Psychoanalytic boundaries and transitional space

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. Psychoanalytic Psychology 1999, Vol. 16, No. 1, 3-20 Copyright 1999 by the Educational Publishing Foundation 0736-9735/99/S3.00 Psychoanalytic Boundaries and Transitional Space Frank Summers, PhD Chicago, Illinois The blurring of the distinction between language and action in contemporary psychoanalytic theories expands the traditional boundaries of psychoanalytic therapy. The current article delineates a conceptualization of psychoanalytic boundaries based on D. Winnicott's (1971) concept that transitional space defines the psychoanalytic process. It is proposed that D. Winnicott's (1971) concept shifts the psychoanalytic paradigm to adaptation, rather than interpretation, as the overriding analytic task. The analyst's adaptation and its limitations define the psychoanalytic dyad, and psychoanalytic boundaries, from this viewpoint, are expressions of the analyst's subjectivity. The clinical implications of this concept of psychoanalytic boundaries are demonstrated in the treatment of a severely regressed patient. Historically, the analytic framework has been conceptualized as a verbal process, and other patient behavior must either be interpreted as symptomatic or contained by limit setting. Patient and therapist talk; action is a violation of this boundary. The critical distinction in this model is between speech and action, and verbalization is the boundary within which the process must operate. Contemporary revisions of psychoanalytic theory blur this distinction to varying degrees. Three separate lines of thought are leading to the breakdown of this traditional shibboleth of psychoanalytic Frank Summers, PhD, independent practice, Chicago, Illinois. Correspondence concerning this article should be addressed to Frank Summers, PhD, 333 East Ontario, Suite 4509B, Chicago, Illinois 60611. Electronic mail may be sent to [email protected]. This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. SUMMERS clinical theory. First, it is now recognized by virtually all contemporary theories of psychoanalysis that speech is action. This principle has become a standard part of American analytic philosophy since J. L. Austin (1965) attacked the distinction between stating and doing. Austin demonstrated that all use of language is action, so that even a declarative statement does something: It informs someone of something for some reason that fits the context of their relationship. More than 30 years after Austin's seminal work, psychoanalysis is coming to the same conclusion: Stating is doing. Second, it is recognized by increasing numbers of analysts, even some classical analysts, that enactments are inevitable. The point for the contemporary analyst is to understand the interactional pattern rather than engage in the illusion that it can be avoided. Third, from a variety of theoretical viewpoints, such as self psychology, object relations theories, and relational psychoanalysis, the paramount role of interpretation in therapeutic action is questioned in favor of the increasing importance accorded the patient-analyst relationship (e.g., Bacal, 1985; Bacal & Newman, 1990; Fairbairn, 1958; Guntrip, 1969; Levenson, 1991; Mitchell, 1988; Stolorow, Brandchaft, & Atwood, 1987; Winnicott, 1963/1965b, 1954/1975b). From all these theoretical perspectives, action becomes a crucial part of the analytic process rather than a deviation from it. This blurring of the distinction between speech and action means that analysts no longer have an easy, clear way to define analytic boundaries. The question becomes, If the speech-action distinction does not define the psychoanalytic boundary, what does? How one answers this question is a function of the theoretical position. Although there are a variety of theories adopted that may be considered, in this article, the issue of analytic boundaries is defined from a Winnicottian-based object relations conceptualization of the analytic process. Psychoanalytic Space From the theoretical perspective of object relations theory, personality development means the unblocking of arrested self-potential. To allow the buried self to become articulated, the analytic relationship must provide the maximum possible space for self-expression. From this viewpoint, because the task is to provide the patient the best possible opportunity to unblock self-arrestation, the analyst's posture is defined by the provision of a space that allows the old modes of being to give way and promotes the creation of a new self-structure (Summers, 1997). If previously buried parts of the self This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. PSYCHOANALYTIC BOUNDARIES AND TRANSITIONAL SPACE are to emerge, an object must be related to in a new way, created in a manner that fits the needs of the patient. Here we can identify Winnicott's (1971) conception of analytic space as an intermediate area between the reality of who the analyst is and the patient's fantasied projections. This space is limited by the analyst's reality but offers the possibility for the patient to experience this reality in a variety of ways and create new meanings within the givens of the analytic setting. Such an analytic stance is not blank because it offers the patient a particular kind of environment, but within these limits, it attempts to provide the maximum space possible for the patient's self to gain expression through a new relationship with the analyst. Whereas in his early work Winnicott (1951/1975a) proposed that the transitional area of human experience was a specific developmental phase, the residue of which endured in pockets of experience such as aesthetic and cultural pursuits, he later applied the concept of transitional space to the analytic process (Winnicott, 1971). His view was that the analyst is a given, a reality limitation outside the patient's fantasied omnipotent control. The patient imbues this reality figure with personal meaning in a manner analogous to the way the child infuses the real blanket with idiosyncratic meaning. The patient in some sense knows the analyst is an other, a person with real qualities, but treats him or her as though he were an object of the patient's creation. In this way, Winnicott saw transitional space as the very essence of psychoanalytic therapy. Although all relationships are to some degree transitional, the analytic relationship is uniquely characterized by the provision of a space the intent of which is to facilitate self-realization in one of the participants. From this viewpoint, some contemporary interpersonal approaches tend to move too far toward the blurring of the distinction between the analytic relationship and typical human interactional patterns. From the object relations viewpoint, the other self, the analyst, enters the relationship, but with a defined intent to attend to the other. Winnicott's concept that the analytic task is to adjust insofar as possible to the experience the patient requires signals a shift from interpretation to adaptation as the paramount psychoanalytic value. From the viewpoint of giving birth to the buried self, the psychoanalytic aim is adaptation to the patient to relinquish defenses and realize the self. To the extent that interpretations aid in the accomplishment of these goals, they are a necessary component of the process, but they subserve the adaptation necessary for the realization of the self. Note that the word therapy is derived from therapeutes, medieval groups of attendants, or servants. When Winnicott redefined the analytic aim from This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. SUMMERS interpretation of unconscious content to adaptation for realizing buried potential, he brought analytic therapy closer to its etymological meaning of attendance to the other. Winnicott did not fall prey to the illusion that the analyst could somehow disappear from the analytic encounter and give the analytic space completely over to the patient. The expression of the patient's self takes place within the givens of the analyst's self. The analyst works to provide the space the patient needs, and the patient must find ways to use what the analyst has to offer to create the needed experience. It must be emphasized that from the viewpoint of transitional space and adaptation, interpretations are not bearers of information to be absorbed, but offerings to be responded to as the patient needs. A good interpretation is submitted for the patient's consideration, a proposal meant to illuminate an aspect of the patient's being that the patient can use to find or create new meaning. This concept of interpretation has its analogue in the development research showing that the child uses the parental response to create meaning from the experience (Demos, 1988, 1992). The analyst offers the interpretation as a bit of reality the ultimate value of which is what the patient creates from it, however that might fit with the meaning intended by the analyst. An ineffective interpretation is an analytic offering from which the patient cannot create a meaningful experience. The implication of this concept of psychoanalytic therapy is that whatever the analyst does to facilitate the realization of the patient's potential is a legitimate component of the therapeutic process. Interpretation is a necessary, even central, means by which this space is created, but it is not the only analytic activity that helps provide the conditions for self-realization. Both interpretation and action are valid analytic techniques to the degree that they facilitate the unfolding of the patient's authentic experience. Transitional Space and the Expansion of Analytic Boundaries From the classical viewpoint, the patient's nonverbal behavior must be translated into the verbal sphere (e.g., Brenner, 1976; Oremland, 1991). Because analysis is defined from this perspective as making the unconscious conscious, to fail to verbalize the patient's nonverbal behavior is to violate analytic boundaries. However, because transitional space is created by both the verbal and nonverbal analyst-patient relationship, insistence on verbalization risks limiting this space to the point that it stifles potential self-articulation, the very aim of psychoanalytic therapy. If the analyst This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. PSYCHOANALYTIC BOUNDARIES AND TRANSITIONAL SPACE insists on verbalization, the patient may be prohibited from a wide variety of experiences that may be critical to the realization of self-development, such as dependence, temporary merger, silence, angry outbursts, the expression of rage toward a surviving object, play, the giving of gifts, sharing of possessions and information, moments of quiet contact, being alone in the presence of the other, and the demonstration of accomplishments (e.g., Winnicott, 1963/1965a, 1963/1965b, 1954/1975b). One example is a case discussed by Michael Balint (1968, pp. 128-129). At a key juncture in the treatment of an overly constricted, cautious young woman, she mentioned that she had never been able to do a somersault. Balint invited her to give it a try, the patient got off the couch and, to her amazement, did a perfect somersault. Balint believed that this event played a pivotal role in freeing this woman from inhibitions in her personal and professional life. In this example, the transitional space was used for an experience that was key to the therapeutic action. Setting limits on the patient's behavior as acting out would have deprived her of a critical therapeutic opportunity. The greater the degree of limit setting, the less is the opportunity for the patient to realize buried aspects of the self. It follows that limit setting must be kept to as minimal a level as the analyst can tolerate. To be sure, some classical analysts would insist their strategy would not be to limit the patient's behavior but to understand it (e.g., Chused, 1991; Oremland, 1991). These contemporary classicists accept enactments as inevitable, but the goal of the analysis, from their perspective, lies in the translation of nonverbal enactments into language. Even this more contemporary version of the classical position, however, defines the boundaries of the analytic as verbalization. The pivotal act of loosening Balint's patient's inhibitions lay in her performing the act of somersaulting, not in any later understanding of it. Such experiences, verbalized or not, are frequently necessary for the realization of previously buried aspects of the self, a fact that signals the need to expand the concept of analytic boundaries. Transitional Space and Boundary Limitations No individual is flexible enough to adapt to any situation the patient may seek, just as no mother is capable of perfect adaptation. Therefore, analytic tolerance becomes the key issue in the delineation of boundaries. Embedded in the analyst's limited tolerance is the reality aspect of the transitional experience. This lack of infinite flexibility means that the patient who attempts to bring the analyst into the sphere of her or his omnipotence will This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. SUMMERS eventually be confronted with the inability to control the analyst. In this way, the analyst's limitations present reality to the patient, just as the mother brings reality to the child by not meeting all the child's needs in the expected manner (Winnicott, 1960/1965c). The experience of analytic boundaries provides the other side of the transitional experience. In this context, it must be remembered that the importance of the blanket to the child is not only that it represents the mother, but that it is not the mother. For the object to be found, it must provide the function of limitation, a reality that must be adapted to. The first not-me possession helps the child traverse the journey from object relating, the world of absolute fantasy, to object usage, the world of making use of the givens of reality (Winnicott, 1963/1965a, 1951/1975a). In transitional experience, the omnipotent control of absolute fantasy is given up for the limited control of using an object. Thus, the reality limitation of the transitional object is essential to the fulfillment of its purpose: the capacity to use objects. In the analytic context, the analyst's inability to adapt absolutely to the patient requires that the patient make some degree of adjustment to the analyst in the service of usability. That is to say, the limits of the analyst's adaptation stimulate her or his transformation from an object of fantasy to a usable object. If one looks at the analytic relationship as a space designed to facilitate the unarticulated self, the analyst's adaptive limitations are the inevitable limits of who he or she is and profoundly express his or her subjectivity. Unfortunately, there is a tendency to hide this fact from patients. Two broad categories of the denial of the analyst's subjectivity can be distinguished. The first avoidance category, limit setting, is often attributed to an abstract rule allegedly imposed on the analytic dyad. Such rules, presumably limiting the analyst's behavior, lend themselves to avoidances of therapeutic responsibility. The invocation of rules that must be followed masks the fact that the analyst's inability or unwillingness to adapt is an expression of subjectivity. The fact that the analyst's decisions are as much an expression of subjectivity as the patient's material is caused Winnicott (1971) to describe psychotherapy as two "overlapping subjectivities" (p. 38). The overlap is the therapeutic space, the arena within which a new object, the analytic object, can be created. To limit this space to rules the analyst must follow is to subject the dyad to artificially imposed dictates behind which the analyst's subjectivity is hidden. In the setting of limits, clinicians are frequently content to rely on vague concepts such as inappropriate, abstractions that limit the patient's This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. PSYCHOANALYTIC BOUNDARIES AND TRANSITIONAL SPACE behavior without invoking the analyst's subjectivity. Such a lack of therapeutic responsibility turns psychoanalytic therapy into a relationship of blind rule following. It is barely imaginable that such a relationship can bring forth an authentically derived new self in one of the partners. If the analyst expects the patient to relinquish old patterns and defenses to articulate a new, more authentic self, her or his own subjectivity cannot be hidden behind unexamined rules. The limits imposed by the analyst on the relationship come from her or his own subjectivity, just as the mother's limited adaptations come from hers. If the analyst takes responsibility for limit setting, the patient experiences the reality of the analyst most poignantly in moments of boundary contact. When externally imposed rules are abandoned, it becomes clear that the analytic relationship is bounded by the reality of the analyst as a person, and the boundaries become a meaningful expression of the analyst's subjectivity, often the deepest expression of it. The second avoidance category is the use of interpretation to escape responsibility for the analyst's limitations. All too frequently analysts decide to interpret patient behavior when they cannot tolerate it (Levenson, 1991). An instructive example of this maneuver is to be found in the recent work of Slochower (1996), an illustration especially illuminating because this theorist has made a major contribution to the theory of holding in therapeutic technique. In her approach, Slochower stresses the importance of the therapist's adaptation to the patient in a manner similar to that being advocated here. In the example under discussion, the patient, Lisa, was chronically rescheduling appointments, and Slochower extended herself to meet the patient's scheduling requirements while efforts to interpret the behavior were fruitless. However, Slochower (1996) became impatient with the constant changing of appointments and her "willingness to reschedule sessions wore away" (p. 74). The therapist decided "to talk directly with Lisa about the nature of our interaction, although I was aware that in doing so I would reintroduce my subjectivity into the exchange" (Slochower, 1996, p. 74). Slochower (1996) told Lisa that they "had to look at what she was saying about her needs and how she expected me to meet them" (p. 74). The patient left treatment indicating that she would seek a more flexible therapist. Slochower wondered whether a more tactful approach had been possible, or if the treatment would have been successful had she been more adaptable. Although Slochower (1996) espouses a relational theoretical perspective, she attempted to use interpretation not to understand the patient, but to extinguish behavior that she found intolerable. When Slochower became This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. 10 SUMMERS unable to endure the rescheduling, rather than communicating her intolerance, she claimed a desire to explore Lisa's needs. By so doing, she hid her motive, her subjectivity, behind an alleged effort to understand. Although Slochower claimed to reintroduce her subjectivity, in fact, she masked her intent to get the patient to stay with an arranged schedule. Nowhere does Slochower consider the possibility that this disingenuousness in her desire to set limits played a role in the patient's decision to end treatment. Precisely because Slochower claims a theoretical perspective that includes the analyst's subjectivity, this illustration suggests that the use of understanding to veil the analyst's intent is widespread. One would expect relational analysts to be least likely to hide their subjectivity behind interpretation. If the analyst expresses subjectivity, the patient experiences an honest revelation of the analyst's intolerance, and psychotherapy becomes an arena in which self-states are genuinely communicated. One wonders if Lisa would have felt it necessary to terminate if her therapist had said, "I'm having a problem with this constant rescheduling. Is there some other way we can handle this?" Although one cannot assume that such a response would necessarily have produced a better outcome, at minimum, the issue of the therapist's intolerance would have been openly confronted. Limit setting is a decisively different experience for the patient depending on whether the analyst conveys prohibitions or refusals as consciously made decisions, thereby assuming responsibility for limiting the patient's behavior, or hides behind an analytic stance. In the latter instance, the patient tends to become suspicious of the analyst's interpretive posture. In my experience, patients feel relieved and safe when they see that the analyst will be honest with them, and they trust interpretations more when they know they will not be used to achieve hidden agendas, such as limiting their behavior. Just as important, when the analyst sets limits openly and directly, the patient knows the reality of the analyst. In this way, the analyst takes himself or herself out of the sphere of omnipotent delusion into the world of reality from which meaning must be derived. From this viewpoint, one can see that the current discussions of enactment focus too much on what the analyst does to the neglect of why she or he does it. For example, some classical analysts, such as Oremland (1991), contend that enactment should be avoided in psychoanalysis in favor of interpretation, although in psychotherapy, enactment may often be necessary. Gill (1991), in his discussion of Oremland's book, contends that the occurrence of enactments is inevitable and that the crucial analytic ingredient lies in talking about such events, so that their meaning to the patient can be elucidated. This dialogue is representative of discussions of This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. PSYCHOANALYTIC BOUNDARIES AND TRANSITIONAL SPACE 11 enactment: The focus is on the analyst's participation and its meaning to the patient. Left out of the exchange between Gill and Oremland is why the analyst acts or does not act in the ways categorized as enactment. The analyst engages in certain behavioral patterns but not others, and each such decision expresses the analyst's subjectivity in a way that has therapeutic potential. The realization of this potential depends primarily on whether the analyst has delved into his or her own subjectivity sufficiently to understand these decisions and to what degree they are expressible to the patient. For example, the analyst refuses the patient's wish for a gift. One aspect of the therapeutic potential of this event is to be gleaned from the exploration of the meaning to the patient of not receiving the gift. However, an equally critical issue is why the analyst refused. The analyst's unwillingness to give to the patient in this way is a significant component of the relationship, the therapeutic potential of which will be lost if this side of the event is unexplored. If such exploration is undertaken, it often transpires that the analyst's behavior is often less important than why it is done. The fact that transitional space is not pure fantasy suggests that efforts at absolute adaptation are not only futile, but antianalytic, violations of the space required for the analytic process. The analyst who would attempt to meet all patient needs would be in danger of creating the delusion that he or she could be controlled in an omnipotent fashion by the patient. From the viewpoint of this conceptualization, a physical relationship between patient and analyst is a violation of the very essence of the analytic relationship. Such a relationship suffocates the space that it is the purpose of psychoanalysis to create and use. Psychological development can occur only if the psyche has the opportunity to play with and create meaning, a process that requires the patient's desire for physical contact to be articulated in transitional space. When the analyst meets the patient's desire with a physical response, symbolization is replaced by concrete gratification, the transitional space disappears, and the essence of the psychoanalytic experience is gone. It is true, therefore, that the maintenance of the transitional arena defines an absolute boundary within which the psychoanalytic relationship transpires. When an analyst feels the urge to meet the patient's desire for a physical relationship, she or he is constrained by the nature of the analytic role and responsibility to that role. Thus, in addition to the boundaries constructed by the analyst's idiosyncratic subjectivity, an absolute boundary is determined by the nature of the analytic dyad. However, even the existence of this absolute boundary does not permit masking the analyst's This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. 12 SUMMERS subjectivity. Although it may well be tempting to invoke this boundary as a predetermined rule in response to the patient's desire for a physical relationship, to do so is to fail to take responsibility for the construction of the analytic relationship. To believe in an absolute boundary for the analytic dyad is to choose a particular way of working with patients. To refuse the patient's desire for a physical relationship on the grounds that it is not permitted by the rules of analysis is failure to assume ownership for analytic boundaries. Within the absolute boundary required for the maintenance of the analytic space, the analyst's idiosyncratic subjectivity defines the boundaries of analytic activity. From this viewpoint, one can conceptualize at least some of the sexual transgressions that have become all too common among psychoanalysts as extensions of efforts at absolute adaptation. I have had limited experience with patients who had sexual experience with previous analysts, but in those cases I have seen, the sexual experience seemed to be an outgrowth of an extended effort by the analyst to meet all patient needs. That is to say, the analyst has indulged in enacting the fantasy of being the omnipotent object for the patient. These analysts seem to fall at the opposite end of the spectrum from the cases discussed by Mayer (1997) who attempt to limit their own participation to such an extreme that any inkling of affective response runs the danger of serious boundary violation. For whatever reason, the analysts under discussion seemed drawn to an interaction in which they would save the patient from the effects of previous trauma, behavior rationalized as being the "good mother" the patient never had. Once the transitional boundary had been blurred by the analyst's uncontrolled desire to be the omnipotent object who would meet the patient's every need, it was a short step to the sexual transgression. To be sure, the reasons for analyst sexual transgressions are varied and complex, and I am by no means suggesting that this dynamic is the only motivation for such serious boundary violations. I am suggesting that violation of the transitional space with attempts at absolute adaptation can result in the most egregious boundary violations. Recognition of the role of both responsiveness and adaptive failure in psychic growth fits developmental research findings from remarkably different traditions. Because the most successful mother-infant dyads demonstrate both affective and timing matching as well as allowance for discontinuity, it appears that excessive coordination does not provide the child with sufficient diversity (Beebe, 1995; Beebe & Lachmann, 1994). The most successful mother—infant dyads have an optimal combination of mutual and self-regulation, a result that indicates the child needs to be This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. PSYCHOANALYTIC BOUNDARIES AND TRANSITIONAL SPACE 13 regulated by the mother and to be left alone (Demos, 1992, 1994). The key for the caretaker is to gauge the right amount of struggle: Too much makes the infant overwhelmed and helpless; too little does not permit the achievement of agency. From a theoretical viewpoint, Winnicott's (1971) and Benjamin's (1995) observations on the development of intersubjectivity are especially relevant. Winnicott (1971) pointed out that the infant must "kill" (p. 92) the mother over and over and see that she survives to become aware that the mother is not an object of fantasy, but a real person, an independent center of initiative. In addition, the mother's failures in adaptation bring reality to the child in manageable doses. Benjamin adds that in the separation-individuation process, the child not only develops a sense of self but also recognizes that the mother has a subjectivity of her own to which the child must relate. The child's increasing separation from the mother provides the opportunity to relate to the mother as a separate mind, a new level of intersubjectivity. Therefore, the mother's difference, her separation from the child, is critical for the development of both the child's recognition of others as others and the sense of reality. The insights of Winnicott and Benjamin are remarkably coincident with the findings of developmental research: The mother's attunement must be mixed with her separateness, her own mental process not matched by that of the child. The analytic process requires an analogous mixture of connection and separation. As much as the patient needs affective resonance for self-development, he or she requires the analyst's separateness for the recognition of the other. One can see from these findings and theoretical observations the powerful importance to the patient of analytic limitations. Every moment of opposition between patient and analyst is a potential intersubjective contact that confronts the patient with the reality of the analyst's mind. Analysts who insist that only interpretation is acceptable limit the space for the articulation of the patient's mind. On the other hand, the analyst who would make efforts at absolute adaptation would be relating as though lacking a mind and therefore rob the patient of the opportunity for recognition of full intersubjectivity, the experience of others as fully developed subjects. Blurring of boundaries robs the patient of this crucial experience, thus arresting the development of intersubjectivity, and the psyche cannot extricate itself from its self-involvement. One of the major benefits of limit setting as the expression of the analyst's subjectivity is its role in the relinquishment of narcissism. Without experiencing the analyst's limitations the patient would be denied the opportunity to find a way to use a real person with real limitations and qualities to form 14 SUMMERS a personally meaningful experience. In the absence of adaptation to the patient and exploration of its limitations, there is not enough reality to use. This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. Clinical Illustration Sharon, a 31-year-old married, childless woman, was referred by her former therapist after both decided that termination of their relationship was desirable. The therapist, who apologized for the referral, described the patient as having undergone a surprising and frightening regression. After about 1 year of psychotherapy, during which she discussed her ambivalence about having children and her job, she began to demand that the therapist physically hold and caress her. He complied with her request for holding, and that activity became a major component of the therapy. In addition, Sharon called the analyst at all hours, often staying silent on the telephone for lengthy periods. During this time, she was dismissed from her job, her functioning became severely impaired, and she ceased working. When the therapist began to hesitate to fulfill her requests for physical contact, Sharon became enraged and threatened to kill him, a reaction that made him so fearful he suggested hospitalization. Terrified by this suggestion, Sharon bolted from the office in a panic. At this point, both decided their relationship had become so tainted that it should be terminated. I found Sharon to be a somewhat older looking, tightly held, inhibited looking, neatly but casually dressed woman. She talked without affect about her lack of functioning and difficulties with her previous analyst. Clutching her chair tightly, Sharon said that she was rarely aware of what she felt and, in an emotionless manner, discussed her difficulty with anger and affect of any type. Despairing of any effort to change herself, Sharon manifested a general hopelessness regarding her condition and her capacity for an emotional relationship. At certain points of extreme despair, her voice changed into that of a little girl, and she asked to be held, cuddled, or played with. She brought in children's books that she either read aloud or asked me to read to her. In one session, she came with her kitten and asked me to engage in play with the kitten and her. When Sharon's voice became "little," she often referred to herself in the third person. Feeling no control over this part of herself, Sharon sought a safe environment in which to express her "little girl side" and felt that verbal responses from me were rejecting and implied that she was "wrong," a response she identified with her mother's coldness. Frequently silent for long periods, Sharon once did not speak for an entire session. In another therapy hour, she kept her back to me and spoke This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. PSYCHOANALYTIC BOUNDARIES AND TRANSITIONAL SPACE 15 only when spoken to. In response to my question, she acknowledged that she wished to be pursued. In a subsequent session, she stood silently, apparently immobilized, in the center of the consulting room for most of the session. Phone calls became frequent, although their purpose was often unclear, and verbal communication was minimal. I felt that I was in the presence of someone desperately trying to form a connection but paralyzed with anxiety. Rather than feeling an inclination to set limits on her behavior, I felt a distance between us I wanted to break through, but I was unsure how to cross the chasm. To that extent, my feeling of a frustrated effort at connection mirrored hers. After a period of time, Sharon established a routine for the sessions. She would bring over my desk chair, sit with her feet on it, cover her body with an afghan and either was silent or spoke in a low, soft, almost pleading voice, often expressing the need to be taken care of. When asked how she was feeling, her usual reply was the utterance of a barely audible "scared." At the end of one session, she refused to leave. Because I felt helpless to end the hour and concerned about the next patient waiting, that was the only incident in which I felt her behavior was unacceptable. I told her I would have to call the police, and she left in anger. There were also celebrations. On birthdays or anniversaries of major events, she brought cakes she had baked, and we shared them. The details of this lengthy and complex treatment have been described elsewhere (Summers, 1988). In this article, I focus on its implications for conceptualizing the boundaries of psychoanalytic therapy. As Sharon became more regressed in therapy and her demands and expectations escalated, I felt a pull toward being controlled and often felt like saying no simply to affirm my autonomy. Nonetheless, I became increasingly aware of the depth of her pain, isolation, loneliness, and depression. I wanted to give to her, but I was not sure how. I began to experience her demands as desperate efforts to gain some nurturing without feeling exposed to injury. It became evident to me that if she could tell me how to give to her, she would feel less vulnerable in her need for me. In any event, I experienced her demands for nurturance as the emergence of needs that made her feel weak, ashamed, and, worst of all, open to being attacked. I responded verbally to this need and her fear of exposing it, but I had no conviction that such comments would have an impact, and I was right. When I told her that she was seeking comfort for the baby in her who had been so long unattended, she nodded in the affirmative, but it was clear that such comments did not alter her need to be nurtured. She required an adaptation to this need, and I did my best to meet it by allowing her the This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. 16 SUMMERS space to be as much of her "little" self as she could in the therapy sessions. I often asked her if there was anything I could do to make her more comfortable while she was curled up in her afghan, often asleep. I stayed with her both physically and emotionally even when she was silent for the entire session. When I did talk, I said that she seemed to need a place where she could "let go" and not feel any sense of burden, that she needed to know I was there. I was mindful of not letting my attention wander from her too much, although it did sometimes happen, because I felt that she needed to be held, not forgotten about. My experience was that if I wandered from attentiveness for a sustained period, I would be abandoning her. This behavior on my part was an adaptation to her need for the experience of being held, a connection in which there was no burden on her to do anything other than be and be taken care of. It was in the midst of this deep regression that Sharon asked if we could read books together. She brought in children's books, such as The Velveteen Rabbit, and I felt very comfortable reading them to her. When she brought in cake for us to share on celebratory occasions, I ate with her as we chatted. During one session of deep regression, she asked in her little girl voice if I would physically hold her. Although this request did not surprise me, it disturbed me because I did not want to meet it, I knew I would not meet it, and I feared for the ruin of all we had accomplished by my unwillingness to comply. However, I felt no desire to accede to her request. My feeling was that physical holding would lose my sense of boundaries and with it the therapeutic process. I felt that my identity, even my sanity, depended on the maintenance of a firm boundary between the physical and the psychological. Nonetheless, I knew she would be injured by my refusal, and I hesitated to give it. I said, "No, I would not. I want to hold you, but not that way. I cannot touch you without feeling that I am losing my sense of what this is and even who I am." Although Sharon was distressed at my refusal of her request, she was surprised at and pleased with the candor of my reply. My explanation gave her pause, and she seemed to think about it for a moment. Nonetheless, she indicated that she needed holding and did not know how "this was going to work" if I could not provide what she needed. I replied that although I would not physically hold her, I had the impression that she had been held during previous sessions. She agreed but still felt she needed the physical experience. I said that I guessed she was not sure of my holding, of my "really being there" without the physical sensation, and that our work was about helping her to feel held, to feel my This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. PSYCHOANALYTIC BOUNDARIES AND TRANSITIONAL SPACE 17 presence, the reality of our relationship, without the physical reassurance of it. Initially she voiced skepticism and questioned whether I was not rationalizing an unwillingness to help her the way she needed, but over time she found a way to adapt to my limitations: She curled up in her afghan and slept while I watched over her. She later told me she felt like a "babe in arms" and experienced the most peaceful sleep of her life during these sessions. After a period of time, she began to play and talk more frequently in the sessions, and it was during this phase that she brought her kitten in and asked me to play with her. I did play rather lightly with the kitten but was not as engaged with it as she wanted me to be. When she questioned my limited interest, I told her that although I found her kitten cute, I did not have a passion for cats and was not as excited about it as she was. Although chagrined at my limited interest in the kitten, she was surprised and pleased that I was happy to read her stories and play other games with her and that I showed genuine enthusiasm for accomplishments of which she was proud. I felt maternal toward her, but I also felt that she was navigating around limitations in my "motherhood." It is striking that this woman who had been so rigid in demanding how she was to be treated and so easily injured when her requests were rebuffed reacted without a dire sense of threat or panic when her most sensitive request was denied. She did not respond with the depression, despair, and hopelessness that had characterized her previous reactions to disappointments. I believe her relatively benign reaction can be attributed to two primary factors: (a) She had by that point felt given to enough that she did not feel totally rejected or empty, and (b) the candor of my response made her feel that the refusal was an authentic expression of who I was, and this made her feel connected to me in a new way. Most important, it motivated her to find other ways to have her needs met. Her ability to do that mitigated her sense of helplessness and eventually led through a tortuous process to functioning. I believe that her ability to accept limitations began her journey back to the world of the give-and-take of normal human intercourse, but this odyssey could not have begun without the adaptation that met some of her needs. In this context, she was able not only to accept my limitations but also eventually to use this experience to engage the adult world of reality. This story has a happy ending. After a long and difficult treatment, Sharon became functional, found a responsible professional job that suited her talents, had one child, and was pregnant with her second when she ended treatment. By chance, I have seen her twice since the termination, This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. 18 SUMMERS and she expressed pleasure in both her work and a family life that now includes three children. Cases like Sharon tempt the conclusion that therapeutic adaptation is a technique reserved for more disturbed patients whereas verbal interpretation remains the modality for the healthier patients. However, I do not wish to draw such a sharp dichotomy. Many highly functioning patients require nonverbal adaptive responses at particular points in treatment. The difference is that for the latter group, these techniques tend to be less central to the treatment. Most patients tend to need a mixture of interpretation and other forms of adaptation. The difference between the treatment of the more highly functioning patients compared with more seriously disturbed patients such as Sharon lies in the relative proportions of these two therapeutic strategies. Conclusion The reconceptualization of psychoanalytic therapy as transitional space for the birth and development of buried self-potential shifts dramatically the concept of analytic boundaries. This model of psychoanalysis extends legitimate analytic activity well beyond speech to the provision of needed experience, not simply for a subgroup of patients, but for many patients at points in treatment. Although a physical relationship is prohibited by the very nature of the analytic relationship as transitional space, all other analytic boundaries are expressions of the unique subjectivity of the analyst. Limit setting, being an expression of the analyst's subjectivity, provides potentially powerful therapeutic opportunities as the "otherness" of the transitional experience from which the patient creates meaning. Therefore, the analyst's willingness to assume ownership of the imposition of limits is a highly significant component of the analytic change process. Adaptation and limit setting work in synchrony to provide the blend of fantasy and reality that forms the uniquely psychoanalytic experience. Limit setting without adaptation does not provide enough space for the patient's self, and adaptation without limits leaves no room for the analyst's mind. References Austin, J. L. (1965). How to do things with words. New York: Oxford University Press. Bacal, H. (1985). Optimal responsiveness and the therapeutic process. In A. Goldberg (Ed.), Progress in self psychology (Vol. 1, pp. 202-226). New York: Guilford Press. This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. PSYCHOANALYTIC BOUNDARIES AND TRANSITIONAL SPACE 19 Bacal, H., & Newman, K. (1990). Theories of object relations: Bridges to self psychology. New York: Columbia University Press. Balint, M. (1968). The basic fault. London: Tavistock. Beebe, B. (1995, April 29). A dyadic system theory of the mother-infant relationship. Paper presented at the Spring Meeting of the Division of Psychoanalysis, American Psychological Association, Santa Monica, CA. Beebe, B., & Lachmann, F. (1994). Representation and internalization in infancy: Three principles of salience. Psychoanalytic Psychology, 11, 127-162. Benjamin, J. (1995). Like subjects, love objects. New Haven, CT: Yale University Press. Brenner, C. (1976). Psychoanalytic technique and psychic conflict. New York: International University Press. Chused, J. (1991). The evocative power of enactments. Journal of the American Psychoanalytic Association, 39, 615-640. Demos, V. (1988). Affect and the development of the self. In A. Goldberg (Ed.), Frontiers in self psychology: Progress in self psychology (Vol. 3, pp. 27-53). Hillsdale, NJ: Analytic Press. Demos, V. (1992). The early organization of the psyche. In J. Barren, M. Eagle, & D. Wolitzky (Eds.), Interface of psychoanalysis and psychology (pp. 200-233). Washington, DC: American Psychological Association. Demos, V. (1994, May 24). Links between mother-infant transactions and the infant's psychic organization. Paper presented at a meeting of the Chicago Psychoanalytic Society, Chicago. Fairbairn, R. (1958). On the nature and aims of psycho-analytical treatment. International Journal of Psycho-Analysis, 39, 374—385. Gill, M. (1991). Indirect suggestion: A response to Oremland's Interpretation and Interaction. In J. Oremland Interpretation and interaction: Psychoanalysis or psychotherapy? (pp. 137-162). Hillsdale, NJ: Analytic Press. Guntrip, H. (1969). Schizoid phenomena, object relations, and the self. New York: International Universities Press. Levenson, E. (1991). The purloined self. New York: William Alanson White Institute. Mayer, E. (1997, February 15). Sexual abuse of patients and problems in our psychoanalytic theory of technique. Paper presented at the Fifteenth Biennial Conference on Psychoanalytic Psychotherapy, Institute for Psychoanalysis, Chicago. Mitchell, S. (1988). Relational concepts in psychoanalysis. Cambridge, MA: Harvard University Press. Oremland, J. (1991). Interpretation and interaction: Psychoanalysis or psychotherapy? Hillsdale, NJ: Analytic Press. Slochower, J. (1996). Holding and psychoanalysis. Hillsdale, NJ: Analytic Press. Stolorow, R., Brandchaft, B., & Atwood, R. (1987). Psychoanalytic treatment: An intersubjective approach. Hillsdale, NJ: Analytic Press. Summers, F. (1988). Psychoanalytic therapy of the borderline patient: Treating the fusion-separation contradiction. Psychoanalytic Psychology, 5, 339-355. Summers, F. (1997). Transcending the self: An object relations model of the therapeutic action of psychoanalysis. Contemporary Psychoanalysis, 33, 411^128. Winnicott, D. (1965a). Communicating and non-communicating leading to a study of certain opposites. In Maturational processes and the facilitating environment (pp. This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. 20 SUMMERS 170-192). New York: International Universities Press. (Original work published 1963) Winnicott, D. (1965b). Dependence in infant care, in child-care, and in the psychoanalytic setting. In Maturational processes and the facilitating environment (pp. 249-260). New York: International Universities Press. (Original work published 1963) Winnicott, D. (1965c). Theory of the parent-infant relationship. In Maturational processes and the facilitating environment (pp. 37-55). New York: International Universities Press. (Original work published 1960) Winnicott, D. (1971). Playing and reality. London: Routledge. Winnicott, D. (1975a). Transitional objects and transitional phenomena. In Through paediatrics to psychoanalysis (pp. 229-242). New York: Basic Books. (Original work published 1951) Winnicott, D. (1975b). Withdrawal and regression. In Through paediatrics to psychoanalysis (pp. 255-261). New York: Basic Books. (Original work published 1954)