Lost for Words:
The Psychoanalysis of Anorexia and Bulimia
by
Em Farrell
| Contents | Preface | Acknowledgements | Chapter: | 1 | 2 | 3 | 4 | Conclusion | References |
CHAPTER TWO
PSYCHOANALYTIC UNDERSTANDINGS
Beginnings
Sours, unlike other historians of eating disorders suggests evidence can be found for anorexia as early as the eleventh century. He refers to Avicenna, a Persian physician and philosopher, who treated a young prince who had anorexic symptoms and melancholia. His delusional system was successfully broken through, and he recovered. (Sours 1980, p. 205). Rudolph Bell in Holy Anorexia makes a convincing case that Catherine of Sienna (1347-1380) was anorexic, if not anorexic bulimic. She made herself vomit, although she did not apparently eat large amounts of food. Bell is sensitive to, and aware of the dangers he faces in such an enquiry, not only in terms of the different cultural, societal and psychological influences, but also because of the very nature of the material which he is dealing with. Whether the 'holy women' of the thirteenth to fifteenth century did suffer from bulimic anorexia is still being debated, (Bell, 1985 ; Bynum, 1987;) not only because of Bell's uncertainties, but because of an argument put forward by Brumberg, that the symptom of restricted eating was only one component of a complex pattern of behaviour, which included violence directed towards the self ( Brumberg, 1988). This presupposes that such behaviour is not part of the experience of late twentieth century anorexics and bulimic anorexics. However, self harm is part of the behaviour of some anorexics and anorexic bulimics. It may take the form of self mutilation, burning, cutting, scratching, overexercising (which can lead to severe joint problems), laxative, diuretic or other drug abuse. There are many different ways of inducing vomiting, from a simple clenching of the stomach muscles to the use of knives and other objects - even fingernails may damage the back of the throat, and the knuckles of the hands may be scraped by the teeth. There is certainly no doubt that these medieval women fasted, vomited, were overactive and had amenorrhea. They also gained political power and prestige by their actions (once they had convinced those around them that they were neither witches, nor possessed by the devil). Questions concerning the degree of body image distortion, the wish to remain thin and the fear of becoming fat are all unanswerable. What was clear was the subjugation of the flesh and a desire for spirituality and otherworldliness which was to be achieved through starvation, and the physical destruction of the body. Many anorexics and anorexic bulimics wish to destroy their bodies and yet preserve the fantasy of continued existence (Bruch, 1978 ; Palazzoli, 1978 ). In this way a link is clear. The fantasy of immortality is as true for some anorexics in the twentieth century as it was for those in the fifteenth century. 'Death becomes a logical, sweet and total liberation from the flesh' (Bell, 1985, p. 13). Bell is writing about Cathering of Sienna, but the same words might well have applied to Catherine Dunbar, an anorexic who starved herself to death in 1983 (Dunbar, 1986).
What has changed is how eating disorder's are understood. The meaning of the symptom of self-starvation has shifted in society's eyes. Psychotherapists view these thoughts as being psychotic delusions, whilst for the 'holy women' they were the source of the prestige and power they had in their society. In the 1990s the context of religious fanaticism is not a feature of anorexia, but physical fanaticism is - not only amongst anorexics and bulimics. Some individuals become severely addicted to exercise and the control they have over their body. Many women with whom I have worked would like a 'touch' of anorexia. These may be women attending workshops on food and eating, fellow psychotherapists and friends. Reports in the media suggest many are willing to pay high prices for an ideal and young body - an idea taken to extremes in Robert Zemeckis' 1993 film, 'Death Becomes Her' , where women keep their youthful bodies at the expense of immortality. They never die. Young women are one of the fastest growing group of new smokers. The wish to restrict their eating and their appetite is a large and accepted part of their motivation (Evening Standard, 25th November 1993 ). Plastic surgery, liposuction and the prescription of drugs to speed up the metabolism, as well as the multi-million pound business of diet food, drinks and books, suggests most of us share in the unconscious phantasy of a perfect body leading to a perfect life.
Symptoms
The presence or absence of unconscious rather than conscious death wish phantasies cannot be worked out in retrospect in relation to Medieval female religious fanatics. The symptom picture is not clear enough, and their families, lives and individual histories cannot subsitute for working with an individual psychodynamically. The impact, influence and effect of their culture and their society on them as individuals is also hard to determine. From the time of Morton (1689) onwards, a clear and relatively stable symptom picture of anorexia has existed. Surprisingly little has changed over time, although some shifts have occurred between the categories, and some symptoms have left the diagnostic picture altogether, but this does not mean they are not clinically in evidence. Morton, in his Pathisologica or a Treatise on Consumption , written in 1689, documented two cases of anorexia and described them as 'a nervous atrophy, a consumption of mental origin, without fever or dyspepsia, with the symptomology of food avoidance, amenorrhea, lack of appetite, constipation, extreme emaciation, and over activity' (p. 206). He recognised the mental origin, the severe weight loss, amenorrhea and over-activity. This symptom picture remained constant over the next two-hundred years or so. Some clinicians noticed the presence of vomiting in their patients and some did not. Lasègue was criticised for failing to recognise the primacy of amenorrhea. He also dismissed the presence of vomiting in the anorexics he saw. 'There is neither vomiting nor any real desire to vomit even in extreme cases, the patient only asserting that a degree beyond would induce this' (Lasègue, 1873 , pp. 146-7). Most recognised the non-physical nature of the disorder and the powerful influence of the mother on the patient.
Bulimic symptoms were recognised from early on, as was binge eating but what was not seen was the distortion of body image and the fear of becoming fat that informs so much of how eating disorders are understood today. Starving was not connected in their minds to bingeing. Sollier (1891) and Charcot (1889) both recognised a desire to be thin - one half of the fear of becoming fat - in anorexics, but Gilles de la Tourette was the first clinician to recognise that although sufferers refused to eat, there was not a loss of appetite, both perceptions of food and the body were distorted. He was the first to correct the misnomer of anorexia; he rechristened it 'anorexia gastrique'.
From the late nineteenth century onwards attempts were made to understand anorexia in terms of physical illness. Both Gull and Lasègue initially attempted to find a physical cause for anorexia. In reports in the Lancet Gull linked it to a gastric nervous malfunction, and similarly Lasègue (1873) believed it was related to a malfunction of the gastrointestinal tract. The presence or absence of constipation as a symptom of anorexia does not have the relevance which it had in the late nineteenth and early twentieth centuries, although peculiar eating habits, or non-habits do effect the digestive tract, as does the psychological state of the patient. The intestinal focus on constipation in anorexics has perhaps been supplanted by an interest in the taking of laxatives and diuretics by anorexic bulimics and bulimics, which causes diarrhoea and can lead to severe constipation and permanent damage of the gastrointestinal tract. Whether this has taken place cannot be determined until the abuse of laxatives is discontinued. The symbolic meaning, or the frequency of the taking of enemas in the context of eating disorders has not been explored, but physically the delicate balance of the gastrointestinal tract can be adversely effected by this procedure. Constipation, bloating and chronic wind are seen more as symptoms of recovery once eating has been resumed, rather than as being part of the primary picture, amongst the bulimic groups.
The complex interaction between eating behaviour, physical symptoms and psychological and emotional states was dispensed with by many when Maurice Simmonds, a pathologist at St. George's Hospital in Hamburg, suggested that emaciation could be the result of pituitary destruction or deficiency (Simmonds, 1914, 1916 .) For the next twenty years many anorexics were diagnosed as suffering from panhypopituitarism and were treated with extracts from the pituitary gland. Simmonds' findings were misleading and meant that by 1916 most cases of anorexia were treated as physical disorders. The concept of Simmonds' disease was challenged in the 1930s,
In 1936, Ryle demonstrated that psychosexual trauma could lead to amenorrhea. His psycho-endocrine thesis was further elaborated by Reifenstien in 1946, when the latter described several cases of amenorrhea demonstrably due to psychophysiological causes. This research clearly established that developmental traumas and interferences, as well as psychosocial stress, can alter hormone patterns and secretions and led to a burgeoning of interest in the psychobiology of anorexia nervosa (Sours, 1980, p. 210).
This discovery led to a renewed search for understanding the links between anorexia, the individual and society, however the dehumanization of eating disorder patients and their symptoms is a trend that chillingly continues today:
The investigation of structural brain abnormalities in patients with bulimia nervosa is a continuing topic of current psychiatric research... Research by Lautenbacher, Galfe, Hoelzl and Prike (1989) on similar lines, revealed delayed gastrointestinal transit in a sample of patients with bulimia (Parry Jones, 1991, p. 140 ).
Alternative Psychoanalytic Understandings of Symptoms
Eating disorders may be seen as psychosomatic illnesses; indeed, Kaufman and Heiman see anorexia as paradigmatic in The Evolution of Psychosomatic Concepts; Anorexia Nervosa, A Paradigm, which they edited in 1964. This does not, however, alter the basic question which remains: how are symptoms described, and what are the meanings that are attached to them? The bulk of the literature on eating disorders originates in the United States where the majority of therapists are ego psychologists, following on in the tradition of Anna Freud, as interpreted by Kris (1950) , Lowenstein (1951) , Hartman (1958) , and Rappaport (1960) . The two main strands represented in working with eating disorders are those who use a conflictual model (Sperling, 1978 ; Wilson et al., 1985, 1992 ) and those who follow a Kohutian deficit model (Goodsitt, 1983, 1985 ; Swift and Letven, 1984 ). For the former group, a symptom 'is viewed as the end product of a complex developmental series of childhood wishes and fantasies, and defensive transfigurations and revision' (Schwartz, 1988 , p. 36). It remains firmly in the hands of the instincts, particularly the sex drive. According to the ego psychologists a symptom represents a repressed wish and denial of that wish, as can be seen with Freud's understanding of hysterical vomiting quoted below. For Kohut and his followers eating disorders are understood as a way to supply missing self-object functions (Brenner, 1983 ; Geist, 1985 ; Gehrie, 1990 ; Sands, 1991 ). In many ways this comes close to an object relations approach, and is a very persuasive way of working with eating disordered patients. Susan Sand's paper on 'Bulimia, Dissociation and Empathy: A Self-Psychological View' (1991) is extremely clear and persuasive, but its emphasis seems to me to fail to give due attention to the violence of the unconscious phantasies of patients and to the importance of coming to terms with adult sexuality and some kind of a working through of the Oedipal situation.
I shall look at the understandings of bulimia and anorexia from these points of view before moving on to an object relations approach as understood by the Kleinian school, where the symptom is seen as occurring within, and emerging as, part of a phantasy (fantasy with a 'ph' always meaning it is unconscious) of an internal object relationship. Accordingly, a symptom is understood as being an expression of an object relationship, rather than being set off by sexual or aggressive drives, although aggression and sexuality may well feature amongst other emotions in the complicated nature of the exchange. This way of understanding anorexic and bulimic symptoms anchors them in the primitive internal world of object relations, unconscious phantasy, primitive anxieties, and pathological narcissism that Kleinians believe occur from birth onwards, unlike the ego psychologists, for whom object relations do not occur until nine months or so (Mahler et al., 1975 ).
Although the ego psychologists call anorexia a psychosomatic illness, their Oedipal and libidinal understanding of the symptom suggests they see it as conversion hysteria. This is strongly argued against by Christopher Dare: 'Psychosomatic symptoms are not hysterical conversions; the elucidation of the fantasies with which the physical symptoms are associated in the patients mind do not reliably provide relief from the symptoms' (Dare, 1993 , p. 12). Despite these objections, a great deal of work on understanding anorexic and bulimic symptoms has been explored in terms of a regression from the Oedipus complex - both positive and negative - and it this territory that I shall look at first.
Psychological Understandings of Anorexia
Pre-Freudian
Prior to the development of psychoanalytic ways of thinking, it was acknowledged that anorexia was an illness whose origins lay in feelings and the mind, rather than being biologically determined. As we have seen Morton , as early as 1689, not only provided an accurate description of anorexic symptomology but also recognised its psychological origins. Gull (1873) and Lasègue (1873) moved from a purely biological to a psychological understanding. As Kaufman and Heiman (1964 , p. 142) have observed, 'Lasègue was so impressed with psychological factors in control of appetite that he tried to explain even the voraciousness of the diabetic as being of psychological origin'.
From the late eighteenth century, the psychological influence of the mother was seen as being a major, although rather mysterious, part of the problem (Nandeau, 1789 ; Gull, 1873 ; Lasègue, 1873 ; Charcot, 1889 ). The historical emergence of the primacy of the mother-child relationship and the relevance of the pathology of the mother is explored in more detail in the chapter on transitional phenomena.
Classical Freudians
Freud (1899) , Osler (1912) and Riddle (1914) took a different tack and linked bulimic symptoms of food cravings and vomiting to hysteria, which, despite Gull and Lasegue's earlier objections, continued to be seen as the aetiology of anorexia. In his Three Essays on The Theory of Sexuality Freud (1905) brings nourishment and sexuality firmly together; they are only separated once weaning occurs. He says of the first psycho-sexual phase, the oral phase:
The first of these is the oral or, as it might be called, cannibalistic pregenital sexual organization. Here sexual activity has not yet been separated from the ingestion of food; nor are opposite currents within the activity differentiated. The object of both activities is the same; the sexual aim consists in the incorporation of the object - the prototype of a process which, in the form of identification, is later to play such an important psychological part (Freud, 1905, p. 337).
The possible divergence of two lines of thought about the meaning of symptoms can be traced from this quote. Does the taking in of nourishment remain for eating disorder patients as a sexually driven activity, or is it a prototype of a way of object relating based on introjective identification, or is it both?
Freud refers to hysterical vomiting in his case history of Dora and in a letter to Fliess.
Because in phantasy she is pregnant, because she is so insatiable that she cannot put up with not having a baby by her last phantasy-lover as well. But she must vomit too because in that case she will be starved and emaciated, and will lose her beauty and no longer be attractive to anyone. Thus the sense of the symptom is a contradictory pair of wish-fulfilments (Freud, 1899 , p. 278).
It is clear that he does not view it as being a fixation at the oral stage but rather as a defensive regression against positive Oedipal wishes.
Freud recognised the displacement of genital wishes to the mouth and the unconscious existence of the mouth-vagina equation. Jones (1927) went one step further and recognised the mouth-anus-vagina equation. These conceptualisations, particularly the former, have been used by many as a template to understand anorexia and bulimia throughout the twentieth century (Greenacre, 1950, 1952 ; Fraiberg, 1972 ; Sperling, 1973 ).
The transformation of a conflict on the genital level to a conflict on an oral basis is part of the ego's effort to gain mastery over a genital conflict by a change of venue, so to speak, by shifting the struggle to a safer, more familiar and more controllable ground (Ritvo, 1984, p. 454).
It provides an accessible route to one way of understanding anorexia as being a defence against a wish for impregnation.
Post-Freud
Following Freud's earlier reasonings, Waller, Kaufman and Deutsch (1940) , in a classic paper, elaborated in much greater detail the nature of the particular oral phantasy which they felt was present in anorexia: 'Anorexics have psychological factors that have a specific constellation centring around the symbolization of pregnancy fantasies involving the gastrointestinal tract' (Waller, Kaufman and Deutsch, 1940, p. 260). Their understanding is also of relevance to anorexic bulimics and to those normal weight bulimics who do have amenorrhea.
We see then, a syndrome the main symptoms of which represent an elaboration and acting out in the somatic sphere of a specific type of fantasy. The wish to be impregnated through the mouth which results, at times, in compulsive eating, and at other times, in guilt and consequent rejection of food, the constipation symbolizing the child in the abdomen and the amenorrhea as the direct psychological repercussion of pregnancy fantasies. This amenorrhea may also be part of the direct denial of genital sexuality (ibid., p. 272).
The complicated nature of the defences, displacements and hidden gratification of wishes is well described by Schwartz.
To the child and the unconscious, food is the paternal phallus, ingestion of which undoes castration and conceives the oedipal baby (Waller et al., 1940 , Lorand, 1943 ; Leonard, 1944 ; Sylvester, 1945 ; Blos, 1974 ). The incorporative act of eating-gorging contains the desire for abdominal distension and impregnation, the defence of upward displacement with oral submission to mother, and the punishment of physical revulsion. The expulsive act of vomiting desexualises the receptive wish, symbolically rejects and restores the ingested phallus-baby, sadistically punishes the thwarting object, and masochistically relieves the guilt evoked by the desire to castrate and possess father (Masserman, 1941 ; Leonard, 1944; Kestenberg, 1968 ) (Schwartz, 1988 , p. 39).
To Have a Penis?
Following this line of thought the question then arises as to the role of father's penis and what it means in phantasy for the female child to have a penis. Is it a wish for a baby from father, which is unconsciously meant to equal a penis, as in the classical understanding of the positive Oedipus complex? Or is it about wishing to have a penis, to identify with the masculine, in effect, to take up what Freud referred to as the negative Oedipal position, a position developmentally prior to the positive Oedipal position? If it is positive, then the wish to replace mother and have a child from father is paramount. If it is negative, the anorexic, or bulimic identifies with the father and wishes to compete phallically with him for the possession and sexual control of mother (Lampl-de Groot, 1927 ; Deutsch, 1930, 1932, 1944, 1945 ; Freud, 1931 ; Brunswick, 1940 ; Nagera, 1975 ). Masserman gives an example of such a patient whose fantasies centred around the acquisition of a penis. She had her periods but was of a relatively low weight, and binged and vomited; she was probably a bulimic:
At this time the material also began to deal with the specific nature of her incorporative desires toward men, namely, to acquire their penises as a symbol of the masculinity desired by her mother and thereby eliminate them as competitors and displace them homosexually in her mother's affections (Masserman, 1941 , p. 330).
In Psychoanalytic Theories of Development: An Integration Tyson & Tyson (1990) suggest that not all girls go through the negative oedipal complex prior to a positive one as Freud thought. They have argued that if it occurs, it occurs after the positive stage and suggests disturbed object relations based on wanting to control mother and deny triangulation. This point is linked to perversion by Prophecy Coles.
I believe Freud was correct when he said some girls are unable to face the unpalatable fact that they have no penis. They continue to enact their "masculine life" and maintain their original relationship to their mother because their rage against her is too great. In this refusal to give up their "masculine life" we see a sexualisation of aggression that is both intense and compelling (Coles, 1989, p. 146).
This suggests there may be a refusal to acknowledge the presence and potency of father in relation to both self and mother. His separateness and his function as part of a copulating couple is to be denied.
To Be a Penis?
Bearing the presence of identification with a masculine ideal in mind, Abraham (1924) suggested that the body itself, by its very thinness, becomes identified with the penis. This idea was taken a step further by Lewin (1933) , who saw vomiting as being a partial identification with an ejaculating, urinating phallus. Sperling (1983) , Wilson (1983) and Sarnoff (1983) saw the flat stomach of the anorexic as being a retreat to a phallic ideal, which was a defence against the feminine wish to be made pregnant by father in phantasy and therefore having to compete with mother. Masserman (1941) and Leonard (1944) both described as central the conflicts of assuming a female role. Schwartz (1988) is one of the few authors who distinguishes between anorexics and normal weight bulimics. He makes the point that although normal weight bulimics have masculine identifications and ideals, they recognise that they do not have a penis, however much they may want one. He contrasts this to anorexics, who by their starving become a penis in phantasy so doing away with father, or the need for a man altogether. This suggests that the distinction is between wishing to compete with father in order to replace him in mother's bed and affections by having a penis, as Masserman (1941) thought, or omnipotently wishing to control mother by becoming a penis, in effect, eliminating all knowledge of father and the possibility of triangulation.
To Avoid Being Either a Man or a Woman
This confusion of gender identity and the possibility of simultaneously holding two positions is reminiscent of Aristophanes, who suggested that originally there were three sexes, not two, as Freud thought. There were not only men and women but also a man and a woman combined. In my clinical experience many bulimics and anorexics attempt to postpone indefinitely the realisation of which sex they belong to, as though it was a decision that could be made by choice alone. They attempt to identify with being both male and female and so imagine they can provide everything for themselves. A retreat into this particular form of narcissism can be seen in some patients, whereas in others an earlier form seems to be present, where the very knowledge of a separate existence from mother is not allowed to reach consciousness. I shall return to this later. As Harvey Schwartz has suggested:
Thus, the stereotyped ritual of gorging on food and forcing one's finger down the throat to induce regurgitation represents in part a simultaneous identification with both parents of the primal scene with an acting out on one's own body of the imagined role of the sadistic phallic father and castrated suffering mother. This defensive bisexual identification denies the humiliation of primal-scene exclusion, undoes "castration" and reverses passive (masochistic) oedipal impregnation wishes (1988, p. 40).
Alternating identification with one parent and the other has been reported by Karol (1980) in regard to asthma patients. Both Moulton (1942) and Sperling (1983) recognised the masturbatory aspect of the bulimic ritual.
Not Oedipal but Pre-Oedipal
This seems to suggest that what matters is being in charge, in control of both parents. No freedom or creativity between self and a parent, or between parents, is allowed. The problems of the unresolved pre-Oedipal relationship to the mother becomes explosive in relation to mother and father during adolescence. It is not just about having sexual intercourse, but is also about how anorexics and bulimics perceive of their role and others in relationships. The interest of many clinicians from Jessner and Abse (1960) to Palazzoli (1978) , Sprince (1984) , Boris (1984a, 1984b, 1988) , Mushatt (1992) and Wilson (1992) , moved from looking at anorexic and bulimic behaviour as representing repressed sexual and aggressive drives at an Oedipal level, to looking at the much earlier pre-Oedipal mother-child relationship. This paralleled a shift from drives to looking at what bulimic and anorexic symptoms were representing in terms of very early object relationships, where the mode of relating is based on control and survival (in Kleinian terms the powerful feelings of the paranoid-schizoid position).
The concrete nature of the thought of bulimics and anorexics and their omnipotent wish to control important objects in their lives is clear in some of their reported fantasies. Maguire noted, 'On the one hand, she wanted to come between them and have exclusive possession of each. She couldn't, she insisted, face the idea of her mother or myself being part of a sexual couple. "It makes me feel ill"'(1989, p. 118).
This is graphically illustrated by a male patient, who imagined being in his parents bed 'lying between them, with his father's penis in his rectum, and his own penis in his mother's rectum - a blissful state of complete union and sole possession of both parents' (Mushatt, 1992 , p. 304). I would disagree with Mushatt: far from blissful, this fantasy could be construed as hopeless, claustrophobic, immobilising and omnipotent - illustrating the perverse desire to deny separation, sexual difference and triangular relationships.
We are getting further and further from classical thoughts on the Oedipus Complex and nearer to a recognition that perhaps the difficulties are more accurately portrayed as being part of the mother-daughter relationship. The scenes above are not apparently about reproductive sexuality; they are about control, power and immobility, rather than creativity. Palazzoli wrote 'The sexual problem is not the basic one. All my women patients were fixated at the pre-genital levels' (1978, p. 76). The Kleinian understanding of part objects and Oedipal relationships creates a soup rather than a linear developmental path, with oral, anal, phallic and genital overlapping and interlocking in ever-fluctuating combinations.
The lack of differentiation between self and mother and vice-versa forms the basic building blocks for understanding both anorexia and bulimia. This is, of course, a description of Klein's (1946) narcissistic state. The focus has moved from oral impregnation fantasies to failures to achieve adequate separation-individuation from mother. This narcissistic difficulty means, in turn, that negotiating Oedipal issues - whether they be seen in a Freudian or Kleinian light - is going to be problematic for these patients.
Problems of Separation-Individuation
The eating disordered individual's relationship to her mother has been strangely absent from the story up to now. Penises and pregnancy phantasies, wishing to replace father or to have or become a penis have been examined. The child's earlier relationship to mother has been in the background, not the foreground, yet eating may not be a substitute for the penis, or for a baby: it may be about eating. Eating originally was about eating mother and not knowing where mother began and baby ended. To understand something of object relations we need to go back to the beginning where symbolic equations reigned unopposed: food = mother and mother = food. The concrete nature of this statement is where the trouble starts. Before a baby can become separate from mother he or she has to recognise that mother is other than herself or himself. Kleinian theory suggests that from birth onwards we have phantasies - unconscious ones, many-layered and multitudinous, stemming from our experiences of our bodies, and this colours how we understand and expect to relate in the real world. We begin by having extremely good and extremely bad phantasies about our mother's body, bits and pieces of it. In the babies mind these experiences are split into two, good experiences and bad experiences, or put another way, an excellent mother who meets our needs and a vicious angry mother who abandons us and does not meet our needs. In phantasy there are two separate people: one whom we have to protect ourselves from in order to survive, the other who provides us with all we need and could ever want.
According to Kleinian theory destructive impulses of our own are projected into mother in order to be rid of these painful and difficult feelings. Mother, in phantasy, then becomes furious and potentially retaliatory. The filter becomes an angry and destructive one. Waller et al., (1940) , Lorand (1943) and Hogan (1992) suggested that the Oedipal castration wishes were so strong, due to the intensity of the child's early oral struggles and the strength of their sadism and aggression. This can be understood as suggesting the proliferation of violent, destroying and destructive internal part-object relationships, at the level of unconscious phantasy. What is vital for successful development is the predominance of good experiences over bad ones, so that when it is realised that mother is but one person, a good and bad mother, good prevails internally. The emergence of concern for the other person surfaces as the good has to be protected from the bad. This is the depressive position. Survival as the aim is superseded. For this to occur, good experiences have to have been allowed and to have been successfully taken in the first place. Kleinians believe that we carry an internal picture of the world, full of part and whole objects to which we relate, and it is this picture, this unconscious filter, through which we see and experience the world and every situation with which we come in contact.
Anorexic behaviour has been understood as being about pathological narcissism, where not only is mother not recognised as being separate from the child, but she is thought to have nothing to offer. Primary envy is seen as so pernicious that it destroys all knowledge, awareness of a good breast, because knowledge of it would mean the recognition of something good outside of person her/himself, which would be intolerable. This is how anorexic and bulimic behaviour is understood by Boris (1984a, 1984b, 1988) . Anorexics become omnipotent in order to survive, as the knowledge of there being anything good which is not in their possession is unbearable. The breast is rubbished until there is little that could be got from it. Instead 'less' becomes 'more', to avoid wanting becomes the clue to existence. To have others want, to be devoid of desire is the aim: to be anorexic. In phantasy, 'no needs' means no separation, for being entirely self-sufficient prevents any awareness of dependency needs in relation to the self. If desire does not exist, mother unconsciously need not exist. The connection of both birth and early nurturing and dependence can be denied. Unconsciously, for many anorexics food still concretely equals mother. Food is the substitute for a longing for fusion, a longing for mother. If the desire was known, then the only result would be enslavement to mother and food. By starving it need never be known. Implicit in Boris's idea of an anorexic's internal world is the presence of a mother who enslaves, who does not want to be separate from her child and cannot bear to know about her neediness. I will return to this later.
Sohn sees it somewhat differently, implying that all connections are likely to fail because envy must not be experienced, yet paradoxically connections and links are continually sought but must never be found:
To my mind, there is a form of envy, which operates by promoting unawareness of the source of pleasure and of pleasure itself so as to be defended against an awareness of envy. The enjoyment giving object is then neither recognised properly nor separated from. Nor will the specificity of the object be noticed or acknowledged, which leads to further strange searchings for a possible repetition' (1985, p. 51).
What is being put forward is that anorexics and bulimics cannot separate from their mothers as their envy of the breast is too great, that to know of this envy would destroy them and their mother, and so in phantasy they possess, control or become mother. An excess of envy is so powerful in Kleinian thought that it prevents the necessary splitting into very good and very bad that is essential for early object relating and growth. Nothing good can be tolerated. There is, I think, something missing here. Unconscious envy may be an element in the difficulties anorexic and bulimic women experience, as may extreme destructive feelings, but by only looking through the filter of the child, the impact of the movements, feelings and actions of the mother seem to fade into relative insignificance. This is not to deny the existence of envy and profound destructive feelings in these patients. What we have to do in order to present a more balanced picture is to enter the arena of muddle between mother and child, where who is who is uncertain. To lay the blame on innate envy or on the mother is neither appropriate nor helpful. The complicated nature of the entwined relationship of mother and child is what needs to be looked at if work with these patients is to be productive.
The bewildering nature of apparently self-destructive omnipotence has already been noted amongst early religious martyrs. Can wilful starvation be understood? Normal weight bulimics rarely suffer from the severity of body image distortion which anorexics and anorexic bulimics do (Schwartz, 1988 ). They rarely die. Death is much lower down their unconscious agenda, although the incidence of conscious depression and suicidal inclinations is much higher in this group. Are anorexics actually trying to kill themselves? Charcot thought so in 1889 , as did Janet in 1929 and Lorand in 1943 . Janet believed it was a response to a rejection, whilst Charcot and Lorand clearly attributed it to something that was taking place in the mother-daughter interaction, as did Nandeau , some hundred years earlier in 1789. What it was that was deleterious in the mother-daughter relationship was not explored. Palazzoli (1978) thought that anorexics identified their physical selves with the bad mother, who then threatened to destroy or overwhelm them from within when they ate. Starvation was for them a way of attempting to avoid, deny and control the bad mother. Yet because of the paranoid-schizoid state they were in, they forgot - were blind to - the impossibility of starving mother without starving themselves. Joyce McDougall , in her book Theatres of the Body speaks of a patient with severe psychosomatic symptoms who believed that by physically damaging her body she could concretely disprove the belief that her body was her mother's body, and if she separated from it she would die (1989 p. 141). Cecil Mushatt sees it differently:
There are instances when the rejection of food expresses the fantasy of omnipotence and invulnerability because of the unconscious fantasy of complete union with mother. Because of this symbolic fantasy, there is no need to eat. To eat food is to acknowledge the fact of separation and of one's mortal being (Mushatt, 1992 , p. 309).
Death simply does not exist. Yet it does. This is an alternative understanding of no needs meaning no separation, as described in Harold Boris's 1984 papers, referred to earlier.
Palazzoli and Bruch described anorexia as being a particular psychosis, mid-way between the paranoid-schizoid and the depressive positions. This is now taken up by the concept of the pathological organisation (Steiner, 1993 ). I do not think all anorexics and anorexic bulimics are in the thrall of a pathological organisation, but some certainly are. Ellen West , (Binswanger, 1944) seems from her diaries and letters to have inhabited such a position, to the degree where her idealisation of death, aided and abetted by an internal gang (Rosenfeld, 1971 ) overran and overpowered her wish for life until she felt she had no option but to commit suicide.
For some anorexics there seem to be only two options, - madness or death. Fenichel (1945) recognised that anorexia could be a defence against psychosis. Eglé Laufer implies that questioning a delusional belief system which operates around the body may lead to a psychotic breakdown.
Any demand on the psyche for change of the body image... can become a threat to the person's ability to relate to the external world in an undistorted form, resulting in a psychotic breakdown (Laufer, 1991 , p. 68).
Eating disorder case histories frequently include Christian imagery of attaining the status of pure spirit by shedding the shackles of the body (e.g., Hogan, 1983 ; Wilson, 1983 ; Wooley and Wooley, 1985 ). An extraordinarily powerful wish for, and sense of, detachment from the physical realm contributes to the common delusion among severely anorexic individuals that radical weight loss will not bring death: The body will die, but not the core self (Cross, 1993 , p. 60).
Many anorexics feel superior and powerful because of their ability to deny their own needs, and so assume they are fundamentally different from other individuals. Omnipotence is about power and control and Palazzoli adds the effect of impact on the 'other' to the anorexics fantasy of immortality.
To begin with, she is prey to a most disastrous Cartesian dichotomy: she believes that her mind transcends her body and that it grants her unlimited power over her own behaviour and that of others (Palazzoli, 1978 , p. 223).
I shall look at this inability to know people as people in the next chapter, whilst not denying the validity of the above approaches. The above examples are taken from individual work with patients where the understanding comes out of the work itself. What I think may be a general trend is the tendency to focus on the narcissistic elements of the anorexic's or bulimic's behaviour. This, in turn, leads to a collusive focus between patient and therapist away from object relating as such. It is this lack which I wish to redress. In order to do this, I want to look at mother, her unconscious and its impact on her child.
Did I Have a Good Enough Childhood?
Babies do have to be cared for physically and emotionally. There is a reality of experience which these women have undergone and which is I think not given enough importance by Boris or Sohn . Mothering is not always good enough. Bruch stressed the real deprivations and deficits which the future anorexic suffered as a child. She led the way forward for the self psychologists such as Goodsitt (1983, 1985) and Swift and Letven (1984) who took her ideas about the failure of these children to recognise and interpret internal signals and external stimuli one step further. They saw over-stimulation as triggering off bulimic and anorexic episodes which prevented anxieties of an annihilatory nature coming to the fore. This theme was also taken up by Krueger (1988) who saw the excessive over activity of anorexics as a way of holding their physical and psychical selves together. Bruch saw this failure as belonging to the parents; the child had not been shown how to know, name, or recognise her internal states. Bruch's work with these patients remained firmly in the conscious realm, but she was perhaps one of the first practitioners to recognise the terror beneath the facade of confidence and insouciance which many anorexics, in particular, present. These models have important implications for the technique of working with this group of patients - of the intricate attention which has to be given to the awareness and naming of patient's internal states - as far as this is possible. This recognition of the presence of powerful psychotic anxieties in eating disorder patients in general, is vital, as is the recognition of the failure of effective parenting.
In the next chapter I wish to focus more on the mother's unconscious phantasies and conscious fantasies to show how they may impact on her daughter, using Winnicott 's ideas of transitional objects and phenomena. They are connected. A number of possibilities about the meanings of anorexia and bulimia have been proposed. Questions remain. What is it about the nature of the mother-child relationship that has meant separation is so hard? Is it, as has been suggested by Freud and others, a question of the an unwillingness to compete with mother, to face the Oedipal situation in a straightforward way? Is it to do with the presence of intense envy in the patient, which prevents separation from mother, for separation would mean having to experience the envy which in phantasy only leads to destruction?
Hilde Bruch' s extensive experience with anorexic patients, and the awareness of the failure of ability to regulate affect are, I think, cornerstones to working with all groups of eating disordered patients. Her clinical experience is closest to my own, yet the annihilatory fears and presence of powerful destructive phantasies seem better fitted to a modern Kleinian technique where the deepest anxieties are interpreted and explored from moment to moment within the session. Both Bruch and her followers, I think, fail in practice to address the power of the negative forces and how they limit and indeed sometimes put these patients' lives at risk.
An alternative description would be to use the model of pathological narcissism, which seems only a small step removed from the idealisation of the death instinct and pathological organisations. Lorand (1943) noted the unconscious death wish fantasies of the mother towards the child and finally added that the whole problem centred around the fact that his patient had been unwanted. I want to take Bruch 's work on the mother, from the arena of the conscious into the unconscious. I hope that it is here that the missing piece of the puzzle may be found. If it is, it should expand our understanding of how psychotherapists work with these patients and give a wider canvas from which to draw individual understandings. I now want to turn to the mother, her pathology, and its possible effect on her foetus, her baby, her child and the eating disordered patient.