no adequate theory to deal with certain psychoanalytic observations, he would not abandon the observations. But in the narcissism paper the theory is stretched to the breaking point. These strains were to be brilliantly resolved in three key papers, all foreshadowed in the paper under discussion. These are “Mourning and Melancholia” (1915), which introduced, in a more coherent way, the theory of an internal world based on identification; The Ego and the Id (1923), which saw the founding of the structural model; and Beyond the Pleasure Principle (1920), which contained the second and final revision of instinct theory. Later in this chapter we will take these papers as starting points in examining the development of the theory of narcissism in the work of Melanie Klein.
Freud’s main motives for writing the paper on narcissism were, broadly, clinical and theoretical, on the one hand, and political, on the other. The principal psychopathological problems that occupied him in this period were homosexuality and paranoid psychoses. The theory of narcissism provided the conceptual tool that linked these two psychopathological states. The principal works dealing with these problems are the Leonardo book (1910a) and the Schreber case (1911). As these works are based on speculations from biographical or autobiographical data, we will first examine some elements of the clinical case that Freud had been treating before the publication of the narcissism paper—namely, the case of the Wolf Man (1918). Although the Wolf Man case was written and published after the narcissism paper, all the clinical work was carried out during the four years that preceded it (1910-14).
In his elaboration of the case, Freud paid particular attention to the relation of narcissism to identification. The Wolf Man, after having been rebuffed by his dearly beloved “Nanya,” turned his affections toward his father. According to Freud (1918, 27), he was “in this way able to renew his first and most primitive object choice which, in conformity with a small child’s narcissism, had taken place along the path of identification.” Freud at this time conceptualized identification as a narcissistic act, and such identifications occurred prior to the more mature situation involving object choice. The Wolf Man subsequently developed a passive sexual aim toward his father: “his father was now his object once more; in conformity with his higher stage of development, identification was replaced by object choice.” The case of the Wolf Man thus presents a study of these crucial trends in the patient’s personality—the passive (masochistic)1 attitude, which was central to his repressed homosexuality, and his narcissistic identifications.
The Wolf Man is, of course, the study of a severe perverse character disorder, and it is clear that, as well as studying the vicissitudes of the libido, Freud is examining the whole question of character development in relation to the various predominant instinctual trends and identifications. The Wolf Man shared with Schreber preoccupations with defecation, religion, and identification with a passive figure in a violent intercourse. Although at the time of his analysis with Freud his symptoms were understood as essentially neurotic, his later breakdown revealed that these fantasies formed part of a psychotic core.
Freud conceptualized the repudiation of the passive (feminine)2 currents in terms of the ego’s need to defend itself. He states, “The ego has only an interest in self-protection and preservation of its narcissism.” A key difficulty arises here, which has remained the subject of some controversy—namely, the relation of narcissism to perversion. Freud, in the Wolf Man case, clearly links them but also makes narcissism a part of the ego’s need to protect itself—an activity that is hard to see as perverse.
In this case study one gains a real impression of the internal representation of a scene (a couple in intercourse and an observer) with a cast of characters in which there are shifting identifications (Meltzer, 1978). The representation (or misrepresentation?) of the parental intercourse as a sadistic act has profound implications for character development. In the Wolf Man case, then, Freud gave his first account of an internalized primal scene with shifting identifications, which enter into narcissistic character formation.
It was, however, in the Leonardo work that Freud made his first statement on narcissism.3 This paper also marks Freud’s increasing preoccupation with character, for here he examines a person’s whole life and work in terms of early experience and psychopathology.
Freud (1910a, 100) writes in connection with Leonardo’s homosexuality: “The boy represses his love for his mother: he puts himself in her place, identifies himself with her, and takes his own person as a model in whose likeness he chooses new objects of his love. … He finds the objects of his love along the path of narcissism.” Freud points out that in this identincatory process the boy “preserves his love for his mother.” Nevertheless, he rather ambiguously states, “What he [Leonardo] has in fact done is to slip back to auto-erotism.” This poses another conceptual difficulty because, for Freud, autoerotism implied a state preceding object-love. Yet in his statement, he describes what is clearly an object relationship. We will return to this complicated question later.
A common theme runs through the Wolf Man, Leonardo, and Schreber papers—namely, the relation of homosexuality to narcissism, and the relation of narcissism to psychosis. There is insufficient space here to address the many details of the Schreber case, but it will be remembered that, after an hypochondriacal illness, Schreber developed a psychotic state in which he built a grandiose delusional system in order to restore omnipotently a world destroyed by a catastrophe. The key features of this system were that “in order to restore the world” Schreber had to achieve “a divine relation” to God. this being effected through his transformation into a woman. Freud demonstrates that the end-of-the-world delusion was a projection of an internal catastrophe —that is, the end of the inner world owing to a withdrawal of “interest”4 from it. He explains that, in the psychoses, the libido is withdrawn from objects (that is, the external world) and is turned onto the ego. In this way he links paranoia, narcissism, and megalomania. He goes on to say that paranoiacs are “fixated” at the stage of narcissism. Here he is using the language of the Three Essays on the Theory of Sexuality (1905), where the libidinal organization can become fixated at various stages of development.
When Freud writes of the withdrawal of libido (or rather “interest”) from the world, it sounds like a very passive, quiet procedure. Indeed, he saw all the “noise” of the paranoid patient as being a result of the patient’s attempts to reconstruct the inner world. Meltzer (1978) has noted that at the very point where Freud talks of the withdrawal of the libido, he quotes Goethe’s Faust: “Woe, woe, thou hast destroyed a beautiful world, with powerful fist, in ruins it is hurled, by the blow of the demi-god shattered!” Thus Freud shows his intuitive understanding that the destruction of the inner world is not such a quiet procedure after all but is brought about by a “demi-god with powerful fist.”
The Schreber case must have been what Freud had in mind when he wrote in the narcissism paper: “just as the transference neuroses have enabled us to trace the libidinal instinctual impulses, so dementia praecox and paranoia will give us insight into the psychology of the ego.” Freud goes on to say that Schreber’s delusions are in a certain sense true and, what is more, are consistent with Freud’s evolving theory of narcissism—namely, that when all the libidinal “interest” is withdrawn from the external world the result is an internal catastrophe, the end of the inner world.
Richard Wollheim (1971) has pointed out that psychoanalytic theory not only offers models of mental functions but also describes how these mental functions are represented in the mind. He makes the point that although this is implicit in Freud’s theory it was not explored further until Melanie Klein. In the Schreber case, however, Freud (1911, 75) gives an explicit statement of this: “It remains for the future to decide whether there is more delusion to my theory than I should like to admit or whether there is more truth in Schreber’s delusion than other people may, as yet, be prepared to believe.” For Freud, Schreber’s delusions were concrete representations of those very mechanisms that constitute the psychosis.
The Schreber case provided Freud with fertile soil for his speculations on the nature of psychosis and, in particular, the relationship of paranoia, narcissism, and megalomania. Schreber’s delusional system thus provided Freud with the model of the ego and object-libido later to be elaborated in the narcissism paper.
As for Freud’s political motives for writing the narcissism paper, Strachey points out in his introduction that the concept of narcissism was to provide an alternative to Adler’s “masculine protest” and Jung’s “nonsexual libido.” We think that these major theoretical difficulties were not really resolved until Freud wrote Beyond the Pleasure Principle—in which the duality of instincts was restored with the theory of life and death instincts. Adler’s aggressive drive could then also be accommodated in a way more consistent with Freud’s theoretical structure.
The narcissism paper is difficult to read because it contains a mixture of two different models of mental life—namely, libido theory and an implicit theory of internal object relations. In struggling with the issues raised in the paper Freud uses terms from the former theory in such a way that they lose their original meaning. For example, the term “ego-instinct” implies a very different use of the word “instinct.”5
“On Narcissism” starts with a statement of the method that bore so much fruit in Freud’s first phase—that psychopathology is a study of developmental arrest, and so the study of psychopathology gives rise to theories of development. The paper is concerned with an account of the phenomenology and dynamics of the “narcissistic neuroses” and also explores why this newly gained knowledge necessitates a different model of the mind.
In the paper the term “narcissism” is first used to describe the relationship in which a person takes his own body as his sexual object. In line with the method of the Three Essays, this situation—a perversion—represents a fixation of the libido at an earlier stage of development. It is, however, immediately clear that Freud realized that he was dealing with something more basic than the other perversions, as he goes on to point out that this “narcissistic attitude” is one of the limits of susceptibility of the individual to psychoanalysis.
Freud then looks for evidence of this primary state recapitulated in the perversion. He obtains this from three sources, all of them rather indirect as they are not based on clinical data. They come from the written reports of a psychotic man (Schreber), the observation of children, and accounts of primitive peoples. Freud states, “the hypothesis of separate ego and object instincts rests scarcely at all on any psychological basis.”
As we have said, Schreber’s delusions provided Freud with important evidence for the existence of primary narcissism. Freud links two important phenomenological features of the psychosis—the turning away from the external world and the presence of megalomania—and assumes that these must have some dynamic relation to each other. The statement that the psychotic turns his interest away from the external world and does not replace this with internal fantasies must strike us as strange and requires further elaboration.
As Sandler and his colleagues (1976) pointed out, when Freud talks of the relation to the external world we must assume he means the relation to the mental representation of external reality.6 Freud, however, speaks of “real objects” and “imaginary objects,” the latter replacing the former as objects of interest in the neuroses. “Real objects” must therefore be accurate mental representations of external reality, whereas “imaginary objects,” presumably, are distortions of external reality (by both fantasies and memories). In effect, Freud is saying that psychotics differ from neurotics in the relationship of the ego to the representations of internal and external reality.7
The connection between these two characteristic features of psychosis (withdrawal from reality and the presence of grandiosity) is described in terms of the vicissitudes of the libido. The libido withdraws from objects and cathects the person’s own ego. Freud, in line with this theoretical model, assumes that this phenomenon must recapitulate a previous stage of development in which the libido primarily cathected the ego. This again gives rise to the concept of primary narcissism. Freud thus deduces (on theoretical grounds) the presence of a primary stage of narcissism that precedes object-love.
Wollheim (1971) has pointed out that the differentiation of autoerotism from narcissism gets perfunctory treatment in the paper. To some extent, it depends on whether the “self” or “the body” is taken as the love object. But as the ego is “first and foremost a bodily ego” (Freud, 1923), this distinction remains unsatisfactory. The debate becomes so theoretical because of the lack of clinical evidence, and Freud’s great dissatisfaction with it is clear. The references to the omnipotence of thought of children and of primitive peoples are more illustrative than supportive.
The extent of Freud’s dissatisfaction with the theory as now postulated —namely, of primary narcissism, object-libido, and ego-libido—is clear in the next part of the paper. One of the central difficulties is that he recognizes he is very near to postulating a single kind of psychic energy. This problem is not only political (as it brings him close to Jung) but also theoretical, for Freud’s model entailed a necessary dualism. Without this, it is difficult to account for repression. Freud states that the notion of ego-libido is not “rich in content,” by which we assume he means that it lacks clinical supportive data—always the acid test for Freud. He shows his awareness of the problem by saying that there is a “total absence of any theory of instincts which can help us find our bearings” (78). Freud appeals to biological distinctions (between hunger and love) and, though tempted by this, states clearly that psychological facts must be kept distinct from biological ones. He argues himself round full circle and ends up by saying, “Let us face the possibility of error … [and be] consistent enough (with my general rule) to drop this hypothesis if psycho-analytic work should itself produce some other, more serviceable hypothesis” (79). Such evidence did, in fact, come later from psychoanalytic work. The questions of masochism and repetition compulsion led to the formulation of a new instinct theory—the theory of the life and death instincts.
The question of hypochondriasis raises problems of a similar nature. Freud has again shown his great clinical intuition in locating the hypochondriacal states closer to the psychoses than to the neuroses (as seen in the Wolf Man and Schreber cases). These states must, therefore, be understood in terms of the functioning of the ego. Freud’s later formulation of a bodily ego links hypochondriasis to deeply unconscious fantasies about the body. In this context, however, he had to deal with the question of hypochondriacal anxiety within the ambit of the first anxiety theory. He suggests that the “dammed-up libido” (the cause of anxiety) is located in the ego. He points out pertinently, however, that it is not easy to explain why this should be unpleasurable, since increased quantities of libido located in the ego should lead to feelings of pleasurable omnipotence. Indeed, as he goes on to say, the more interesting question is why libido should leave the ego in the first place—in other words, why infantile narcissism, and the omnipotence associated with it, should ever be abandoned.
Here Freud turns again to literature and finds an answer in purely psychological terms in Heine’s poetic and insightful statement that we need to create in order to be healthy. The whole question of the capacity for object-love (as opposed to self-love) becomes central to mental health. Freud, however, again turns to libido theory in saying that one needs to release libido in order not to suffer neurosis; thus he robs this notion of its central psychological impact.
When Freud turns from the psychoses and the attendant difficult theoretical questions to the whole question of the erotic life, he seems to be much more comfortable. He says that it is in the study of erotic life that we have the “strongest reasons for the hypothesis of narcissism.” The reasons are strong because they are based on clinical data. It is worth pointing out, however, that all the examples given refer to secondary narcissism. He clearly has in mind the Leonardo case, supported by clinical work, when he talks of the man who looks for his own (projected) self in the object of his love. In postulating two different types of love, the anaclitic (love of the object that nourishes) and narcissistic, or self-love, Freud is struggling to define a type of love relation that is not narcissistic. In his description of “complete object-love” of the attachment (therefore nonnarcissistic) type, Freud describes what amounts to enslavement to an idealized object. To our minds, this type of relation, though it contains an acknowledgment of the need for the object, still bears strong narcissistic features. As we will discuss later, such enslavement is brought about through projection of aspects of the self into the object.
Some of the difficulties in this section of the paper arise from Freud’s use of the term “object” in the sense of external objects. Elsewhere, however, he makes it clear that the external object obtains its character from what has been projected onto it. Earlier in the paper he calls this a “transfer of narcissism” to the sexual object, by which he means a projection of the idealized self. He makes further use of this idea in his discussion of the idealization of children.
When we move on to Freud’s first elaboration of the ego-ideal we find him talking more explicitly about an internal world where identifications and projections take place—this being a necessary prior step to external projection (for example, of the ego-ideal) onto external objects. Here he differentiates an internal object that is not in the ego; thus, in effect, he distinguishes self and ego. The ego-ideal, a relic of infantile development, “he projects before him as his ideal.” So Freud shows that an internal scenario in which ideal aspects of the self can be projected underlies the narcissistic object relation so clearly described in the Leonardo case. The capacity for forming an ego-ideal and for projecting it onto other objects clearly has important implications for object choice that are not fully explored in the paper.
Toward the end Freud again comes up against the limitations of instinct theory when he tries to delineate the features of nonnarcissistic love. When he talks of the “return of love” from the object, this seems to be closer to the hydraulic model. He is, as yet, unable to discuss what sort of internal preconditions must exist in order for the ego to obtain the enrichment ensuing from the capacity to love. The question becomes increasingly confusing when Freud writes of the normal happy state of being loved as a return to the primal condition in which object and ego cannot be distinguished—that is, to primary narcissism. When, at the end of the paper, Freud writes of the patient’s preference for cure by love over cure by analysis, he is presumably talking of the patient’s preference for an overwhelmingly narcissistic love, destined to failure as it takes so little account of reality.
The narcissism paper, then, comes at the point where the limits of Freud’s first model have been reached. Implicit in the paper is an object-relations theory of development and a growing interest in the question of character and the internal world. These are, however, not developed further. We suggest that the paper has a number of theoretical problems: (i) the model of the ego- and object-libido, (2) the question of primary narcissism, and (3) the nature and function of the ego.
The model of ego- and object-libido threatened the duality vital to Freud’s system, because the principal mental conflict, as presented, is now between instinctual forces that have the same origin. As Wollheim (1971) put it, “remove the duality and the whole theory of psychoneurosis would surely crumble. … It was precisely this duality that the discovery of primary narcissism appeared to threaten.” The question continued to preoccupy Freud and finds its final resolution in Beyond the Pleasure Principle, where the proposition of the life and death instincts reinstates the duality. The idea of a primary destructive force was taken up most emphatically by Melanie Klein. Its relation to narcissism was explored by Herbert Rosenfeld (to be discussed below).
Primary narcissism—a state that precedes both the formation of the ego and object relations—remains in our view a most unsatisfactory concept. The difficulties with this theory are, to some extent, demonstrated by the different meanings Freud and various authors have given it. For example, it is sometimes referred to as a state between autoerotism and object choice, or an objectless undifferentiated state, that is, preceding autoerotism—a state assumed to be close to the intrauterine state. Here we would agree with the objections raised by Laplanche and Pontalis (1983)—that if one is to accept the existence of an objectless state, then it would be incorrect to call this narcissism, as Narcissus from his point of view perceived an object with whom he fell in love. Phenomenologically this is a state of object relations in which parts of the self are felt to be in the object.
The last section of the narcissism paper boldly presents the beginnings of the structural model. The ego and its functions became an increasingly important preoccupation for Freud after he had written the narcissism paper. In “Mourning and Melancholia” (1915) he was able, for the first time, to give a full account of an internal object relationship that involved projection and identification. This paved the way for the theory, expounded more fully in The Ego and the Id, of an ego built up from “abandoned object cathexes.” This is really the point of departure for Melanie Klein, who went on to explore the constant interplay between projection and introjection in creating an inner world. As we have said, the proposition of the ego-ideal and the observing agency both foreshadowed the structural model with the formulation of the superego. Freud was later to acknowledge Klein (Freud, 1930) when he agreed that the murderousness of the superego bears no resemblance to the reality of the actual parents and must be based on the projection of murderous impulses arising from within. Toward the latter part of his life Freud became increasingly preoccupied with the role of aggression and gave it a much more prominent position.
THE DEVELOPMENT OF THE THEORY OF NARCISSISM IN THE WORK OF MELANIE KLEIN
Although some of the contradictions present in the narcissism paper were partially resolved by Freud’s later work, he still maintained his belief in a primary narcissistic stage that precedes object relations. Klein, in one of her few explicit references to primary narcissism, clearly demarcates herself from Freud. She states:
The hypothesis that a stage extending over several months precedes object relations implies that—except for the libido attached to the infant’s own body—impulses, phantasies, anxieties and defences are not present in him, or are not related to an object, that is to say they would operate [in a vacuum]… . there is no anxiety situation, no mental process which does not involve objects external or internal… . furthermore, love and hatred, phantasies, anxieties and defences are indivisibly linked to object relations. (Klein, 1952; italics added)
In the same paper she goes on to say that states of “narcissistic withdrawal” are actually states in which there is a withdrawal to internalized objects. She thus explicitly departs from Freud’s notion of an autoerotic and narcissistic stage that precedes object relations. We have already said that Freud is by no means unambiguous on this matter.
Klein believed that from the beginning there is a rudimentary ego that alternates between states of relative cohesion and states of unintegration and disintegration. This rudimentary ego forms intense relations to objects and uses defense mechanisms. At first these objects are primitive “part” objects, but with further development these become more integrated. The importance of this is not only theoretical but also clinical, for from this perspective there is no mental state, no matter how regressed a patient may be, that is objectless and conflict-free.
Klein emphasized the constant interplay of projection and introjection in the building up of an internal world of objects to which the ego relates and which are also experienced as relating to each other. It was in her painstaking study of these processes that she demonstrated the rapid oscillations that can occur in the state of the ego and of the internal objects.
Following Freud in the last phases of his work, Klein put great emphasis on the anxiety consequent on the infant’s perception of his own violently destructive impulses toward his objects. She was much impressed by the archaic, murderous qualities of the infantile superego, which bore so little relation to external reality, and saw this situation as based on the projection of the infant’s own destructive impulses—ultimately derived from the death instinct.
A major innovation was her concept of “positions” rather than stages of development. This concept refers to states of the ego, the anxieties that are present, the defenses against these anxieties, and the internal object relations. She described two positions that represent two phenomenologically distinct states: the paranoid/schizoid and the depressive positions.
As we will discuss later, narcissistic object relations are characteristic of the paranoid/schizoid position. In this state the world is deeply split between good and bad objects; this splitting takes place internally and is also projected externally. The dominant anxieties are of a paranoid nature, and the defenses are aimed at protecting the self and the idealized objects from the murderous objects that contain split-off and projected aggression originating in the infant’s self. The defenses of denial, splitting, and projection are characteristic of this position. The basic developmental task is the building up of a secure enough good object in order for further integration to occur. If this is achieved the infant will be better equipped to face and cope with the anxieties inherent in the depressive position in which the infant develops a relation to “whole” objects.
In recognizing that the “good” and the “bad” objects are not in reality separate, the infant requires the inner strength to bear separation, fear of loss, and the guilt ensuing from the recognition of the damage felt to have been inflicted on the good object. The capacity to bear this guilt enormously strengthens the ego. The concern for the internally damaged objects leads to a wish to repair them rather than to deny their existence, and so the infant enters the ethical world. The successful negotiation of these depressive anxieties leads to a much firmer relation to internal and external reality and the capacity to differentiate self from object.
Klein used the term “positions” to emphasize that these are not only developmental stages but two different ways of relating to inner and outer reality, which, to some extent, are always present. Traumatic situations can cause some regression from the depressive to the paranoid/schizoid position, but if there is a secure enough inner good object, such a regression is temporary.
“Notes on Some Schizoid Mechanisms” (1946) is the key paper in which Klein presented her understanding of narcissism. The schizoid object relations she describes are of the type that would today be termed “narcissistic.” In the paranoid/schizoid position the capacity for accurate perception of inner and outer reality is obscured through the mechanisms of denial, splitting, and projection. Internal and external reality are constantly in danger of collapsing into each other, and when they do, the outcome is psychosis. It is for this reason that Klein asserted that the “fixation point” for psychosis is located at a phase of development that precedes the negotiation of the depressive position. Like Freud, she considered that narcissism and psychosis are rooted in a stage of development preceding mature object relationships, but, unlike Freud, she describes this state not as objectless but as involving more primitive object relationships.
In this important paper Klein gave her first detailed description of the mechanism of projective identification. In projective identification aspects of the self are omnipotently denied and projected into the object, which then becomes identified with these projected aspects of the self. When aspects of the self that are felt to be good are projected, this leads to idealization of the object, while the hostile destructive impulses are split off and projected elsewhere. Because of the splitting, paranoia and idealization always coexist, the idealization often being used as a defense against paranoia. Such processes underlie the defense mechanisms of “reversal” and “reaction formation.” It is because the object becomes identified with aspects of the self, to the extent that its real properties are obscured, that this mechanism is the basis of narcissistic object relations. It is a familiar clinical finding that narcissistic or “borderline” patients are equally prone to idealize or denigrate their objects. One relation can quickly change into the other. In both situations the subject has a profound incapacity to see his objects as they “really are.” From the subjective point of view, of course, the objects that become identified with the self are not experienced as part of the self. When Narcissus gazes at his reflection in the water he does not know that what he sees is himself.
Patients who make excessive use of projective identification are trapped in a world made up of projected aspects of themselves. The profound denial and projection lead to a weakening of the ego, which becomes less able to cope with anxiety, leading to further splitting and projection: a truly malignant vicious circle. As well as exploring the basis of these narcissistic object relations, Klein examined what she called “narcissistic states” in which there is a withdrawal from reality toward an idealized internal object. If we recall Freud’s description of Leonardo, it could be said that Leonardo identified himself with his idealized object (mother), while he projected another aspect of himself (the needy, dependent self) into the young men he pursued.
Projective identification of a more extreme type is frequently seen in psychotic patients. The process can be so massive that the patient loses his whole identity and takes over the features of his object (such as delusional beliefs that he is another person, usually someone powerful and famous).
To illustrate, Mrs. A, a psychotic patient, lay rigidly on the couch in an attitude of frozen terror for much of the first year of her analysis. She was later able to explain that when she walked behind the analyst on the way to the consulting room she found herself forced to stare at his behind. She experienced this as a violent assault on him and felt terrified of him. This patient felt she had made a violent intrusive attack with her eyes. The result was that she felt that the violent and intrusive aspects of herself were now present in the analyst, so she felt trapped with a terrifying object from whom she expected retaliation. It was characteristic of this patient that she often needed to hear the analyst talk so that she could determine from the sound of his voice whether he was the terrifying object she took him to be.
Such a situation is common in less ill patients and often underlies acute claustrophobic anxiety. These patients have a deeply held conviction that the attacked object can only retaliate. In this sense all the objects are felt to be as incapable of coping with the situation as they feel themselves to be. Patients who use projective identification to this extent are overwhelmingly preoccupied with the state of their objects. They listen to interpretations attentively, not in terms of what they communicate in words but as revelatory statements about the state of mind of the analyst.
In the latter part of her 1946 paper Klein gives a detailed phenomenology of schizoid object relations. These patients often feel themselves to be unreal or artificial. They can appear remote, having to keep a distance from their objects, which they feel contain terrifying projected aspects of themselves. Alternatively, they develop clinging compulsive ties to their objects, feeling that the loss of the object amounts to an annihilation of parts of the self. It is because projective identification leads to a depletion of the ego that these patients so often complain of feelings of emptiness.
There are many patients who experience love as a threat, fearing that it will deplete them. Male patients sometimes experience this concretely and have theories about becoming incapacitated by the loss of semen. Such patients, in fact, believe themselves to function in a way that corresponds to Freud’s first hydraulic model of the libido. They feel that love is a substance of which they have a finite amount, and so they have to prevent themselves from losing it into their objects. In a certain sense they are correct, because in the love relation they fear, they do lose part of themselves through projective identification. Freud attributes some of these features to anaclitic love. It is for this reason that we have said that anaclitic love, as he describes it, has a strong narcissistic component.
To illustrate, Mr. B led a very limited life marked by profound sexual inhibitions. Although he did well educationally, he had been unable to use his abilities and worked in a rather menial job. Once in his life he had a girlfriend whom he desired. He became madly preoccupied with her and couldn’t bear her being out of his sight. He felt he was breaking down. Following the end of this relationship he lived in a withdrawn, remote way. He proclaimed his self-sufficiency and maintained that he could not understand why anyone should allow another person to become important to him, that such a thing wasted one’s time. Essentially he believed that all objects should be replaceable, and he attempted to live his life accordingly.
Sometimes he would describe situations in which he perceived two peopie who had a real interest and enthusiasm for each other. Such a perception was accompanied by intense pain, with a feeling that something was lacking in him. This momentary awareness, however, was quickly replaced by a state of arrogant superiority in which he would mock such people as being “very infantile.”
He was late for practically every session of the first eighteen months of his analysis, intent on not “wasting time in the waiting room.” He had a long-standing relationship with a black woman whom he saw as being totally dependent on him and for whom he often felt a great deal of contempt. He projected into her all the hated dependent parts of himself. He often experienced his analyst in a similar way, as someone dependent on him. Although Mr. B tried to be self-sufficient, at times he experienced a devastating sense of having wasted his life, accompanied by a terror of “growing old and being alone.” A very envious man, he was always preoccupied with who was better off or worse off than himself. He constantly dreaded discovering that someone he had decided was “no good” was, in fact, more able than he.
The session reported below followed a period in which he was making some steps toward integration and moving away from his narcissistic attitude. He was beginning to feel that his analysis was of some importance to him. The session took place a few days before a break.
He was late for the session and after a brief apology went on to talk of something that had happened the previous evening. He had seen a car behind him which, “from the configuration of the headlamps,” he realized was the same model as his analyst’s. (It was also the same as his own, for a few weeks after starting his analysis, he had purchased a car of the same type and model.) He became intensely preoccupied with the need to know if it was the analyst’s car. Some features of the car were different, and the driver was a woman (the analyst was a man). He felt that his need to establish to whom the car belonged was driving him mad.
An interpretation was made along the lines that he wanted to take a really good look inside his analyst to see what sort of a person he was, in particular to see if he was different from himself—something he had previously thought unlikely. He seemed relieved and also interested in this interpretation and went on to discuss a situation he had described before but never so vividly. He said that whenever he saw someone he believed to have certain valued qualities that he himself did not possess, he felt an immediate need to merge with or get inside this person. He called this process “colonization.” He described the urge to get inside the object as unbearable. In such situations he had an intense urge to masturbate, but he tried to resist it as he felt it wasted something. He also explained that this often happened when he suddenly saw someone he had thought was “no good” in a new light.
This material gives a vivid description of this patient’s difficulties and of how he coped with them. Because he functioned so much through projective identification (for example, by projecting the needy aspects of himself into his girlfriend and into the analyst who waited for him), his life was dull and repetitive. All his objects appeared similar because, in reality, they were containers for the projected aspects of himself. In this session he seems to have been very struck by the interpretation that was made, which he thought was new and which allowed him to see his analyst in a new light, as someone with qualities that were valuable and important to him. In this situation the analyst was experienced as being separate from him and not controlled by him. This time he did not, as he often did, quickly mock the interpretation and render it meaningless, nor did he take it over and make it his own (in the same way that he had made the analyst’s car his car). He felt separate from the analyst and was immediately faced with unbearable feelings of desire for an object that he did not possess. The wish to get inside the object through “colonization” was a wish to wipe out the separation from the object and to possess it quickly and greedily. It is likely that he also projected good aspects of himself into his object and was desperate to maintain contact with them, through “colonization.”
This material represented an important move forward in which he could retain the capacity to see a desired object that was separate from himself. But this was immediately followed by a wish to repossess it in masturbatory fantasy.
An important aspect of this material is the question of envy—Mr. B’s preoccupation with establishing whether people were better off or worse off than he. He moved with remarkable facility from seeing a good esteemed object to representing the same object as worthless. This procedure seemed to be brought about by unbearable envy and it also defended him against envy.
The role of envy in narcissistic disorders has received increasing attention from Kleinian authors. The question of primitive envy received its first full treatment in the short book Envy and Gratitude (Klein, 1957). In this work Klein shows that envy is a psychological manifestation of the most destructive human impulses. She quotes Chaucer: “It is certain that envy is the worst sin that is; for all the other sins are sins only against one virtue, whereas envy is against all virtue and against all goodness.” The envious person cannot accept things from the object, for to do so means to acknowledge its worth and separateness. Such people quickly devalue anything that is potentially useful to them (as Mr. B often did). They “bite the very hand that feeds them.” Envy is fundamentally so incapacitating because it is the very goodness of the object that is hated, so that nothing useful can be obtained from it. Beyond this, however, an envious person is also constantly persecuted, for when he enviously attacks his objects, they turn, through projection, from loving objects into envious persecuting ones. Such patients are often extremely anxious about their own possessions, as they constantly feel that others will be envious and rob them.
In Envy and Gratitude Klein shows the close connection between envy and projective identification. The attack on the object is motivated by envy but also defends the subject against envy. She emphasizes that primitive envy is often hidden, split off, and silent. In the transference situation such envy often severely limits the patient’s ability to use any analytical work and underlies severe negative therapeutic reactions.
The difficulties associated with excessive envy, and therefore excessive use of projective identification, are characteristic of the paranoid/schizoid position. Freud, in the paper under discussion, recognizes that the narcissistic attitude limits the individual’s susceptibility to psychoanalytic treatment. From our point of view, this is because such patients have great difficulty in allowing themselves to receive anything of worth from their objects; instead of seeing and making use of it, out of envy they destroy the object and also their own capacity to recognize it as separate from themselves. In “Instincts and Their Vissicitudes” (1915) Freud, discussing primary narcissism, talks of a state in which the infant feels himself to be the source of all satisfaction: “When during the stage of primary narcissism the object makes its appearance, the second opposite to loving, namely hating, also attains its development” (136). He states in the same paper that “hate, as a relation to objects, is older than love. It derives from the narcissistic ego’s primordial repudiation of the external world” (139). If one holds with the theory of primary narcissism, the discovery of the goodness of the external object comes relatively late and leads to narcissistic rage. If, with Klein, one holds that from birth there is a capacity for awareness of the external object, this narcissistic rage is an expression of envy.
So, for Klein, envy is a fundamental attitude and part of the paranoid/ schizoid position. As the title of her book suggests, the polar opposite of this attitude is gratitude. The person who can be genuinely grateful to his objects and who can acknowledge the fact that he is separate from them can develop the capacity for genuine creativity. Because he is less envious he is less persecuted by envious objects and has a more secure relation to a good internal object and therefore can learn from experience. In other words, his object relations are predominantly those of the depressive position. Inherent in this move from the paranoid/schizoid to the depressive position is the lessening of narcissistic omnipotence. The patient’s need to control and take over the analysis lessens with his increasing ability to tolerate his awareness of the analyst as a separate person who can be of use to him.
For Klein, envy is intimately related to the primary destructiveness of the death instinct. The struggle between the life and death instincts is seen as a continuous conflict in development and is represented psychologically by the struggle between love and gratitude, on the one hand, and hate and envy, on the other. These issues have been further considered by Rosenfeld (1971), who made explicit the link between envy and the death instinct. He explored the deep split that exists in some patients between the libidinal or needy part of the self, which wishes to be understood and helped, and the violent, destructive, envious part of the self, which seeks to dominate and triumph over the object and the hated dependent self. Such patients tend to keep their external objects in a devalued state, enviously undermining them while idealizing their own omnipotent destructiveness. To acknowledge a need for help is to put themselves in an unbearably humiliating situation. Whenever the analyst talks to them of anything needy in themselves, they experience this as an attempt to make them dependent—that is, to forcefully reproject dependency into them. Rosenfeld describes these patients as in the grip of a “powerful gang” that seeks to control them and that advertises itself as superior to the analyst. Such patients, if they do allow themselves to be helped, often feel that they are in terrible danger from this powerful gang.
A schizophrenic patient, Mr. C, spent many sessions in a state of blissful, mocking serenity, silently staring at his analyst. Occasionally, he would giggle at something that was said to him or would respond with a superior and patronizing comment. He seemed to be self-sufficient, to be narcissistic omnipotence personified. In one session, however, he informed his analyst that he was in communication with a group of people called “the Scientists” who advised him not to “talk to Dr. Bell [the analyst], who is quite mad.” They told him that dying would be a good thing, for if he were to die, he could live forever. That the patient brought this up was obviously of enormous importance: he wished to escape from the grip of this powerful gang. Shortly after this session he became acutely distressed and threatened to throw himself under a train. He felt in enormous danger, as he had betrayed the secrets of “the Scientists.” His suicidal impulses were based on his terror of “the Scientists” but also on a painful state of confusion; he was no longer sure who was mad and who was sane. This also illustrates the closeness of envy to the death instinct.
It is possible to see this process in a less florid form in many patients who are less ill. Miss D, though clearly often experiencing desperate states of anxiety, would constantly talk about them in a superior way, inviting the analyst to join her in contempt for her own needy self. For example, she would talk of feeling terrible anxiety associated with a fear of dying. She would punctuate these accounts by remarking, in a sarcastic and superior voice, “how peculiar.” She experienced any interpretations aimed at understanding her fears as attempts by the analyst to force her to be dependent and so make her inferior. She would bring letters to her session that she had written at home, which contained long accounts of her desperate need for help. She behaved as if she were not allowed to bring this extremely needy aspect of herself into the analytic situation and could only, so to speak, slip it in as a secret message. She seemed able to believe only in those relationships in which “one person is up and the other person is down.” In line with this, when she actually did feel in great need—for example, before breaks that affected her deeply—she felt that she was with an analyst who was secretly triumphant and mocking, which meant that she had no one who could help her. It was a long time before she could even contemplate the idea that her analyst could be anything else but triumphant if he was leaving her in a dependent and needy state. In this sense she had a deep conviction that her analyst was identical to herself.
Rosenfeld thought that the “narcissistic organization” is both an expression of and defense against envy. He makes the point that awareness of separateness from objects, an awareness that causes frustration, inevitably leads to envy. He goes on to say that “aggressiveness toward objects therefore seems inevitable in giving up the narcissistic position and it appears that the strength and persistence of omnipotent narcissistic object relations is closely related to the strength of the envious destructive impulses.”
The narcissistic omnipotent aspects of the self often exert a powerful and seductive influence that makes it increasingly difficult to make any contact with the sane, needy aspects of the patient. This is particularly so in patients who are more psychotic, who often hate life and idealize death as a solution to all problems. It is as if they are being lured toward death as a state in which they will be free of all need and frustration. These patients often feel that the analyst is burdening them with the will to live, which they hate. Technically, this is, of course, an extremely difficult situation. Rosenfeld says, “Clinically it is essential to find and rescue the sane dependent part of the patient from its trapped position.”
Mrs. A, whose case was presented earlier, had pursued a career in which she looked after psychiatric patients. In this situation she identified all the patients with the hated dependent parts of herself. She started analysis for “help with some problems,” but she later explained that she believed she was having analysis in order to become a psychoanalyst. This matter was further complicated by the fact that she knew her analyst was undergoing analysis as part of his psychoanalytic training. One year after starting her analysis she had an acute breakdown and was admitted to the hospital, where her analysis was continued. From the patient’s point of view this represented the enactment of her wish to enter the analyst and return to an idealized intrauterine state in which she could surrender all the burdens of living.
So this patient offered two alternatives in order to be able to continue to live. In one, she abolished all separateness and lived inside the analyst, “the psychoanalytic hospital,” a state mat was idealized but, in fact, represented helpless invalidism. The alternative to this was to take over omnipotently all the capacities of her analyst and become him, so mitigating her dependency and envy of the good object. This patient had taken many overdoses and was often in the grip of a delusional belief that the best solution to her difficulties was death, which she felt was an ideal state, promising complete freedom from frustration and the burdens of living. This apparently represented unconsciously the longed-for return to the intrauterine state. To achieve this, however, she had to die.
Segal (1983, 1984) has further explored the idealization of death in these patients. Freud (1924) used the term “nirvana” to describe the seductive pull of the death instinct. Again, to quote Laplanche and Pontalis (1983), “‘nirvana’ evokes a profound link between pleasure and annihilation.” One of the most difficult technical problems with such patients is to distinguish the sane part of the patient, the part that genuinely wishes the analyst’s help, from the destructive part, which construes help as total possession and control.
Six months after leaving the hospital, Mrs. A made some important steps in arranging to get herself looked after by attending a day-center (something she had previously refused to do) rather than, as she put it, “lying in bed all day,” as she felt tempted to do. Attending a day-center meant to her cooperating with her analyst, but to do this also put her in great danger. In a session shortly before a break, she said after a prolonged silence, “I think I don’t want you to know that I need you.” When asked what had happened that had led her to say this, she replied that she had imagined herself in the hospital having a fight with a big powerful patient who suddenly turned on her and pinned her down. She then imagined a particular nurse (someone to whom she had previously turned for assistance and whose help she had appreciated) asking her if she was okay. She imagined herself replying that she was all right. It was this fantasy that had led her to say that she did not want the analyst to know that she needed him.
This material illustrates the constant difficulty such patients present, for when the patient has cooperated and felt helped in a realistic sense, she suddenly feels overpowered by an envious, ruthless, violent part of herself, against which she is helpless and which silences her wish for assistance. This material also illustrates how important it is that the helpful object be active in trying to get access to the silenced, needy part of the patient and not be bought off by the patient’s statement that “everything is okay.”
Further work in this session revealed that the patient believed—or as she put it, told herself—that if she spoke to the analyst about her fear of the break and her feeling that she really needed him, she would be accused of wanting to possess him and not permit him to go away. In this sense she was saying that the analyst would be unable to distinguish the sane, needy parts of herself from those aspects that sought ruthlessly to possess her objects.
The question of the “narcissistic gang” or “pathological organizations” that form the core of these narcissistic difficulties has received increased attention from Kleinian authors during the past fifteen years. Basically, all of them agree that there is a profound split between the sane, dependent parts of the self and a narcissistic, destructive organization that attempts to dominate it. Such patients often feel they can omnipotently take over various aspects of their objects and possess them to avoid dependence and envy. Sohn (1985) describes this as the formation of the “identificate.” These authors agree, too, that the split is also a split between a psychotic and nonpsychotic part of the personality, which is present in all of us (see especially Bion, 1957). Steiner (1979) has emphasized the relative stability that some of these patients achieve; they remain stuck “on a border” between the paranoid/schizoid and the depressive positive. He has also illustrated (1982) how the more perverse parts of the personality try to seduce and corrupt the patient’s sanity. Sometimes these patients feel they have to “do deals” with the destructive parts of themselves. Mrs. A, for example, often had to engage in obsessive counting rituals at various points in the session to ward off “catastrophe,” which usually meant her death, the analyst’s death, or her parents’ death. She felt that if she attended to these procedures for a given amount of time the omnipotent force would be placated and she might then be allowed to turn toward her analyst and listen to him. Sometimes, however, she became identified with this destructive omnipotent organization and in this state violently attacked the analyst’s words, breaking them up into syllables that went round and round in her mind until they became meaningless. This procedure was associated with considerable triumphant excitement.
CONCLUSION
In this essay we have followed the various threads that Freud brought together in his narcissism paper. We have tried to show that it is only with his later formulations, in “Mourning and Melancholia” and Beyond the Pleasure Principle, that he resolved some of the theoretical difficulties inherent in this paper. With the restoration of a dualistic theory of life and death instincts and his increasing realization of the importance of aggression, the whole question of narcissism took on a different complexion. To quote Wollheim (1971) again: “It is only in Beyond the Pleasure Principle that the problem raised by primary narcissism receives its dramatic resolution. From the evident relief with which Freud presented his new position, we can infer the strain under which he and his theory had been placed these last few years” Freud did not abandon the notion of ego and object instincts but grouped them together as part of the life instincts now seen as together opposing the death instinct (Freud, 1940); and, although he accepted at certain times that the first erotic object is the mother’s breast, he continued to assert the existence of a primary narcissistic stage preceding object relations. We have contended that we find this concept clinically not useful and theoretically ambiguous. To quote again from Laplanche and Pontalis (1983): “From a topographical point of view it is difficult to see just what is supposed to be cathected in primary narcissism.”
Some writers have suggested that self-esteem is a healthy remnant of primary narcissism. To our way of thinking, however, healthy feelings of self-esteem have more to do with an internal situation in which there is a secure relation to a good internal object rather than an ideal one.
We have also elaborated Freud’s more intuitive notion of the destructive-ness of narcissism. This is seen in his linking of narcissism to a fundamental resistance to analytic work, his recognition that the narcissistic attitude is opposed to all creativity, and last, but perhaps most important, his understanding of the close link between narcissism and psychosis. These issues have been further explored within the Kleinian school.
From this perspective, stable nonnarcissistic object relations can be achieved only when the depressive position has been negotiated, for it is in this process that there is a differentiation of the self from object. The move toward the depressive position is a move in the direction of a situation in which love and gratitude toward the external and internal good object can oppose the hatred and envy of anything that is good and is felt to be external to the self. The increasing integration and separation resulting from a withdrawal of projections allow love for an object to be objectively perceived. It also means allowing the object to be out of the subject’s control and acknowledging it in relation to other objects. So, by definition, the capacity to negotiate the depressive position also involves a capacity to negotiate the Oedipus complex and allow an identification with a creative parental couple.
In the work of Rosenfeld, Sohn, Segal, and Steiner, the relationship between the narcissistic and nonnarcissistic parts of the self becomes a central focus of the analytic work not only with psychotic patients but also with the less disturbed. The sane awareness of the need for nourishment from an external object that cannot be controlled by the self is the basis of libidinal love, and this bears some relation to Freud’s description of anaclitic love. The narcissistic aspects of the personality do all they can to deny this reality (the reality of dependence) and advertise a superior state of narcissistic self-sufficiency. In some patients this idealization of narcissism takes the form of an idealization of death and a hatred of life.
To conclude, we would like to return to the original Narcissus myth. Narcissus is trapped, gazing at something that he subjectively believes is a lost loved object but that objectively is the idealized aspect of his own self. He believes himself to be in love. He dies of starvation, however, because he cannot turn away toward a real object from whom he might have been able to get what he really needed.
Notes
1. The masochistic identification with a damaged mother was extensively explored in this paper.
2. For Freud at this time, passivity and femininity were synonymous.
3. Although there is a footnote concerning narcissism in Three Essays on the Theory of Sexuality, this was added in 1910, the date of the publication of the Leonardo case.
4. It was in the Schreber case that Freud for the first time used the more general word “interest” rather than “libidinal investment.”
5. This marks a move away from the concept of instincts as having a source, aim, and object toward a more exclusive interest in the aim of an instinct. Indeed, it is hard to say what would constitute the source of an ego instinct (see Wollheim, 1971).
6. Given that the withdrawal from reality really means withdrawal from the mental representation of reality, it becomes clear how this withdrawal leads to an internal catastrophe.
7. As mentioned above, Freud understood the withdrawal of interest from external reality as a passive procedure. One year later, however, he stated in “Instincts and Their Vicissitudes” that the apparent indifference of the ego to the external world in primary narcissism is, in reality, hatred of it. He says, “Indifference falls into place as a special form of hate or dislike.”
REFERENCES
Bion, W (1957). Differentiation of the psychotic from the non-psychotic personalities. Int. J. Psycho-Anal., 39:266-75.
Freud, S. (1900). The Interpretation of Dreams. S.E. 4 and 5.
———. (1905). Three Essays on the Theory of Sexuality. S.E. 7.
———. (1910a). Leonardo da Vinci and a Memory of His Childhood. S.E. 11.
———. (1910b). The psycho-analytic view of psychogenic disturbances of vision. S.E. 11.
———. (1911). Psycho-analytic notes on an autobiographical account of a case of paranoia (dementia paranoides). S.E. 12.
———. (1915). Instincts and their vicissitudes. S.E. 14.
———. (1916). Some character-types met with in psycho-analytic work. S.E. 14.
———. (1918). From the history of an infantile neurosis. S.E. 17.
———. (1920). Beyond the Pleasure Principle. S.E. 18.
———. (1923). The Ego and the Id. S.E. 19.
———. (1924). The economic problem of masochism. S.E. 19.
———. (1930). Civilization and Its Discontents. S.E. 21.
———. (1940). An Outline of Psycho-Analysis. S.E. 23.
Jones, E. (1955). Sigmund Freud: Life and Works. Vol. 2. London: Hogarth Press.
Klein, M. (1946). Notes on some schizoid mechanisms. In The Writings ofMelanie Klein. Vol. 3, Envy and Gratitude and Other Works. London: Hogarth Press.
———. (1952). The origins of transference. In The Writings of Melanie Klein. Vol. 3, Envy and Gratitude and Other Works. London: Hogarth Press.
———(1957)- Envy and gratitude. In The Writings ofMelanie Klein. Vol. 3, Envy and Gratitude and Other Works. London: Hogarth Press.
Laplanche, J., and Pontalis, J. B. (1983). The Language of Psycho-Analysis. London: Hogarth Press.
Meltzer, D. (1978). The Kleinian Development. Strath Tay, Perthshire: Clunie Press.
Rosenfeld, H. A. (1971). Clinical approach to the psycho-analytical theory of the life and death instincts: An investigation into the aggressive aspects of narcissism. Int. J. Psycho-Anal., 59:215-21.
Sandler, J., Holder, A., and Dare, C. (1976). Narcissism and object love in the second phase of psychoanalysis. Brit. J. Med. Psychol. 49:267-74.
Segal, H. (1983). Some clinical implications ofMelanie Klein’s work: Emergence from narcissism. Int. J. Psycho-Anal., 64:269-76.
———. (1984). De l’unite clinique du concept d’instinct de mort. In La pulsion de mort. Paris: Press Universitaire de France.
Sohn, L. (1985). Narcissistic organization, projective identification and the formation of the identificate. Int. J. Psycho-Anal. 66:201 -13.
Steiner, J. (1979). The border between the paranoid schizoid and depressive positions. Brit. J. Med. Psychol., 52:385-91.
———. (1982). Perverse relationships between parts of the self: A clinical illustration. Int. J. Psycho-Anal., 63:241-51.
Wollheim, R. (1971). Freud. Glasgow: Fontana.