ÇÁ·ÎÀ̵å Á¤½ÅºÐ¼® ¿¬±¸¼Ò

»çÀÌÆ®¸Ê


Ä¿¹Â´ÏƼ


Ŭ¶óÀÎ ÇÐÆÄ »çÀ¯ÀÇ ¹ß´Þ °úÁ¤
°ü¸®ÀÚ(cjlee7600@hanmail.net) http://www.freudphil.com
2009³â 04¿ù 12ÀÏ 12:33 5910
Developments in Kleinian Thought: Overview and
Personal View
Elizabeth Bott Spillius


The following paper was originally published in Psychoanalytic Inquiry, 1994, Vol. 14, No. 3, Contemporary Kleinian Psychoanalysis, pp. 324-364. It is reproduced here by kind permission of the Editor-in-Chief Joseph Lichtenberg and the Issue Editor Helen Schoenhals.


Elizabeth Bott Spillius, whose pre-analytic background was in anthropology, is a training analyst of the British Psychoanalytical Society. She has been extensively involved in research and writing about the development of Kleinian and post-Kleinian thinking, and has edited Melanie Klein Today Volume 1: Mainly Theory and Volume 2: Mainly Practice, both published by Routledge in 1988.



Developments in Kleinian Thought: Overview and
Personal View
Elizabeth Bott Spillius, 1994


I. DEVELOPMENTS IN KLEIN'S THINKING

Of crucial importance in Klein's work is that it began in the study and treatment of children. Klein was not the first analyst to treat children, having been preceded by Freud and the father of Little Hans and by Hug-Hellmuth; Anna Freud had started analytic work with older children at about the same time as Klein though along rather different lines. But Klein invented an analytic way of using the technique of play, which gave even very young children under three years of age a suitable medium for expressing their thoughts and feelings, a medium which could easily be combined with their developing capacity to express themselves in language. The invention of this new technique uncovered new data and slowly gave Klein an unshakeable conviction in the reality of the clinical facts she was discovering.

I find the clinical material of her early papers about children absolutely compelling. These papers were among the first things I read in psychoanalysis, and certainly the first that seemed real. Rita, Trude, Peter, Ruth, Fritz, Felix all became persons in a new but somehow familiar world. Later on I found it puzzling that many of my analytic colleagues found Klein very hard to read and her theory sometimes even preposterous. Her style is not particularly felicitous and I had difficulties with the more 'theoretical' papers, but each of the early clinical papers seemed to me like a good anthropological monograph - social anthropology being my profession at this time and my first intellectual love. There were vivid data, just enough theory to make sense of the data, sudden jumps of imagination and theoretical understanding that led on to the next paper. I did not find her clinical descriptions or her theory hard to understand; unconscious phantasy, internal objects, early sadistic superego, psychotic anxiety, a world of utterly good and utterly bad objects, attacking one's mother's insides, desperate wishes to repair the damage - it all seemed familiar, deeply consonant with a vague unformulated sense of things I might have felt as a child. Perhaps I found her work so convincing because I knew very little about psychoanalysis and had no prior attachment to other theories. Indeed, it was reading Klein that stimulated me to read Freud - long overdue - and I found the same sense of discovery of something half-known already. Then came Bion, and gradually many other psychoanalysts, but none had the impact of these first three.

All her life, in spite of the controversies and furore raging around her, Klein thought of her work as following in the footsteps of Freud, as an extension of his work. And indeed, in my view too, there is a consistent allegiance throughout Klein's work to what she regarded as the essential spirit of Freud's approach and technique. But she was an innovator. She regarded the play of a child as the counterpart to the free association of adults. In this play Klein was fully prepared to enact many (though not all) of the roles suggested to her by the child in order to arrive at an understanding of the child's motives and feelings. She was critical in 1927 of Anna Freud for introducing educational elements into child analysis and for emphasising the positive transference and not interpreting the negative transference (Klein 1927a). The descriptions of Klein's technique with adults years later in 1943 (King and Steiner 1991, pp. 635-638) and in 1952 ('The origins of transference' 1952a) are basically very similar to her technique as she described it in 1927, and she clearly thought that both were closely based on Freud. (Anna Freud, however, thought otherwise. See King and Steiner 1991, pp. 629-634.)

Klein was an innovator in ideas as well as in the technique of child analysis. I find it convenient to divide her work into two phases. Up until 1935 she was basically working within the theoretical framework of Freud and Abraham, though she made many changes in it, some of them inadvertent. After 1935, with the two papers on the depressive position (1935, 1940), the paper on the paranoid-schizoid position (1946), and Envy and Gratitude (1957), she developed a new theory of her own. (For a general introduction to the work of Klein, see Segal 1974. See also Hinshelwood 1989, Meltzer 1978, Caper 1988. Petot 1990 and 1991 gives a detailed textual analysis of the development of Klein's thought. Greenberg and Mitchell 1983 discuss her position as an object-relations/drive-structure theorist. Kernberg 1969 and Yorke 1971 present critical reviews.)


A. The First Period of Klein's Work, 1920-1935

The work of this period is innovative, complex, and piecemeal. Klein was discovering new data and working out new conceptualisations of it so quickly that her formulations were bound to be inconsistent, especially as she was holding fast at the same time to the libidinal phases theory of Freud and Abraham (Freud 1905; Abraham 1924). I have found it convenient to summarise her work during this period under a number of conceptual headings, which, in keeping with her explosion of findings and ideas at this time are somewhat unconnected with one another. Further, some are descriptions of her findings and ideas, whereas others are my own inferences about her approach. Some of the ideas of this early period were retained throughout all her work, others were dropped or reformulated. (For a chronological account of this early period, see especially the Editor's notes in The Writings of Melanie Klein 1975, and Petot 1990.)

1. Freud's drives and Klein's drives. During this early period Klein seems not to have seen any difference between her conception of drives and Freud's. But imperceptibly she was making an important conceptual shift. Where Freud thinks of drives as biological forces which become almost fortuitously attached to objects through post-natal experiences, for Klein drives are inherently attached to objects. During this period she had not yet wholly rejected the idea of primary narcissism (which she does later, 1952a), but she was moving in that direction. Even in the early years hers is an object relations theory. Further, she conceives of the individual's own body not as the source of biological drives but as the medium by which the psychological drives of love and hate - mostly hate was explored in these early years - are expressed (Greenberg and Mitchell 1983). Hence Klein's approach is simultaneously a drive theory and an object relations theory, though her 'drives' are becoming increasingly psychological rather than biological, and the role of anxiety in affecting their expression becomes increasingly important as her work develops.

2. Phantasy. Klein hardly mentions phantasy conceptually and gives little sign of realising that she was using the concept differently from Freud. Freud uses the term in different ways, but in his central usage phantasy is resorted to when an instinct is frustrated (1911). For Klein unconscious phantasy accompanies gratification as well as frustration, but, further, it is the basic stuff of all mental processes; it is the mental representation of instincts. This view was not formally stated conceptually, however, until Susan Isaacs' paper 'The nature and function of phantasy', first given in 1943 during the Controversial Discussions (King and Steiner 1991) and later published (1952). It is Klein's view that fantasising is an innate capacity, and that the content of phantasies, although influenced by experiences with external objects, is not entirely dependent on them. She thinks that hate is innate; later she would stress that love too is innate.

Throughout this early period it is implicit that Klein believes that the infant also has innate unconscious knowledge, however hazy, of objects - breast, mother, penis, womb, intercourse, birth, babies - although she does not state this unequivocally until much later.(But see 1927b, pp. 175-176.)

3. Internal Objects and the Inner World. Klein vastly develops the concept of the 'inner' world of internal objects, once again, in this early period, without much conceptual emphasis. In her early clinical work with children she was very much struck by the fact that the internal imagos of parents were very much more ferocious than the actual parents appeared to be. Gradually she developed a conception of internal objects and the inner world as built up through the mechanisms of introjection and projection which she believes operate from the beginning of life. Thus the inner world is not a replica of the external world; experiences of the external world help to shape the inner world, and the inner world affects the individual's perception of the external world. Unlike Freud, she does not restrict the idea of 'internal object' or superego to the single internalisation of parental figures after the passing of the Oedipus complex. (See Hinshelwood 1989 and Greenberg and Mitchell 1983 for detailed descriptions of internal objects.)

4. The Early Superego and the Oedipus Complex. Klein thinks that the children she treated showed clear signs of an early and very sadistic superego (as well as a more developed conscience) which did not correspond to their real parents and which Klein thinks is based on their own sadistic phantasies. (Freud acknowledged this statement of Klein's in Civilization and its Discontents, 1930, p.130.) Klein dates the Oedipus complex progressively earlier and earlier, finally linking it to weaning. At times, like Freud, she links the development of the superego to the Oedipus complex; at others she says that the first introjected object can assume superego functions.

5. Sadism and psychotic anxiety. In her very first papers Klein emphasised libidinal drives and their expression in unconscious phantasy in every activity. (See especially Petot's discussion, 1990.) Klein here means 'libidinal' not in the general sense of 'loving' or 'life-giving', but in the sense of sexuality, involving a somewhat ruthless pursuit in phantasy of sexual aims. Soon afterwards she began, with characteristic enthusiasm, to explore a new terrain, that of aggression and destructiveness, which at this period she almost always called sadism. Up until Freud's Beyond the Pleasure Principle' (1920) and even later, aggression was generally neglected in psychoanalysis as a phenomenon in its own right; it was usually spoken of as a component of the libidinal instinct. In this early period and indeed throughout her work Klein thinks that the mother's breast, her body, and the parental intercourse are the main targets for the projection in phantasy of destructive impulses. This means that the breast, the mother, and the parental intercourse come to be felt as cruel persecutors, and they are then aggressively attacked. During this early period Klein develops the concept of the 'combined object' - the phantasy of a hostile mother containing a hostile penis. Sadistic phantasies arouse intense anxiety, which Klein feels can be the basis of childhood psychosis and of adult mental illness. She develops, in this connection, a new conception of obsessional neurosis as a defence against early psychotic anxiety instead of regarding it as a regression to a fixation point in the anal phase of libidinal development. (See especially 1932, pp. 149-175.)

Klein's concentration on sadism must certainly have been affected by the change in Freud's theory of instincts outlined in Beyond the Pleasure Principle (1920) and by Abraham's work on oral and anal sadism (1924), but I think the main reason for her stress on it came from her clinical work with children, for she found that the children she analysed had extremely ruthless sadistic phantasies about which they characteristically felt very guilty. She then extends her ideas backwards to construct a theory of sadism in infancy, and she thinks of sadism as an important root of the epistemophilic instinct. Towards the end of this phase of her work she begins to distinguish descriptively between anxiety and guilt. But she makes little use of the idea of love during this early period. And in spite of her emphasis on sadism, it is not until 1932 in The psychoAnalysis of Children that she begins to use Freud's idea of the death instinct, and to mention the conflict between life-instinct and death-instinct. Even so, she does not really use the idea of the conflict conceptually until the later phase of her work.

6. The epistemophilic instinct and symbolism. In her very earliest papers Klein talks about the epistemophilic instinct as rooted in libido and expressed in all the child's activities. Gradually she comes to think of sadism as a crucial element in the urge to know. She thinks the infant feels the mother's body to be the source of all good (and bad) things, including the father's penis, and in phantasy the child attacks the mother's body both out of frustration and in order to get possession of her riches. Such phantasied sadistic attacks arouse anxiety, which can be a spur to development. Combined with phantasies of projecting sadism into the mother, anxiety about attacks on her body means that her body is felt to become dangerous. The child is then constantly impelled to find new and less dangerous objects, to make new equations, a process which forms the basis of symbolism and the development of interest in new objects. Klein makes it clear too that such equations are what gives life to children's play, and that the same processes are the basis of transference. If anxiety about attacking the mother's body becomes excessive, it leads to inhibition, neurosis, and in very severe cases to psychosis, as in the case of 'Dick' discussed in 'The importance of symbol-formation in the development of the ego' (1930).

7. The Development of the Boy and the Girl. Here Klein puts forward new ideas of development, emphasising the importance of the phantasied sadistic attacks on the mother's body described above, with their accompanying fear of retaliation and the formation of a severe superego. She thinks that both boys and girls go through a 'feminine phase' in which, out of frustration by the mother and fear of her retaliation for their attacks on her, they turn away from the mother to seek satisfaction from the father and his penis; the phantasied relation with the mother during this phase is one of identification in which the child 'becomes' the mother in order to take her place with the father (1928), a forerunner of at least one form of her later idea of projective identification (1946). In girls this phase is the basis of future femininity, in boys it is normally overcome as Oedipal desires increase. Klein thinks the girl has a lasting fear of damage to the inside of her body because of the sadistic attacks she has made on her mother, and that this is for girls the counterpart of castration anxiety in boys. In The psychoAnalysis of Children (1932) Klein further discusses the complexities of the development of the boy and the girl, stressing, like some other female analysts of the period, the girl's awareness of her vagina. In the later phase of her work Klein revised some of her early views on sexual development (1945). Klein's views on sexual development have interested certain analysts and some feminists, but so far none of her British colleagues has taken up her work in this area.

8. Phases. In her theoretical formulations of this period Klein stuck to the idea of the phases of the libido outlined by Freud (1905) and to the further divisions within them propounded by Abraham (1924), but it is beginning to be apparent that she thinks that anal and phallic phantasies may occur alongside oral ones. One gets the impression that in her clinical observations she largely disregarded the phases, which creates a certain discordance between her theory and her clinical reports.

9. The effect of external parents. In spite of developing an object-relational theory rather than a solely biological-drive theory, Klein does not stress conceptually the actual external parents' personalities and behaviour as part of her theoretical system. She frequently mentions the importance of parents, and her clinical work shows that she related children's behaviour and phantasies to the behaviour and character of the actual parents (see especially Part 1 of The psychoAnalysis of Children 1932 and later The Narrative of a Child Analysis 1960b), but in her theory, especially in the early period, she tends to stress the role of parents as correctives and mitigating factors modifying the anxieties arising from the child's inherently sadistic phantasies. In the later period of her work she explicitly states the importance of the environment (1935, p. 285; 1952c, p. 94, p. 98; 1955, p. 141n3; 1957, p. 175, p. 185n2, pp. 229-230; 1959, pp. 248-249; 1963, p. 312). But it is clear even in the later period that Klein thinks that, even though the character and behaviour of parents is extremely important in shaping the child's development, the child's constitution is also an extremely important factor and the child himself is a very active agent. This view of Klein's has frequently been mistaken as meaning that she thinks the parents (the 'environment') are unimportant, and she has been much criticised for it.

10. Klein's approach to Freud's theories. Klein appears not to have been explicitly interested in the more abstract aspects of Freud's theories, especially his early theories. The idea of the System Unconscious with its own special logic of the primary process does not seem to have caught her imagination; she left it to her colleagues to point out that many of the qualities of the System Unconscious are worked into her concept of unconscious phantasy. She does not distinguish between ideas and feelings that are descriptively unconscious from those that are dynamically unconscious. Perhaps because her early work was not rooted in the topographical model, she does not make a point of the fact that Freud's development of the structural approach in place of the earlier topographical model meant a major change in his basic model of the mind. Her theoretical ideas begin from the structural model of Freud's The Ego and the Id (1923), though she uses his terms somewhat differently and her use of the structural model differed from his because she incorporated her ideas about object relations as an integral part of it. The superego, as described above, is for her earlier and more complex than for Freud. Her idea of the 'id' is not so rooted in biology as Freud's. In the case of the ego, Klein never really distinguishes between the ego and the self, and throughout her work she uses the terms interchangeably, though of course Freud often did this too. Klein does not seem to have realised how important in Freud's thinking was the change from conceiving of anxiety as dammed-up libido to thinking of anxiety as a signal. Even in the early period of her work she was beginning to think of anxiety as a response to destructive forces within the personality.


B. Klein's New Theory: The Paranoid-Schizoid and Depressive Positions, 1935-1960.

The work Klein had done up until 1932, piecemeal and incomplete, was followed by a great leap of imagination which brought her previous work into a new and more integrated synthesis. This was not occasioned by a new method such as the play technique as her earlier work had been; it involved the forming of new thoughts about already known clinical facts and partly worked-out concepts. It is a remarkable achievement of theoretical formulation, perhaps surprising and even mysterious in that Klein was never preoccupied with theory-building as an end in itself.

The new theory consists basically of the delineation of two sets of anxieties, defences, and object relations which Klein calls the 'paranoid-schizoid position' and the 'depressive position'. It has not been easy for other analysts to understand, and many, especially in the United States, have not considered it important or plausible enough to be worth the effort. In Britain, much of Europe, and in South America, however, the theory has had considerable influence, and it is the theory of this later period that has been the basis for most of the developments worked out by Klein's contemporary and later colleagues.

The theory is expounded in four main papers: 'A contribution to the psychogenesis of manic-depressive states' (1935); 'Mourning and its relation to manic-depressive states' (1940); 'Notes on some schizoid mechanisms' (1946); and Envy and Gratitude (1957). A concise statement of the theory is given in 'Some theoretical conclusions regarding the emotional life of the infant' (1952b), though of course without including the concept of envy.

The new theory makes two main changes in the conceptions I have described above as typical of the first period of Klein's thought. I believe that these changes are both necessary for the formation of the new theory but also, somewhat paradoxically, are a consequence of it.

First, she reformulates her earlier descriptions of sadism and aggression in terms of an interaction of life and death instincts as expressed in love and hate. In her view of the death instinct Klein follows Freud quite closely, especially when she is making formal theoretical definitions of it; in clinical contexts she often speaks of 'destructive instincts' or 'aggressive instincts' and sometimes 'self-destructive instincts' without explaining each time the way in which such instincts are derived from the death instinct. In keeping with her view that instincts are inherently attached to objects, Klein's formulation of the death instinct is more clinically-directed and less biological and philosophical than Freud's. Where Freud thinks that the Unconscious contains no idea of death or annihilation (1923, p. 57; 1926, p. 129), Klein thinks that '...there is in the unconscious a fear of annihilation of life' (1948, p. 29). For Klein, this fear of annihilation is the primary anxiety, more basic than birth anxiety, separation anxiety, castration anxiety. Where Freud attributes the deflection of the death instinct to 'the organism', Klein attributes it to the ego (1948, pp. 28-30; 1957, pp. 190-191; 1958, p. 237). Klein thinks that part of the death instinct is projected into the primal object, the breast, which thereby becomes a persecutor, while part is retained within the personality; some of this remaining internal death instinct is turned against the persecuting object as aggression (1946, pp. 4-5; 1958, p. 238n). Like Freud (1923, p. 54) she thinks that some of the internal death instinct is bound by libido, but she also thinks that some of it remains unfused and continues to be an active source of anxiety to the individual about being annihilated from within.

Accompanying her reformulation of sadism and aggression in terms of their derivation from death instinct, Klein greatly increases her use of the idea of love, libido, and of the conception of the good object as the core of normal ego development. Klein had noted the interplay of love and hate in the late 1920s and early 1930s, but at that time she did not make much conceptual use of it. In the later period this interplay becomes central to her new conceptions of the paranoid-schizoid and depressive positions.

The second major change after 1935 is that Klein greatly reduces her adherence to Freud's and Abraham's conception of instinctual phases in favour of a theory of development based on changing modes of internal (and external) object relations. She continues to think that oral expressions of love and hate come first, but she thinks that they overlap with rather than are sequentially followed by anal, urethral, phallic, and genital modes of expression. But instead of phase, in her new theory she speaks of 'position', that is, an organisation of typical anxieties, defence mechanisms, and object relations. She thinks that in infancy the paranoid-schizoid position comes first and is then followed by the depressive position, but she uses the word 'position' rather than phase to emphasise that throughout childhood and indeed also in later life there is fluctuation between the two positions (1952b). Positions as she conceives of them are thus not phases which one passes through and leaves behind. This changed conception made it possible for some of her colleagues, especially Bion, to loosen the connection with literal developments in infancy still further, to the point where the positions are conceived as 'states of mind' regardless of the chronological age at which they are experienced. This emphasis has helped many analysts to look for moment-to-moment shifts in a session from integration and depressive anxiety towards fragmentation and sometimes persecution rather than looking only for major shifts of character and orientation.


The paranoid-schizoid position. At first Klein used the term 'paranoid position'; later she added the word 'schizoid' in recognition of Fairbairn's work (1941, 1944) on splitting of the ego and its relation to schizoid states.
Klein thinks that the normal paranoid-schizoid position occurs in the first three months of infancy and is characterised by persecutory anxiety, that is, fear of annihilation from within and, because the feared malignancy is in phantasy projected outwards, from without as well. She assumes that the infant experiences sensations as 'caused' by malevolent or benevolent objects. Thus hunger in her view is likely not to be just an experience of 'no-food-is-here' but to be something like 'that object is starving me to death', or 'something terrible is attacking me'. A feeling of comfort would be attributed to the benign motive of a good object. It is clear that Klein thinks infants distinguish between self and object, between me and not-me, from birth, though the distinction is based on perceptions shaped by phantasy and by phantasied attributions of motive, and are thus, presumably, very different from the perceptions that would be made by an adult observer. Of course any phrasing of such early perceptions in words is misleading. Susan Isaacs assumes that these very early events are first experienced as sensations, then gradually draw upon plastic images - visual, auditory, kinaesthetic, touch, smell, taste - before becoming linked up with words (Isaacs 1952).

The concern in this very early period is for oneself, not yet for one's object. Klein assumes that anxiety about being annihilated from within is dealt with by splitting and projection. The infant splits his good from his bad feelings and in phantasy projects both into objects felt to be external, 'not me', so that both the ego (self) and object are split. The infant thus lives in a world in which he and some of his objects are extremely bad whereas other objects and other aspects of himself are extremely good. Emotions are labile; good rapidly changes into bad and vice versa, and there is no recognition of the fact that the good and the bad object are the same person. The infant thus lives in a world of 'part' objects, in the sense that what would to an outside observer be one object is to the infant at least two (good and bad). Further, Klein assumes that the first object is a part object, the breast, but in Klein's view this 'breast' is not just a purveyor of food, a satisfier of instinct; it is the source of love, of life itself. She tacitly assumes that in early infancy anatomical part-objects are normally perceived and treated as if they were whole objects and that whole objects may be treated as if they were parts. Full recognition of the identity of objects as wholes and of oneself as a whole in her view comes later, in the depressive position.

Klein uses the term 'projective identification' to describe a complex set of processes by which part of the self is split off and projected into an object to which the individual reacts as if the object were the self or the part of the self that has been projected into it. The individual who projects in this way will then in phantasy introject the object as coloured by what he has projected into it. It is through such constant interplay that the inner world of self and internal objects is built up. Splitting, projection, and introjection are the characteristic mental mechanisms of the paranoid-schizoid position, accompanied by idealisation, denigration, and denial. Omnipotence of thought is thus characteristic of the paranoid-schizoid position. Klein notes that when projection is excessive, objects and the self become fragmented but in her 1946 paper she does not explain why projection should be excessive in some individuals and much less pronounced in others.

Klein thinks that failure in working through the persecutory anxiety and tendency to split of the paranoid-schizoid position are basic pre-conditions for paranoid and schizophrenic illness.

In later papers Klein makes important additions. In Envy and Gratitude (1957) she states that a more than usually marked degree of primary envy, which she regards as a constitutional factor, leads to a pathological paranoid-schizoid position. Because envy attacks the good object, it arouses a premature experience of depressive anxiety about damage to the good object, and interferes with the primal differentiation between good and bad in the object and in the self. Hence it is likely to result in confusion and in very severe cases to confusional states. (See also Rosenfeld 1950.) Such a breakdown of normal splitting leads to difficulty in working through the paranoid-schizoid position and in proceeding to a normal experience of the depressive position.

In a late paper (1958) Klein suggests that the bad objects of the paranoid-schizoid position are not the most terrifying objects; the most terrifying figures are split off into an area of the deep unconscious which remains apart from the normal developmental processes that give rise to the superego. She does not, however, fully work out this idea or integrate it with her other work.


The depressive position. Klein believes that at about three to six months the infant's object relations change from relation to a part-object to relation to a whole object. Although she does not explicitly say so, Klein seems to base this dating on the well-known observation that at some time between three and six months infants begin to look more 'human' and to behave in a much more integrated way. Klein supplements this sort of casual observation with more systematic observations by Ribble (1944). Klein made her own observations of infants (1952c) but these are examples based on her theoretical formulations rather than raw data from which her formulations were derived.

In Klein's view of the depressive position the good and the bad mother are seen to be the same person; the infant begins to feel that the good mother he loves has been damaged by the attacks he has made and continues to make on the bad mother, for they are one and the same. This realisation is extremely painful and gives rise to what Klein calls 'depressive anxiety' as distinct from the persecutory anxiety of the paranoid-schizoid position. It consists of a mixture of concern for the object, fear of its being damaged beyond repair, guilt and a sense of responsibility for the damage one has done. The individual is afraid of losing his object and has a strong urge to repair the damage. The actual state of the external object is extremely important; if the mother appears to be damaged, the child's guilt and despair are increased. If she appears well, or at least able to empathise with her child's problems about her state, the child's fear of his destructiveness is decreased and trust in his reparative wishes is increased. The idea of reparation, already introduced in 'Infantile anxiety situations reflected in a work of art and in the creative impulse' (1929b), now becomes a key concept. The pain of the new integration is sometimes so great that it leads to defences characteristic of the depressive position such as manic and obsessional reparation, denial, triumph, and contempt. If these defences fail, the individual may retreat temporarily or for longer periods to the defences characteristic of the paranoid-schizoid position.

The favourable outcome of the depressive position is the secure internalisation of the good object, which in Klein's view becomes the 'core of the ego', the basis of security and self respect. The individual's future mental health and capacity to love depend on this internalisation. Failure to achieve it constitutes the psychic basis of manic-depressive illness.

In her 1940 paper Klein also adds normal mourning to the phenomena of the depressive position. Mourning in later life reactivates the depressive position of infancy and indeed leads for a time to a feeling of losing all internal goodness. Mourning that is successfully worked through leads to a deeper and stronger establishment of the good internal object.

In her new theory Klein makes a crucial and most interesting link between the Oedipus complex, Freud's 'nuclear complex of the neuroses', and the depressive position. She notes that the onset of the depressive position coincides with the beginning of the Oedipus complex, and says that the sorrow about feared loss of good objects in the depressive position is the source of the most painful Oedipal conflicts, for attacks on one's Oedipal rival are simultaneously attacks on one's loved object (Klein 1940 p. 345; 1952b, p. 110; 1957, p. 196; 1958, p. 239).

In later papers Klein makes additions to some of her early findings; among several others she notes that the dreaded combined object of her earlier work is modified, in the depressive position, by a conception of internal and external parents in a happy relation with each other (1952b). She revises her earlier views of the Oedipus complex (1945). She notes too that transitory experiences of depressive anxiety and guilt can occur in relation to part objects in the paranoid-schizoid position (1948, 1960a).

The delineation of the paranoid-schizoid and depressive positions, combined with the role of early envy in exacerbating the difficulties of the paranoid schizoid position, comprise Klein's final theoretical statement, integrating most of her earlier ideas into a new constellation. The concepts of the paranoid-schizoid position and the depressive position have proved to be exceedingly rich, so much so that their expressions and implications are still being explored.


II. DEVELOPMENTS BY KLEIN'S COLLEAGUES IN BRITAIN

A central feature since the 1950s in Britain has been a decline in the amount of psychoanalysis of children by Kleinian analysts, although child psychotherapy has developed rapidly as a profession. Analysts who continue to work with children are especially interested in trying to bring together developments in child analysis with technical developments that have been worked out with adults.

Interest shifted first to analysis of psychotic patients, especially evident in the papers of Bion, Rosenfeld, and Segal in the 1950s. Work with such patients has continued, though fewer papers have been written about psychotic patients since the 1950s, and the number of papers involving borderline and narcissistic patients has greatly increased. Many developments have occurred through continued work with these and other types of patient: refinements in the concept of projective identification; development of new theories of thinking; new ideas about the paranoid-schizoid and depressive positions; and developments in technique. (My discussion of these topics is closely based on the introductions I have already written to the various sections of Melanie Klein Today, 1988.)


1. Studies of Psychosis. Working with psychotic patients gave many of the analysts who undertook it a deep conviction that the thinking of psychotic patients could be comprehensible and that Klein's ideas about the anxieties and defences of the paranoid-schizoid position were profoundly useful in understanding the way very disturbed infantile object relations inhabit the inner world of the psychotic patient and that these relations could be understood as they were lived out in the relationship with the analyst. This work led Segal and Bion to develop ideas about the process of thinking, and Rosenfeld to productive studies of many topics including confusional states, homosexuality in relation to paranoia, narcissism, and borderline states. (See Rosenfeld 1965 and 1987.)


2. Projective identification Although Klein defined the term 'projective identification' almost casually and was apparently always somewhat doubtful about its value because of the ease with which it could be misused (Segal 1982), the term has gradually become the most popular of her concepts, the only one that has been widely accepted and discussed by psychoanalysts generally, even though this discussion is sometimes incompatible with Klein's conception.

As I have described elsewhere (1988 pp. 81-6, 1992) there has been much discussion about whether the term should be used to refer only to instances where the recipient is emotionally affected by the projection. In my view such restriction would be most unwise, for it would greatly limit the usefulness of the concept and is in any case totally contrary to the way Klein herself used it. I think the term is best kept as a general concept broad enough to include both cases in which the recipient is emotionally affected and those in which he is not. It might be useful, however, to have distinguishing adjectives to describe various subtypes of projective identification; 'evocatory' might be used to describe the sort where the recipient is put under pressure to have the feelings appropriate to the projector's phantasy.

Most of the other questions that have developed in the use of the concept are best answered in the same way, that is, by using the concept as a general term within which various subtypes can be differentiated. The many motives for projective identification - to control the object, to acquire its attributes, to evacuate a bad quality, to protect a good quality, to avoid separation (Rosenfeld 1971a) - all are most usefully kept under the general umbrella.

It is perhaps unfortunate that Bion did not develop a special term for the behaviour the individual uses to induce the other person to behave in accordance with his phantasies of projective identification. Especially when analysing psychotic patients Bion spoke to them in very concrete language because that was the way his patients thought; thus he would say, for example, 'You are pushing your fear of murdering me into my insides' (Bion 1955). This led for a time to a fashion, especially among relatively inexperienced analysts, of speaking conceptually of phantasies actually being concretely put into the analyst's mind. Such usage has been sharply criticised by Sandler (1987), who uses the useful terms 'actualisation' and 'role-responsiveness' to describe the processes by which individuals behave in such a way as to get their object to feel and behave in a way that will satisfy the projector's unconscious wish (1976a, 1976b, Sandler and Sandler 1978). The current practice among British Kleinian analysts, partly because of the criticisms of Sandler and others and especially because of the work of Betty Joseph (1989), is to distinguish conceptually between projective identification as a phantasy and the behaviour unconsciously used by the individual to get his object to behave in accordance with it.

Another change in thinking about projective identification is that the term used to be used almost entirely to characterise a very pathological, primitive defence. It continues to be used in that way when the patient being described is functioning mainly at the level of the paranoid-schizoid position, but it is also used to describe less pathological attributions of self and internal objects to external objects, attributions that are the basis of empathy and characteristic of the depressive position. This distinction between 'normal' and pathological projective identification has occurred largely through the work of Bion (1962a, 1962b, 1963).


3. Work on symbolism, thinking, and experiencing. Two of Klein's ideas have been important starting points for later work on thinking. The first is her theory of symbols described above (p. 00) and the second is the idea of projective identification.

In a seminal paper on symbol formation developed from Klein's ideas
about symbolism, Hanna Segal distinguishes between symbol formation in the paranoid-schizoid position, which she calls symbolic equation, and symbol formation in the depressive position, which she calls symbolism proper (Segal 1957). In symbolic equations the symbol is confused with the object to the point of being the object; her example is a psychotic man who could not play the violin because it meant masturbating in public. In such a state of mind the ego is confused with the object through projective identification; it is the ego which creates the symbol; therefore the symbol is also confused with the object. In the depressive position, where there is greater awareness of differentiation and separateness between ego and object and recognition of ambivalence towards the object, the symbol, a creation of the ego, is recognised as separate from the object. It represents the object instead of being equated with it, and it becomes available for use to displace aggression and libido away from the original objects to others, as Klein described in her symbolism paper (1930).

Bion uses the idea of projective identification in developing a theory of thinking that has had a profound effect on the conceptual and technical repertoire of many analysts (1967a; 1962b; 1963; 1965). In this body of work Bion suggests three models for understanding the process of thinking.

The first model is similar to Segal's idea of an unconscious phantasy being used as a hypothesis for testing against reality (Segal 1964). In Bion's formulation of it, a 'pre-conception', of, for example, a breast, is mated with a realisation, that is, an actual breast, which gives rise to a conception, which is a form of thought. He thinks of the pre-conception as part of the individual's inherent mental equipment, an idea that has affinity with Freud's inherited phantasies (1916-1917), with Klein's notion that the infant has an innate idea of the mother and the breast (1952c and 1959), and the developmental psychologists' idea of predesigning (Stern 1985).

In the second model, a pre-conception encounters a negative realisation, a frustration, that is, no breast available for satisfaction. What happens next depends on the hypothetical infant's capacity to stand frustration. Klein had pointed out that in earliest experience an absent, frustrating object is felt to be a bad object. Bion took this idea further. If the infant's capacity for enduring frustration is great, the 'no-breast' perception/experience is transformed into a thought, which helps to endure the frustration and makes it possible to use the 'no-breast' thought for thinking, that is, to make contact with, and stand, his persecution. Gradually this capacity evolves into an ability to imagine that the bad feeling of being frustrated is actually occurring because there is a good object which is absent and which may or may not return. If, however, capacity for frustration is low, the 'no-breast' experience does not develop into the thought of a 'good breast absent'; it exists as a 'bad breast present'; it is felt to be a bad concrete object which must be got rid of by evacuation, that is, by omnipotent projection. If this process becomes entrenched, true symbols and thinking cannot develop.

The third model has come to be called the formulation of the container and the contained (Bion 1962b; see also O'Shaughnessy 1981a). In this model the infant has some sort of sensory perception, need, or feeling which to him feels bad and which he wants to get rid of. He behaves in a way 'reasonably calculated to arouse in the mother feelings of which the infant wishes to be rid' (Bion 1962a). The projective identification in itself is an omnipotent phantasy, but it also leads to behaviour that arouses the same sort of feeling in the mother. If the mother is reasonably well-balanced and capable of what Bion calls 'reverie', she can accept and transform the feelings into a tolerable form which the infant can reintroject. This process of transformation Bion calls 'alpha function'. If all goes reasonably well, the infant reintrojects not only the particular bad thing transformed into something tolerable, but also, in time, the function itself, and thus he has the embryonic means within his own mind for tolerating frustration and for thinking. Symbolisation, a 'contact barrier' between conscious and unconscious, dream thoughts, concepts of space and time can develop.

The process can, of course, go wrong, either because of the mother's incapacity for reverie or the infant's envy and intolerance of the mother being able to do what he cannot. If the object cannot or will not contain projections - and here the real properties experienced in the external object are extremely important - the individual resorts to increasingly forceful projective identification. Reintrojection is effected with similar force. Through such forceful reintrojection the individual develops within himself an internal object that will not accept projections, that is felt to strip the individual greedily of all the goodness he takes in, that is omniscient, moralising, uninterested in truth and reality testing. With this wilfully misunderstanding internal object the individual identifies and the stage may be set for psychosis.

Of all Bion's ideas, the notions of container and contained and alpha function have been the most widely accepted and more or less well understood. Their adoption has led to a less pejorative attitude towards patients' use of projective identification and to a better conceptualisation of the distinction between normal and pathological projective identification. The container/contained model of the development of thinking has lessened the divide between emotion and cognition. Further, to Bion the external object is an integral part of the system. As described above, Klein has often been accused, wrongly I think, of paying no attention to the environment. (See above p. 00.) Bion shows not only that the environment is important, which Klein also stated, but how it is important. The importance of the environment had been stressed by many other British analysts (Rayner 1991) especially Fairbairn (1941; 1944), Bowlby (1944; 1951), and Winnicott (1945; 1950-1955; 1952; 1956; 1960; and Rodman (ed.) 1987 pp. 89-93 and 144-146) before Bion's formulation of the container/contained model of thinking. The distinctive feature of Bion's construction is that it uses the ideas of projection and introjection to formulate a conception of the internal dynamic involved in the mutual interaction of the container and the contained. He puts a particular emphasis on mental understanding: mental understanding by the other, in his view, is what makes it possible for the individual to develop mental understanding in himself and thus to move towards having a mind of his own and an awareness of the minds of others. Further, he focuses attention not only on the effect of the container on the contained, but also of the contained on the container. His is an 'internal' notion, very much concerned with the modification of thoughts and feelings by thinking. It is a model that he describes rather than an empirical description; it can be applied not only to a mother giving meaning to an infant's fear, or to an envious infant developing an envious superego, the particular mother/baby examples Bion describes, but to many other forms of interaction, including of course the analytic process.

In 'A theory of thinking' (1962a), and indeed in his later work, Bion did not do as much as he might have to link his three models of thinking. It is surely repeated experiences of alternations between positive and negative realisations that encourage the development of thoughts and thinking. And the return of an absent mother gives rise to a particularly important instance, repeated many times in childhood (and in analysis), of a mother taking in and transforming, or failing to transform, the bad-breast-present experience.

In subsequent work (1962b) Bion further elaborates the model of container/contained and thinking as an emotional experience of getting to know oneself or another person, which he designates as 'K', in distinction from the more usual psychoanalytic preoccupations with love (L) and hate (H). He also describes the evasion of knowing and truth, which he calls 'minus K'. He says that K is as essential for psychic health as food is for physical well-being. In other words, K is synonymous with Klein's epistemophilic instinct, though in a more elaborated form.

Bion also develops the idea of fluctuation between the paranoid-schizoid and depressive positions, which he represents by the sign Ps<---->D, as a factor in the development of thinking (1963). This movement back and forth from the paranoid-schizoid to the depressive position was originally pointed out by Klein herself, but Bion focuses on the dimension of dispersal/disintegration (Ps) on the one hand and integration (D) on the other, ignoring for the time being the other elements of the paranoid-schizoid and depressive constellations as described by Klein. Further, Bion's formulation draws attention to the positive aspects of the paranoid-schizoid chaos, to the need to be able to face the possibility of a catastrophic feeling of disintegration and meaninglessness. If one cannot tolerate the dispersal and threatened meaninglessness of the paranoid-schizoid position, one may of course break down; or one may push towards integration prematurely or try to hold on to a particular state of integration and meaning past its time. (c.f. Eigen 1985.)

Bion's work on thinking is used by many analysts and is still being developed and explored, particularly in Britain by O'Shaughnessy (1981a, 1992) and Britton (1989, 1992b and this journal).

Elsewhere (1988 vol. 1 p. 158 and 1989 pp. 107-109) I have briefly described Esther Bick's theory that there is in infantile development a phase of 'unintegration' and 'adhesive identification' which precedes the processes of projection and introjection so crucial to Klein's theory of the paranoid-schizoid and depressive positions and to Bion's theories of thinking (Bick 1968, 1986; see also Anzieu 1989). Although many of Bick's students have used some of her ideas in clinical work, only Meltzer (1975; Meltzer et al 1975), Tustin (1972, 1981, 1990), and Ogden (1990) have attempted to incorporate her ideas into their conceptual system.


4. The Positions and the Concept of Pathological Organisation

The depressive position has continued to be a central conception, though changes have occurred in ideas about it, sometimes through careful clinical and conceptual analysis (J. Steiner 1992) and sometimes without people realising they were occurring. In her own descriptions Klein stresses the integration of part objects - breast, face, hands, voice, smell - to form the whole object; she also stresses the integration of the goodness and badness of the object and of the subject's own love and hate. These features have been retained, but use of the idea of the depressive position in the study of borderline, psychotic, and very envious patients has led to a gradual and increasing emphasis on recognition of the object's separateness and independence as another hallmark of the depressive position. (Recently (1991) Jean-Michel Quinodoz has written specifically on this topic and its connection with loneliness.)

Studies of thinking and artistic endeavour have also shown the very close, indeed, intrinsic relationship that exists between the depressive position, symbolic thought, and creativity (Segal 1952, 1957, 1974, 1991).

A third aspect of the depressive position that has received even more stress than Klein gave it is the intrinsic connection between the Oedipus complex and the depressive position (Britton 1989, 1992a; O'Shaughnessy 1989). As described above, Klein herself drew attention to this connection; I believe that the increased stress on the intrinsic nature of the connection between the Oedipus complex and the depressive position has come about because of the focus on recognition of the object's separateness as a crucial aspect of the depressive position. Once the other person is perceived to be separate, he (or she) is felt to have a life of his own which the subject does not control; the relationship with a third object is the essence of one's primary object's 'life of his own'.

Further explorations of psychosis, addiction, sexual perversion, perverse character structure, but especially studies of narcissism and borderline states, have led to refinements in the understanding of the paranoid-schizoid position and the relation between the paranoid-schizoid and depressive positions. Klein herself made a distinction between the normal paranoid-schizoid position (1946) and the pathological developments that occur when primary envy is very strong (1957). Bion took this further, outlining, especially in the container/contained model of thinking, the processes that can lead to pathology in the paranoid-schizoid position. In his model he mentions two factors: deficiencies in the mother's capacity for reverie, and overwhelming envy in the infant. He implies that other factors in the hypothetical infant may be involved, but envy is the only one he discusses. Gradually the idea of an 'organisation' of interlocking defences has been evolved to order the clinical phenomena encountered, especially those involved in narcissistic and borderline states. Many authors have contributed to the development of the concept, and the word 'organisation' has been in use for some time, first as 'defensive organisation' (Riviere 1936; O'Shaughnessy 1981b), also as 'narcissistic organisation' (Rosenfeld 1971a; Sohn 1985), more recently by John Steiner as 'pathological organisation' (1982, 1987, 1992). In addition, a great many other analysts have used the idea without using the term (Spillius 1988, vol. 1, pp. 195-202); obviously they have influenced one another though evidently without being aware at the time of having a common theme.

There are two main strands of thought in the idea of the pathological organisation. The first is the dominance of a bad self over the rest of the personality; many authors point out a perverse, addictive element in this bondage, indicating that it involves sado-masochism, not just aggressiveness. The second strand is the idea of development of a structured pattern of impulses, anxieties, and defences which root the personality somewhere between the paranoid-schizoid and depressive positions. This pattern allows the individual to maintain a balance, precarious but strongly defended, in which he is protected from the chaos of the paranoid-schizoid position, that is, he does not become frankly psychotic, and yet he does not progress to a point where he can confront and try to work through the problems of the depressive position with their intrinsic pain as well as their potential for creativity. There may be shifting about and even at times the appearance of growth, but an organisation of this sort is really profoundly resistant to change. The defences appear to work together to make a rigid system which does not develop the flexibility characteristic of the defences of the depressive position, and efforts by the individual to make reparation, so characteristic of the depressive position, are usually too narcissistic to bring lasting resolution. There is considerable variation in the psychopathology of pathological organisations, but the analyses of these patients tend to get stuck, either to be very long, only partially successful, or sometimes interminable. The various authors are concerned with the question of whether the destructiveness of these organisations is primary or defensive. Often it is both, and indeed it is implicit in the work of many of the authors that the organisations they discuss are compromise formations, that is, they are simultaneously expressions of inherent destructiveness and systems of defence against it.


5. On Technique.

Strong feelings are experienced about the technique as well as the ideas of Klein and her colleagues. Analysts who are sympathetic to her point of view find the technique rigorously psychoanalytic. Those who are unsympathetic find it unempathically rigid.

(a) Basic features of Klein's technique. As Segal notes (1967) the basic features of Kleinian technique are closely derived from Freud (1911-1915): rigorous maintenance of the psychoanalytic setting so as to keep the transference as pure and uncontaminated as possible; an expectation of sessions five times a week; emphasis on the transference as the central focus of analyst-patient interaction; a belief that the transference situation is active from the very beginning of the analysis; an attitude of active receptivity rather than passivity and silence; interpretation of anxiety and defence together rather than either on its own; emphasis on interpretation, especially the transference interpretation, as the agent of therapeutic change. There is also an emphasis on the totality of transference. The concept is wider than the expression in the session towards the analyst of attitudes towards specific persons and/or incidents of the historical past. Rather the term is used to mean the expression in the analytic situation of the forces and relationships of the internal world. The internal world itself is regarded as the result of an ongoing process of development, the product of continuing interaction between unconscious phantasy, defences, and experiences with external reality both in the past and in the present. The emphasis of Klein and her successors on the pervasiveness of transference is derived from Klein's use of the concept of unconscious phantasy. She conceives of unconscious phantasy as underlying all thought, rational as well as irrational, rather than there being a special category of thought and feeling which is rational and appropriate and therefore does not need analysing and a second kind of thought and feeling which is irrational and unreasonable and therefore expresses transference and needs analysing.

Klein and her successors believe that when patients regress, analytic care should continue to take the form of a stable analytic setting containing a correct interpretive process; the analyst should not attempt to recreate or alter infantile experiences in the consulting room through non-interpretive activities. Even in the development of play technique with children Klein adhered to these principles, except that play as well as talk was the medium of expression. Similarly, in work with psychotic patients, some changes enforced by the patient have been contained without loss of overall method.

(b) Developments in technique. Certain changes of emphasis have taken place in Kleinian technique in the last thirty years or so, partly through belonging to a psychoanalytic society in which there are other points of view, and partly through constant exploration, through being prepared to discard existing accepted procedure. Developments in technique and in ideas have gone along together, each influencing the other. Most of these changes have developed piecemeal and without anyone being very much aware of them at the time; they have been 'in the air' rather than the product of conscious striving.

(i) The interpretation of destructiveness. Both Klein and her colleagues have often been accused of overemphasising the negative. Certainly Klein was very much aware of destructiveness and of the anxiety it arouses, which was one of her earliest areas of research, but she also stressed, both in theory and practice, the importance of love, the patient's concern for his objects, of guilt, and of reparation. Further, in her later work especially, she conveys a strong feeling of support to the patient when negative feelings were being uncovered; this is especially clear in Envy and Gratitude (1957). It is my impression that she was experienced by her patients not as an adversary but as an ally in their struggles to accept feelings they hated in themselves and were therefore trying to deny and obliterate. I think it is this attitude that gave the feeling of 'balance' that Segal says was so important in her experience of Klein as an analyst (Segal 1982). Certainly that sort of balance is something that present Kleinian analysts are consciously striving for. In this respect some of the authors of early clinical papers in the 1950s and 1960s, many of them given to the British Society but not published, took a step backwards from the work of Klein herself, especially from her later work. This was also a period when stated 'belief in the death instinct' was tacitly used, in my opinion, as a sort of banner differentiating Kleinians from the other groups of the British Society. (Perhaps other groups used their opposition to the idea of the death instinct in similar fashion.) Since that time there has been a change, not in the emphasis on destructiveness and self-destructiveness, which have continued to be considered of central importance both clinically and theoretically, but in the way they are analysed, with less confrontation and more awareness of subtleties of conflict among different parts of the personality over them. This change has been influenced not only by the work of Bion but also by Rosenfeld's continued stress on the communicative aspect of projective identification and by Joseph's emphasis on the need for the analyst to become aware of subtleties of the patient's internal conflict over destructiveness and thus to avoid joining the patient in sado-masochistic acting out.

Although the actual term 'death instinct' is now probably used less frequently than it was thirty years ago, there is basic agreement on its importance. There are two emphases, not mutually exclusive. One idea is that individuals with a particularly strong tendency towards inherent destructiveness and self destructiveness tend to attack or to turn away from potentially life-giving relationships, wishing to oblate any awareness of desire that would impinge on their static and apparently self-sufficient state. Another idea, closely related, emphasises what Rosenfeld, following Freud, calls 'the silent pull of the death instinct', which promises a Nirvana-like state of freedom from desire, disturbance, and dependence (Rosenfeld 1987). Both Joseph and Segal also stress the conflict among different parts of the personality over the voluptuous lure of withdrawing into despair, masochism, and perversion.

There are differences in the extent to which analysts believe that marked tendencies to attack positive relationships and/or to withdraw into self-sufficiency are innate or acquired, inherent or defensive. In my view this is a false opposition. From the perspective of treating a particular patient, I think it is impossible to tell what is innate, what has been acquired through interaction with others, and what is the continuing product of that interaction. What one can tell is how deep-rooted the patient's negative tendencies are in the present analytic situation, but this does not tell one whether the deep-rootedness is innate or acquired. And, of course, it is part of the analyst's job to tease out how much his own behaviour may exacerbate his patient's negative tendencies. It is equally important for the analyst to avoid an attitude of blame, whether it be blame of the patient, of his innate

¡ã Go to top