Treating Patients with Symptoms -- and Symptoms with Patience:
Reflections on Shame, Dissociation, and Eating Disorders

Philip M. Bromberg, Ph.D.

I offer the view that the symptom picture found in most patients with eating disorders, as well as in the symptomatology of many other socalled difficult patients, is the end result of prolonged necessity in infancy to control traumatic dysregulation of affect. I propose that the central issue for an eating-disordered patient is that she is at the mercy of her own physiologic and affective states because she lacks an experience of human relatedness and its potential for reparation that mediates self-regulation. She is enslaved by her felt inability to contain desire as a regulatable affect and is thus unable to hold desire long enough to make choices without the loss of the thing not chosen leading to a dread of self-annihilation. Trauma compromises trust in the reparability of relationship, and for symptoms to be surrendered, trust in reparability must be simultaneously restored. Because felt desire is the mortal enemy of an eating-disordered patient, this fact becomes a central dynamic in the analytic field, leading analyst and patient into a struggle over who shall hold the desire and whether the issue of control over food is allowed to become a subject for negotiation. I discuss the inevitability of the analyst's own dissociative reactions in response to the patient's internal war over desire and control, and the different types of interpersonal enactments into which an analyst is drawn. In this tension, as illustrated through clinical vignettes, analyst and patient slip in and out of a constantly shifting array of self- states and thereby have an opportunity to coconstruct a transitional reality within which the patient's impaired faith in the reliability of human relatedness can be restored, and eating can become linked to appetite rather than to self-protection.

Virginia Woolf (1928), with characteristically understated perceptiveness, casually observed in her novel Orlando that "these selves of which we are built up, one on top of another, as plates are piled on a waiter's hand, have ... little constitutions and rights of their own":

One will only come if it is raining, another [will emerge only] in a room with green curtains, another when Mrs. Jones is not there, another if you can promise it a glass of wine-and so on... . [E]verybody can multiply from his own experience the different terms which his different selves have made with him-and some are too wildly ridiculous to be mentioned in print at all [pp. 308-309].

I've begun with this passage because rarely has the wonderous nature of the self been portrayed with such matter-of-fact honesty and simplicity. Woolf's evocative description is easily embraced by the reader with feelings of pleasurable recognition, and even her use of the phrase "different selves" has a felt rightness to it that is accepted without resistance. Perhaps her language, so personal and down-toearth, allows her such a warmly congenial relationship with the reader that it helps overcome the potential for discontent that sometimes can be evoked by the notion that we each comprise different selves. As Popeye put it, flexing his muscles belligerently, "I yam what I yam and that's all that I yam," and any idea that suggests otherwise, whether offered by an analyst to a patient or by a writer to a reader, feels "right" only if presented in a manner that simultaneously supports one's basic experience of self-continuity. It is when an analyst fails to provide such support that he is most apt to experience his patient as "difficult."

If we were to place our "difficult" patients into one group and our so- called good patients (more and more difficult to find these days) into a second group, there is a one particular characteristic of most patients in the first group that transcends the individual personality traits that we use in making differential diagnoses between the members of this group. Put simply, the thing our "difficult" patients most require from their therapists for growth to occur is, paradoxically, the same thing that is most difficult for the therapist to provide-interpersonal engagement that combines affective honesty and safety. Although it is certainly true that this same combination is needed by patients in the second group, with that group its provision is part of the natural give-and-take of the relationship and is often unnoticeable except as "background music."For patients in the "difficult" group, this natural give-and-take is minimal, sometimes totally absent. Most of them, with good reason, have come to mistrust signs of genuine relatedness from another person as though these signs were "really" omens of potential betrayal. Such patients are difficult, and they feel difficult to a therapist because they deprive a therapist of what he most counts on in order to sustain hope-a working relationship that will grow in depth and security as the work progresses. With these individuals, such a relationship does not as a rule exist, at least for a long time, because their mental structure has been shaped too extensively or for too long by the effects of trauma and dissociation. Their capacity to trust a human relationship must first be slowly restored-in some cases, built for the first time and, without this taking place, any attempt at psychoanalysis results in a "pseudoanalysis."If such a person is ever to truly feel recognized within a relationship, the therapist must comprehend that, when his patient behaves as if the only self that she feels is "really" her is the one that s there at the moment, his patient is not being simply capricious, inconsistent, or, even worse, resistant. The therapist's task is to allow himself to slowly discover all her selves or self- states and to form relationships with each, allowing that each holds a different version of "truth" and its own agenda for treatment. p. 894The title of this paper, "Treating Patients with Symptoms-and Symptoms with Patience," is one I used (Bromberg, 1995b) at a conference celebrating the centennial of Freud's publication of Studies on Hysteria (Breuer and Freud, 1893-1895). In that paper, I discussed Freud's treatment of his famous patient Frau Emmy von N from a vantage point 100 years later. Freud's refreshingly honest assessment was that "the therapeutic success on the whole was considerable; but it was not a lasting one" and that "the patient's tendency to fall ill in a similar way under the impact of fresh traumas was not got rid of (Breuer and Freud, pp. 101-102).

From my perspective (Bromberg, 1996), the symptoms that Freud called hysteric are more usefully seen as dissociative, and the reason Emmy was still susceptible to falling ill was that her need to maintain the dissociative structure of her mind was kept alive by the present, not by the past, and thus remained untouched. Like any trauma survivor, she protected herself against the future by treating the present as if it were nothing but a replica of the past. In this context, her "illness" was not just her symptoms, but the plundering of her life-a foreclosure of the "here and now" on behalf of the "there and then," effectively preventing her from living life with spontaneity, pleasure, or immediacy. In other words, to understand why Emmy's cure was not a lasting one is also to understand that we do not treat patients such as Emmy to cure them of something that was done to them in the past; rather, we are trying to cure them of what they still do to themselves (and to others) in order to cope with what was done to them in the past. And so, I concluded, our success as clinicians in treating patients with symptoms lies, at least in part, in our ability and willingness also to treat symptoms with patience.I believe that the same point can be made equally validly with regard to most symptoms, including those associated with eating (which, by the way, Emmy had to no small degree). That is, they represent the basic adaptational function of dissociation in foreclosing the possibility of holding in a single state of consciousness two incompatible modes of relating. As Freud put it, "it is impossible to eat with disgust and pleasure at the same time" (Breuer and Freud, 1893-1895, p. 89). In its most general form, the person is protected from the felt impossibility of responding self-reflectively with feelings of fear and security toward the same object at the same moment. On the basis of my own clinical observations and those of others (McCallum et al., 1992; Gleaves, May, and Eberenz, 1996; Katz and Gleavesr 1996), 1 offer the view p. 895that most of the symptoms associated with eating disorders can be most usefully understood as an intrinsic outcome of dissociation. The ability to act from a cohesive sense of self-agency is not destroyed but is "repackaged" in unlinked states of mind, leading to a personality dynamic whereby certain self-states are dramatically purposeful-often stubbornly intractable-whereas others seem inhibited but are simultaneously on alert.

A temporary cognitive narrowing is experienced during a binge as the bulimic refocuses attention on to the immediate stimulus. This refocusing allows a reduction in negative affect or a general reduction in self-awareness. . . . This pattern of "escaping from awareness" appears to involve cognitive processes similar to those that underpin the concept of dissociation [Everill, Waller, and Macdonald, 1995, p. 155).

But Freud never followed through on his early insights, and in his war with Janet¹ ended up minimizing the significance of trauma and dissociation in human mental functioning. In this context, Havens (1973), in a wonderful book, Approaches to the Mind, had some things to say about Janet and Freud that are highly relevant:

It was obvious to both Janet and Freud that the conscious ideas of the patients did not encompass the phenomenon of hysteria. Freud then searched, first by hypnosis and later by the method of free associations, for unconscious ideas and was led forward to the idea of unconscious yearnings, attitudes, convictions, and expectations. Janet searched for what besides ideas was dissociated, and in what ways. He left behind the old conception of single ideas, resulting from trauma, and splitting off from mental life, for that of dissociated functions or systems within which many sensations, acts, fears, and ideas were included-separate, organized centers of attention, receiving impressions and able to be communicated with; in control of the personality (as in the somnambulisms and fugue states) or capturing a leg, arm, or the function of eating. . . . Each hysterical function had its own consciousness, organizing principles, and capacities for communication [pp. 59-60].

With regard to patients suffering from eating disorders, this issue is particularly salient, as was Janet's contribution. For the anorectic, wrote Janet (1907), the act of eating is an amnesia, "a somnambulistic phenomenon which can only be effected in the ... somnambulistic state ... [and] is lost to the normal and waking consciousness" (p. 243). The most powerful and enduring significance of Janet's work in this area, Havens (1973) asserted, is not simply in recognizing the centrality of dissociation, but in demonstrating that dissociation involves an organized system of self-experience rather than just a single idea or sensation, and that in the case of anorexic illness this cutoff system has captured and redefined the meaning and function of eating (pp. 49-50). It is a hypnoidally isolated complex of physiological events, fears, movements, sensations, and ideas that work together as a separate center of attention that is able to be communicated with, and is in control of the total personality when it needs to be.

What Janet called a "system or complex" is what I see as a dissociated self- state, or a self organized by its own dominant affect, its own view of social phenomena and human relationships, its own moral code, its own view of reality that is fiercely held as a truth, and, with patients suffering from eating disorders, its own relationship to food and to the body.In the contemporary analytic community, Janet's ideas have begun to take hold only during the past 10 to 15 years, influenced mainly by hard data provided from outside psychoanalysis. At this point in time, however, an increasing number of analytic clinicians, researchers, and theoreticians are arguing that, and presenting evidence that, the human personality begins and continues as a multiplicity of selves or self-states, each with its own dominant affect and sets of characteristics that are always shifting in configuration and moment-to-moment availability to one another. This is not, however, the same as saying that everyone is really a "multiple personality" or, as we now term it, a "dissociative identity disorder." Why not? Because one's different states are subjectively linked together by the developmentally necessary illusion of being "one self," and, if all goes relatively well early in life, one's own self-state shifts are normally as unobservable as the beating of one's heart, and self-continuity goes on without disruption. Most of us can access a broad range of self-states that can participate in even the most emotionally complex and psychologically conflicted situations. But not all are so lucky. In the face of psychological trauma, self-continuity is threatened, and this threat, for most human beings, is countered by the use of dissociation as an evolutionary survival response that is equivalent in survival priority to certain genetically coded response patterns of lower animals to a life-threatening attack by a predator.

For instance, there is a strange-looking sea creature, the holothurian, that is best known for its disconcerting ability, when attacked, to divide itself into unlinked parts and regenerate from that which escapes death. "Non omnis moriar--I shall not wholly die!" Consider, if you will, the following brief excerpt from a poem by a woman awarded the 1996 Nobel prize for literature when she was in her late 70s, the Polish poet Wislawa Szymborska.² The poem's title, "Autotomy," is the biological term for this capacity of certain living things to give up wholeness in order to preserve life:

In danger the holothurian splits itself in two: it
offers one self to be devoured by the world and
in its second self escapes.

In the middle of the holothurian's body a chasm opens
and its edges immediately become alien to each other.

On the one edge, death, on the other, life.
Here despair, there, hope.

To die as much as necessary, without overstepping
the bounds.
To grow again from a salvaged remnant.

Here a heavy heart, there non omnis moriar,
three little words only, like three little plumes ascending.
[Szymborska, 1983. pp. 115-116]

I shall not wholly die! With these words in mind, now contemplate a piece of advice offered more than 140 years ago by George MacDonald (1858), the Scottish clergyman and author whom C. S. Lewis acknowledged as the primary source from which his own capacity to bridge fantasy and reality was shaped. MacDonald wrote that "the best way to manage some kinds of painful thoughts is to dare them to do their worst, to let them lie and gnaw at your heart till they are tired, and you find you still have a residue of life they cannot kill" (p. 55). MacDonald's words sometimes come to me when hope seems far away with certain patients. But all too often it's just too damned difficult to do what MacDonald proposed. It's difficult even for those of us who haven't been seriously traumatized as children. But for those who have, trying to hold painful thoughts, letting them "gnaw at your heart till they are tired," and surviving without dissociating is an outright impossibility.

Laurie, age 26, had been obese as a child and became bulimic during adolescence, at which point she shed most of her weight and for a brief period became anorectic. At the point I first met her, she appeared slightly underweight but not anorectically thin. Her older sister, who had also been obese as a child, remained that way, never having developed a bulimic solution. She became for Laurie the apotheosis of greed-an object of contempt and a constant reminder to Laurie of who she was not. Laurie entered treatment with a variety of symptoms along with the eating disorder. Some were classically dissociative, such as fugue states manifested in such things as her history of "forgotten" appointments, not going to class, not coming to work-all without awareness. She also suffered from flashbacks, traumatic nightmares, and a body experience sometimes blurred and always unstable. She frequently started sessions by asking me either "What do I look like today?" or "Do you think I look different from last time?" She was also prone to sudden state-shifts during sessions, preceded by attentional disturbances in which she seemed to disappear from whatever interaction was taking place between us. She also presented with other symptoms, less classically dissociative, but often found to accompany the former. The major ones were migraine headaches so severe that they could virtually incapacitate her, and a compulsive "hair- twirling--leading-to-hair-pulling" habit that started at age 14, when she also developed amenorrhea, which lasted until she left home at age 20. The bingeing and hair pulling had both been described by her as reducing tension because they "make her mind a blank."

On this particular day, Laurie began her session, as she frequently did, sitting silently and staring, trancelike, into space. She then began to talk blandly and without affect about having pulled out her hair the night before. "I discovered I haven't stopped. You thought I had." As I listened to her, I became aware of my own feelings-an odd blend of apathy and irritation. I might have ignored it and just launched into another dead-end inquiry about the details of what had led to the hair pulling, except that I noticed her sitting back contentedly, like someone settling in to watch a movie. I commented on this, and she responded by telling me the dream she had the night before-a dream in which she was at her sister's birthday party, which was taking place in an insane asylum. "There was a big gooey birthday cake-my sister loves cake; I hate it-and she's getting mad at me because I don't want any. I finally start yelling at her to 'shove it.' I woke up just as I had my face up against hers, screaming, 'I don't have that body, it's yours, not mine, you pig!'"

I asked Laurie what thoughts she had about the dream, and she said in an offhand manner, "None! I don't like to come up with ideas. I'd rather wait for you to have some. That way I can either agree or disagree, and I don't have to risk being wrong." There was a look in her eye as she spoke that made me feel she was trying to pull me into a battle. But, as was not the case earlier in the session, I could vaguely sense she was conscious of what was going on between us at least to some extent, and, contrary to feeling apathetic and irritated earlier, I was not put off by her manner this time. In fact, I was feeling playful, and I found myself saying to her:

Actually, what you usually do is agree and disagree at the same time, so I never can really be sure what you feel. It's like I slave over a hot couch all week. I cook you my best interpretations. Do you eat them? No! You taste a piece here, a piece there. I throw out three quarters of what I cook. There are starving patients in Europe who would be grateful for what I throw away that you don't eat.

Laurie, though not a "couch" patient, began to laugh, and I could feel the atmosphere shift-lending some support to Lewis's (1992) observation that laughter is a mechanism by which shame can be reduced or eliminated. "Laughter," Lewis wrote, "especially laughter around one's transgression as it occurs in a social context, provides the opportunity for the transgressing person to join others in viewing the self. In this way, the self metaphorically moves from the site of the shame to the site of observing the shame with the other" (p. 130). In my own preferred idiom-as I don't see the shift as metaphorical laughter allows more of one's selves to get into the act. A part of Laurie that was lively, animated, and almost enthusiastic was now clearly participating.

I said to Laurie that at this stage of our work, the odds of each of us being wrong were pretty high because there's so much we don't know yet. But the odds get better when we compare notes about what each of us is feeling about what we are doing. It reduces the amount of guessing. After a moment of silence, she replied:

I think the insane asylum was your office. Sometimes I feel you want to make me fat like my sister. All these interesting things you say to me make me afraid to hear more. Yeah, I guess I was sitting back like I was getting ready to take in a great movielike a great meal I could have without getting fat because I didn't even know I was eating it. I think I know what made me start pulling my hair out again.

She then described an event that took place two days before this session. "I was walking to a restaurant with my father, and he holds my hand in this weird way-he won't let go. I had to pretend I was fixing my hair to remove my hand." This was her first concrete association that could potentially shed some light on her hair twirling/ pulling. Then, suddenly, in the voice- of a preadolescent child, she said, "He never touches Mommy that way. I wonder sometimes if, people who see us think I shrunk."

"Who would they see?" I ask her.

Now back in the other voice, "They'd see a 10-year-old girl walking with her father. My husband lives with her most of the time. Most of the time he likes it. But he doesn't like it when I change. He s says, `Why do you have to be different people on different days?'"

"Well, I'm glad I had a chance to meet her," I reply, "even if it was only very briefly."

"Yes," she said pointedly. "She went away again as soon as you asked me to tell you about her." A bit dazed as the end of the session approached, I mumbled what I hoped would be a supportive response and a "good" note on which to stop: ""Maybe if I talk to her more directly, she will stay longer. Do you think so?"

Naively anticipating that this would be our "marker" for next time, I was shocked when she ignored what I thought were my obvious cues that the time was up, and she began a monologue that seemed nonstop. She began to talk about how afraid she is of offending people at whom she secretly scoffs-people who think that what they say matters to her when it really doesn't. "It's so strange," she went on, "even though some people don't seem to get angry about it. I crash when it's over. I'll go home and binge and vomit."

I felt paralyzed at this point. Even though I felt she was talking about us and about what it feels like for her after she leaves sessions, I also wanted to end the session. I kept hoping that if I gave her just a little more time I would think of the "right" thing to say, so I let her go on ... and on . . . and on . . . hoping against hope that if I didn't stop her before she stopped herself, it would end in the "right" way. Well, I finally abandoned that hope and stopped her 10 minutes past the end of the session.

She said, sounding a bit miffed, "I haven't finished yet." I replied, probably sounding a bit contrite, "I know, but our time is up for today, and I'd love for us to have a moment to reflect on what just happened here." She retorted, "I never reflect on what I'm saying when I'm like this." I answered, now a bit more composed, "But later when you get home, a different part of you does think about it-all alone. And then you decide you were horrible and end up bingeing and vomiting." At that moment, something "clicked into place" for me, and I added, "But in a funny way, you only did what I asked you to do, didn't you? Remember when I said, `Maybe if I talk to her more directly, she'll stay longer'? I just didn't expect it to happen now! But why wouldn't it?" And then, shifting realities, "After all, you were just being you. 'Longer' means until you've finished, right?" She, stood up, grinned, and left. I breathed a grateful sigh of relief.

Psychosomatic dissociation can be seen as the core of eating disorders. It results in the separation of the mind from the body, with the absolute hegemony of the mind and_ the negation and abandonment of the body in anorexia, and in bulimia with either a continuous struggle between the mind and the body, with their alternately prevailing cycles, or eventually a consolidated withdrawal so that the mind is no longer able to control the body or keep in touch with it [Ciocca, 1998, p. 52].

In the context of the preceding vignette, Ciocca's statement underlines the intricate relationship among trauma, dissociation, damaged capacity for affect regulation, and the need for an analytic relationship in which growth and repair are inseparable components. As we know, successful interpersonal transactions between infant and caregiver mediate, at the brain level, the capacity for affect-regulation and secure human-relatedness. One could, if one wished, translate this into the language of attachment theory, whereby the phrase "a secure attachment bond" could substitute for the phrase "secure human relatedness." When these early patterns are relatively nondisruptive and, most important, reparable, they create a stable foundation for internal affect-regulation that is largely nonverbal and unconscious.

Tronick and Weinberg (1997), in a seminal paper on research into affective regulation and the "architecture of mother-infant interactions," emphasized how disruptions in the mutual regulatory process create a break in the development of intersubjectivity. In the face of chronic failure to repair the interactions, the infant is unable to achieve social connectedness and develops dysregulated affective states that it is unable to control. An early coping style begins to develop in which most of the infant's activity is enlisted into stabilizing out-of-control affect. Most significant is Tronick and Weinberg's conclusion that

reparation of interactive errors is the critical process of normal interactions that is related to developmental outcome rather than synchrony or positive affect per se. That is, reparation, its experience and extent, is the "social-interactive mechanism" that affects the infant's development . . . [because] the infant develops a representation of him- or herself as effective, of his or her interactions as positive and reparable, and of the caretaker as reliable and trustworthy [pp. 65-66, italics added].

Much adult psychopathology may thus be the end result of prolonged necessity in infancy to control physiologic and affective states while lacking an experience of human relatedness and its potential for reparation that mediates it. An early foundation of reparability in relationships allows for further successful negotiation of interpersonal transactions at increasingly higher levels of adult self-development and interpersonal maturity, affecting not only the richness of one's life but determining the difficulty an individual will experience in attempting to negotiate and use a psychotherapeutic relationship. Trauma creates the experience of nonreparability, and, in those areas in which trauma has occurred, the experience of trust in the continuity of human relatedness must be restored.

With regard to those individuals for whom this state of affairs leads to an eating disorder, Boris (1984, 1986) offered the view that eating disorders arise when the dysregulation of desire is linked in infancy with the dysregulation of appetite. This leads to what Boris called an "unevolved state of mind," in which one wishes and hopes to have everything all the time-a state of mind commonly known as, greed. Greed is a state that attempts to eliminate the potential for traumatic rupture in human relatedness by replacing relationship with food-a solution that is largely self-contained and thus not subject to betrayal by the "other." But it is by no means a perfect solution. The particular problem with greed is that its presence is inevitably tarnished by the existence of choice and the shadowy pressure of the need to make one. The realization of the need for choice, Boris wrote, either stimulates a refusal to endure it, leading to the decay of appetite back into greed and an experience of vast frustration, or stimulates the making of the choice, leading to the satisfaction of appetite but always accompanied by the feeling of profound loss of the thing not chosen. In Boris's (1986) words, "Appetite . . . makes manifest the infant's first encounter with actuality and, as such, makes actual experience for the first time a player in the process. The quality of the appetitive experience will now play a role in whether the feeling of loss is modulated by compensatory and consoling experience-or is not" (pp. 48-49).

Boris (1986) was saying, in other words, that the essence of the human condition is having to recognize one's insufficiency and that the degree to which one draws satisfaction from human relatedness will keep him from seeking nonhuman solutions (e.g., food) as a means of compensating for the experience of loss. For some people, because the recognition of insufficiency is unbearable, choice becomes unbearable, and in the infant who later in life develops an eating disorder, the capacity to make a choice is impaired because the experience of loss connected to appetite is not modulated by the compensation and consolation of human relatedness. That is, what in adulthood could develop into appetite and healthy, regulatable desire, . because it is denied the relational context on which that transformation depends, freezes the experience of being an affectively out- of-control infant within a dissociated self-state that takes on an imperious life of its own. The person who eventually shows up at the therapist's office, no matter how she chooses to define her "presenting problem," is someone whose real problem is that she is totally at the mercy of her own feelings-someone who is enslaved by her felt inability to contain desire as a regulatable affect. Boris (1984) wrote, "By the simple expedient of declaring `less is more,' greed for the breast is metamorphosed into a gluttony for punishment, yearning into abstinence, retention into elimination (in bulimia) via each and every alimentary orifice, indeed, by exercise and sweating, through the very pores themselves" (p. 317). The renunciation of desire is what we see as the hallmark of anorexia and, in a different way, of bulimia. But at its core, it is a loss of faith in the reliability of human relatedness.

I have found that, in patients with eating disorders, the transmutation of desire into renunciation is most frequently mediated through the mechanism of dissociation. Sands (1994) suggested that "dissociative defenses serve to regulate relatedness to others... . The dissociative patient is attempting to stay enough in relationship with the human environment to survive the present while, at the same time, keeping the needs for more intimate relatedness sequestered but alive" (p. 149). In other words, dissociative defenses are not designed simply as an impermeable suit of armor. No matter how walled off the patient may be from intimate contact with others, the broadest purpose of a dissociative mental structure, including its place in most eating disorders, is not just insularity but regulation. It is above all else a dynamic mental organization designed for affective self-regulation--a mental structure tailored to anticipate trauma but sufficiently permeable to be a potential doorway to therapeutic growth. Its insularity reflects the necessity to remain ready for danger at all times so it can never-as with the original traumatic experiences arrive unanticipated. Its permeability reflects a capacity for authentic but highly regulated exchange with the outside world and similarly regulated spontaneity of self-experience (Bromberg, 1995a, pp. 194195). Let me put it in the words of a patient who did not have an eating disorder:

When I was little and I got scared-scared because Mommy was going to beat me up-I'd stare at a crack in the ceiling or a spider web on a pane of glass, and pretty soon I'd go into this place where everything was kind of foggy and far away, and I was far away too, and safe. At first, I had to stare real hard to get to this safe place. But then one day Mommy was really hitting on me, and without even trying I was there, and I wasn't afraid of her. I knew she was punching me, and I could hear her calling me names, but it didn't hurt and I didn't care. After that, anytime I was scared, I'd suddenly find myself there, out of danger and peaceful. I've never told anybody about it, not even Daddy. I was afraid to because I was afraid that if other people knew about it, the place might go away, and I wouldn't be able to get there when I really needed to.

As treatment progresses, the patient will often reveal the existence of an inner life dominated by a never-ending war between parts of the self, each denouncing the other around the issue of appetite and desire-a war that more often than not manifests itself experientially through the presence of internal voices, often sadistic and unrelenting, that the patient desperately needs to still by finding ways to give each some of what it wants. And the war between self-states never ends, because as Boris wrote, desire is undeniable -and durable, and for someone with an eating disorder it means that everything else is obliterated.

How does this get expressed in treatment, and- how does it relate to dissociation? Zerbe (1993) wrote, "Perhaps one test of the integration of self comes from how consciously we hold varying perspectives of ourselves and other people in full awareness from moment to moment. The more easily we do this, the better we seem to know ourselves and the more we experience self-cohesion" (p. 320). Zerbe continued, "Given the frequency with which dissociative states and eating disorders may occur together, it behooves clinicians to keep an open eye to making both diagnoses if one or the other is found" (p. 321).

In bulimia, for instance, bingeing-and-purging as a cohesive act is by and large done in a dissociated (not-me) state. Muller (1996) wrote that such patients "are attempting to set a marker at the edge of the self so that they can experience a limit and not become fragmented in a diffuse unnamable scatter" (p. 85). This purpose could not be served if the bulimic were fully conscious because it would be a self-experience that was being done by me to me and thereby would be useless in its ability to ward off autonomic hyperarousal of affect. The trance state, through dissociation, allows this "marker" to be set between areas of self that trauma has made incompatible. But this protection comes at a huge price because it forces the self to severely limit what can be experienced as "me" at any given moment. How are we to account for a solution that can be so fiercely self-destructive? I don't pretend to have an answer to this, but there is something I find convincing in a comment made more than 30 years ago. Erikson (1968) wrote of such individuals as creating a "self-decreed moratorium, during which they often starve themselves, socially, erotically, and last but not least, nutritionally, in order to let the grosser weeds die out and make room for their inner garden. But often when the weeds are dead, so is the garden" (pp. 41-42, quoted in Muller, 1996, p. 81). And even more tragically, as we are all too painfully aware, so, sometimes, is the person herself. As Ciocca (1998) has put it, we must find the reasons for

her intolerance of a ... conflict related to ... being herself. The aim of every therapeutic encounter ... is to lead her to a meeting with herself, as that which she is, and that which she could become.... [Foremost] in therapy is her impossibility to deal with the dissociation itself. In such a situation it is useful to highlight the way in which her mind works, and how it influences the way in which she is living [p. 54].

If a therapist is to work with a patient's here and now experience of safety by negotiation within the enactment-while he is highlighting the way her mind works-it is essential that the therapist be especially attuned to his patient's potential for dissociated hyperarousal of affect created by the relationship itself, addressing each patient's ongoing experience of emotional safety as an intrinsic and inseparable part of the analytic interchange. For those patients in whom the effect of trauma on the organization of psychic structure has been most pervasive, the self- reflective ability to work in the here and now is least likely to be present at the beginning. These patients tend to use each session to process nonprocessible experience that has occurred in prior sessions. In other words, each session becomes for at least one part of the self a kind of commentary (through derivatives, dreams, and enactments) on the preceding session or sessions. The therapy proceeds that way, with the therapist's job being to try to enable the processing to be safer and safer so that the person's tolerance for potential flooding of affect goes up-that is, her threshold for triggering increases. This allows her to increasingly hold onto the ongoing relational experience (the full complexity of the here and now with her therapist) as it is happening, with less and less need to dissociate, and, as the processing of the here and now becomes more and more immediate, it thereby becomes more and more experientially connectable to her past.

In any treatment of a patient for whom trauma is an issue (and, as I said, I believe this encompasses every patient in at least some areas of his or her personality), a patient will attempt to talk about the trauma, always with hidden shame, because "talking about" evokes "reliving." The patient, however, will usually reveal enough data to stimulate the therapist's curiosity, setting off a process in which the therapist's accelerated attempt at exploring events for which the patient has no narrative memory leads to an enacted reliving of the trauma, into which the therapist is drawn like a moth to a flame. As the enactment continues, the patient's dissociated shame escalates, and the therapist finds himself feeling things about his patient and about his own role that make him increasingly uncomfortable, often triggering his own dissociative processes. I believe this collision of subjectivities to be an inevitability, not a sign of a technical error. It draws the work into a dialectic between the here and now and the there and then, allowing the mutual construction of a transitional reality in which both the patient's and the analyst's dissociated experience have an opportunity to coexist as a perceived event different enough from the patient's narrative "truth" about relationships to permit internal repair to take place and the patient's reliance on dissociation to be gradually surrendered. To be fully in the moment is to be fully allowing new (as yet unprocessed) experience to interface perceptually with episodic memory, thus optimizing its potential for integration into narrative memory and, ultimately, enriching self-narrative--the goal of any form of treatment. We are at those moments, in effect, "standing in the spaces." Dissociation short-circuits this possibility.

But the term eating disorder can become a real handicap to therapy if it is embraced unreflectively by the therapist as simply a handicap to the patient. The therapist must simultaneously recognize and respect as an achievement the means by which a patient has constructed her eating "disorder" so as to preserve its dissociative structure and thereby give each part of self some of what it wants without unbearable conflict.

In an earlier paper (Bromberg, 1994) 1 described a case being presented to me in ongoing consultation (pp. 247-250). The patient was a woman who had been in treatment for quite a few years and who had made major changes in her life and her self-experience, except with regard to the thing that had brought her into treatment in the first place, her obesity. The analyst, having had as much as he could handle of "getting nowhere" with it no matter how hard he tried, stopped addressing the issue of her weight, hoping that she would eventually bring it up on her own. The war over desire was in a new phase. He allowed long silences to develop in which he hoped that she might ultimately put what she was feeling into words, which she did, but not in the way that he had hoped. As he was finding the silences increasingly difficult to tolerate, she began to take him to task for his "failure" to mention her weight when she herself wasn't mentioning it. Without the least regard for logic, she told him he had no right to stop trying to find out what she was feeling. "What did you think you were doing?" "You ought to know," she insisted, "that when I'm talking about anything else, as long as I'm still fat, it's only my good self that's talking and that I'm doing something self-destructive that you're not even caring about." It was in the course of their dealing with the apparent "no-win" quality of his "failure" that he was able to begin to find a small island of shared experience on which he could plant at least one of his feet.

"Only in my silence," she declared, "do I feel real. The only way I can get out of here [meaning her inner world] is to be silent for a year." How could it make any sense that the only way she recognized that she could release herself from the trap of her dissociated mental structure was without words, by remaining silent for a year? The point she was making was not that silence itself mattered, but silence in the presence of her analyst. Why? Because her silence in his presence could have a communicative impact as long as he hadn't given up trying. The analyst has to get fed up; it is important that he get fed up; he should get fed up. But he shouldn't get so detached from his own "fedupness" that he cannot perceive the retaliatory component of his behavior. If he is open to that, he will feel the communication from the patient as it is pressed into his soul through her silence as well as into his brain through her words. The patient was ultimately able" to put into words this remarkable insight:

When I'm not talking to you, and you don't realize that my silence is talking, I feel like I'm hurting myself and you don't care. I hurt myself by being fat in order to call attention to the inside "me." And if you don't notice or seem not to, it's like you're mad that I'm still fat and will let me hurt myself because I'm fat instead of putting why I'm fat into words. But if I do talk, it's not my fat self that's talking. So you have to find her by noticing the fat and not pretending you don't. If I get thin, no one will ever look for her, because if I stop calling attention to her existence, you will settle for my good self, which looks healthy because it is thin, and you will never know it isn't real to me. I'm like Dr. Jekyll and Mr. Hyde.

In other words, by "noticing," through the impact of forced involvement with what the patient needs to call attention to without communicative speech, the dissociated self can start to exist, and a transition begins to take place to what this patient so evocatively described as a growing awareness of "Mr. Hyde" (Hide). But the success of the transition depends on the ability of the patient to destroy successfully the analyst's unilateral experience of "what this is `really' all about"-a piece of "pathology" that has become an adversary--and thus destroy the analyst's limited image of the patient in which Mr. "Hide" is imprisoned. The problem for the analyst, of course, is that his own self-image, which is a part of all this, is also dismantled, and it is this destruction he must "survive," in Winnicott's (1969) conceptualization of object usage.

The analyst's own dissociative processes enter the picture in this context. If the patient's eating behavior is held by the therapist for too long as simply a piece of pathology to get rid of-and the patient, of course, makes it very easy for us to hold that view-nothing much changes, and what we have come to call "resistance" starts to fill up all the space. It becomes very easy to grow to hate your patient's eating disorder and then, without recognizing it, to hate that part of yourself that is trying but failing to "cure" it. The analyst begins to feel victimized by his own desire and then feels the patient's "pathology" as an adversary. The mortal enemy of an eating-disordered patient, as Boris cogently observed, is desire. Because she does not wish to want, her solution is to stimulate desire in the other; to become object, not subject; to be the object of the other's wants. In treatment, this leads to a situation with which we are all painfully familiar. The patienttherapist relationship is pulled into the patient's internal drama, which has become a substitute for living. The war over who shall hold the desire is externalized as an interpersonal war and is fought dissociatively, calling into play a constantly shifting array of the patient's and analyst's self-states. It is a war that, in one respect, the analyst has to lose in order to win. A transitional reality has to be constructed in which trust in human relatedness begins to become possible, and this can happen only through the analyst's surrender to his own dissociated self-experience.

Not so simple! And especially not so simple at those moments when reality feels like a kaleidoscope. For example, after a session that seemingly went well, a depressed patient with a long-standing eating disorder left a message on my answering machine late that night:. "Memories are beginning to come up that I've never had before, and it's very disturbing. It's like I'm watching them from a different part of my brain," she said. "It's very weird." Her voice sounded upset, but not in a panic.

Next morning, someone I hardly recognized showed up for her session, and growled menacingly:

I'm the one you need to ask the permission from! Who do you think is going to pay the rent if you keep going the way you are going? You said that I would be able to carry on with my life and my work if we agreed to do this therapy. This is bullshit! There is nothing to be gained from this. This work changes nothing. It's expensive and a waste of time.

You remind her of how alone she is, how alone she has always been, and this is supposed to be of help? She's nothing but a fat, ugly, poor kid in pain, and she has suffered enough! I won't let her suffer anymore! She knows that no one will support her if I don't. Not even the shrink will be there if the bills don't get paid. WHO DO YOU THINK PAYS THE BILLS ANYWAY???

I won't allow this! I will not allow this! I WILL NOT ALLOW THIS! As long as you threaten to disable me, I will not allow this. I am not nice, and I don't care what you think of me.

As an example of working with state-change phenomena, this incident is not as dramatic as it may sound. I selected it because it puts into high relief the point made by Stechler (1998) that "the possibility of the emergence of new states and new organizations arising during times of dysregulation and apparent disorganization or chaos, has become one of the hallmark principles of contemporary theories of self- organizing systems" (p. 8). In therapy, Stechler wrote,

It is almost entirely a renegotiation to try to facilitate the creation of new organizations and new states . . . [p. 10]. Whether that new state will be a richer, more complex, and more appropriate foundation for further development, or will be the less advantageous choice in the sense of narrowing via toxic adaptation, may depend on whether the partner in this selforganizing system biases it in one direction or the other. The more toxic adaptation can stem from an interactional partner who reacts as if his destabilization is toxic to him. That is, if the therapist's ... primary aim is to reduce his own destabilization and its accompanying anxiety as if it were toxic and intolerable, it will bias the partner's choice in the same direction [pp. 1617]. If the therapist can stay connected with his own and with the patient's destabilization, and can bias his own subsequent state choice towards openness and affective authenticity, then the patient's will be similarly biased. If the patient 'feels the freezing or the pretense of the therapist at those critical moments, the work of the therapy cannot proceed well [p. 17].

What Stechler called openness and affective authenticity require an abiding respect not only for a patient's autonomy, but an equally abiding respect for what might be called, more poetically, a patient's autotomy-- the dissociative unlinking of parts of the self in the face of potential trauma so that non omnis moriar! -- "I shall not wholly die!"

NOTES

1) See Breger's (2000) breathtakingly honest biography, Freud: Darkness in the Midst of Vision, for the most comprehensive understanding of Freud's "anti-Janet" stand and its constricting effect on the evolution of psychoanalytic theory over the past 100 years.

2) I thank Dr. Marcel Hudon for making Wislawa Szymborska's writing known to me.

REFERENCES
Boris, H. N. (1984), The problem of anorexia nervosa. Internat. J. Psycho-Anal., 65:315- 322.
____(1986), The "other" breast: Greed, envy, spite and revenge. Contemp. Psychoanal., 22:45-59.
Breger, L. (2000), Freud: Darkness in the Midst of Vision. New York: Wiley.
Breuer, J. & Freud, S. (1893-1895), Studies on hysteria. Standard Edition, 2. London: Hogarth Press, 1955.
Bromberg, P M. (1994). "Speak! that I may see you": Some reflections on dissociation, reality, and psychoanalytic listening. In: Standing in the Spaces: Essays on Clinical Process, Trauma, and Dissociation. Hillsdale, NJ: The Analytic Press, 1998, pp. 241-266.
____(1995a), Psychoanalysis, dissociation, and personality organization. In: Standing in the Spaces: Essays on Clinical Process, Trauma, and Dissociation. Hillsdale, NJ: The Analytic Press, 1998, pp. 189-204.
____(1995b), Treating patients with symptoms-and symptoms with patience. Paper presented at New York University Postdoctoral Program in Psychotherapy and Psychoanalysis, New York. p. 912____(1996), Hysteria, dissociation, and cure. In: Standing in the Spaces: Essays on Clinical Process, Trauma, and Dissociation. Hillsdale, NJ: The Analytic Press, 1998, pp. 223-237.
Ciocca, A. (1998), Psychosomatic dissociation and eating disorders. In:
Psychotherapeutic Issues on Eating Disorders, ed. P Bria, A. Ciocca & S. de Risio. Rome: Society Editrice Universo, pp. 49-55.
Erikson, E. (1968), Identity: Youth and Crisis. Austen Riggs Monogr., No. 7. New York: Norton.
Everill, J. T, Waller, G. & Macdonald, W. (1995), Reported sexual abuse and bulimic symptoms: The mediating role of dissociation. Dissociation, 8:155-159.
Gleaves, D. H., May, M. C. & Eberenz, K. P (1996), Measuring and discriminating dissociative and borderline symptomatology among women with eating disorders. Dissociation, 9:110-117.
Havens, L. L. (1973), Approaches to the Mind. Boston: Little Brown.
Janet, P (1907), The Major Symptoms of Hysteria. New York: Macmillan.
Katz, B. E. & Gleaves, D. H. (1996), Dissociative symptoms among patients with eating disorders: Associated feature or artifact of a comorbid dissociative disorder? Dissociation, 9:28-36.
Lewis, M. (1992), Shame: The Exposed Self. New York: Free Press, 1995.
MacDonald, G. (1858), Phantastes. Grand Rapids, MI: Eerdmans, 1981.
McCallum, K. E., Lock, J., Kulla, M., Rorty, M. & Wetzel, R. D. (1992), Dissociative symptoms and disorders in patients with eating disorders. Dissociation, 5:227-235.
Muller, J. P (1996), Beyond the Psychoanalytic Dyad. New York: Routledge.
Sands, S. H. (1994), What is dissociated? Dissociation, 7:145-152.
Stechler, G. (1998), Affect, the Heart of the Matter. Presented at Paul Russell Symposium, Cambridge, MA.
Szymborska, W. (1983), Autotomy. In: Postwar Polish Poetry, 3rd edition, ed. & trans. C. Milosz. Berkeley: University of California Press, pp. 115-116.
Tronick, E. Z. & Weinberg, M. K. (1997), Depressed mothers and infants: Failure to form dyadic states of consciousness. In: Postpartum Depression and Child Development, ed. L. Murray & P Cooper. New York: Guilford, pp. 54-81.
Winnicott, D. W. (1969), The use of an object and relating through identifications. In: Playing and Reality. New York: Basic Books, 1971, pp. 86-94.
Woolf, V (1928), Orlando. New York: Harcourt Brace.
Zerbe, K. J. (1993), Selves that starve and suffocate: The continuum of eating disorders and dissociative phenomena. Bull. Menn. Clin., 57:319-327. 300 Central Park West New York, NY 10024

Philip M. Bromberg, Ph.D. is Training and Supervising Analyst, William Alanson White Institute; Clinical Professor-of Psychology, New York University Postdoctoral Program in Psychotherapy and Psychoanalysis; author of Standing in the Space: Essays on Clinical Process, Trauma, and Dissociation (The Analytic Press, 1998); and member of the editorial boards of Contemporary Psychoanalysis and Psychoanalytic Dialogues. An earlier version of this essay-"Out of Body, Out of Mind, Out of Danger" originally presented as part of a panel discussion at an October 1999 conference sponsored by the Eating Disorders and Substance Abuse Service of the William Alanson White Institute, will be published by Jason Aronson as a chapter in Hungers and Compulsions, a volume of the proceedings of that conference. The more extended version published here was presented at the Self Psychology Conference of the Institute of Contemporary Psychotherapy, Washington, DC, April 2000; at the "Healing in Eating Disorders: Biological, Humanistic, and Spiritual Dimensions" conference sponsored by the Center for Complementary and Alternative Medicine, State University of New York, Stonybrook, October 2000; at the Westchester Center for Psychoanalysis, White Plains, New York, November 2000; and at the Toronto Institute for Contemporary Psychoanalysis, Toronto, Canada, September 2001.

© Philip M. Bromberg, Ph.D. 2001