Almost every introduction to the concept of the borderline state (or the borderline disorder) begins with remarks about how problematic, confusing, and at times even embarrassing this concept has become. This predicament is understandable when we consider that the borderline concept originated in a strange admixture of inclusive and exclusive criteria (those of neurosis and psychosis), and that it was refined in the context of manifest theoretical tensions as well as latent personal investments in them. Hence to trace the development of a term which has come to function as the umbrella for a number of different psychopathological entities is a daunting task that will inevitably involve condensation, simplification, and sacrifice of subtlety and nuance.
Since borders are delineated by the two lands they divide, we can think of a geographical analogy for at least the early origins of the concept (Leichtman 1989). The lands, descriptive psychiatry and depth psychology, have at varying times either joined forces or considered each other with suspicion. And, although the term “borderline” is undoubtedly associated with psychoanalysis (the most elaborate expression of depth psychology), the clinical conditions it represents must have existed before psychoanalysis.
The first enlightened modern psychiatrist, Philippe Pinel (1801), described the manie sans délire, a subcategory of one of the four types of insanity. This excitement without delirium is the closest to what we would today call a severe personality disorder in a regressed state.
A quarter of a century later, John Pritchard (1835) added a new clinical syndrome that he called moral insanity, a form of psychopathology that did not entail cognitive impairment. Originally his term covered many individuals whom we would now call borderline, though later it gradually became restricted to psychopathic characters.
In 1890, thirty-five years after his groundbreaking description of manic depressive illness (folie circulaire), Jean-Pierre Falret observed another type of madness, the folie hystérique. It is fascinating to recognize in his description of splitting, impulsivity, and affective lability many of the traits of today’s borderline.
The two great descriptive psychiatrists who were both born the year after Freud, Emil Kraepelin (1905) and Eugen Bleuler (1911), were also intrigued by that middle territory between plain psychosis and unconventional normalcy. And in the narrow confines of this paper, the names of Ernst Kretschmer (1925), and Kurt Schneider (1923) ought to be mentioned. Both made significant contributions to the study of personality disorders but are largely ignored in the modern bibliography.
We may observe two trends in the early attempts of descriptive psychiatrists to come to terms with the diagnosis and understanding of severe personality disorders: first, the tendency to view these disorders as pre-morbid forms (or formes frustes, or good-outcome cases) of the main psychoses, particularly schizophrenia; second, and more intriguing, the tendency to approach with a broad outlook the entire spectrum of personality disorders and then to narrow the focus of clinical research to those with strong psychopathic traits and antisocial behaviors. It is as though, lacking the interpersonally derived data made available through the psychodynamic approach, the academic psychiatrist-clinician had no alternative but to restrict himself to the more loud and persistent antisocial behaviors (Stone 1986).
Finally, there is Karl Jaspers’ General Psychopathology (1913, 1923, 1946), that grand oeuvre of phenomenological psychiatry, long neglected in the Anglo-Saxon world (it was translated into English fifty years after its first edition). Here we find, with surprise, the description of two personality types, the self-insecure, who would correspond to the narcissistic personality of our time, and the incurable personality, who would approximate what we would call the severe borderline.
The adjective “incurable” reveals the limitations of the descriptive-nosographic approach. With few exceptions, this descriptive approach remained dormant till the late ’60s.1 At that time in the United States the proliferation of mental-health centers, the availability of psychotherapists, and the pressing need to become more rigorous in psychiatric diagnosis resulted in a renewed undertaking of empirical studies, this time with the collaboration of psychoanalysts.2 Space does not permit me to go into details of the descriptive approach, other than to underscore that, in recent years, empirical research on the comorbidity of the borderline disorder has shifted its focus from schizophrenia to affective disorders. This by no means suggests that the association with affective disorders is clear and definite; many borderlines never develop a clear affective disorder and their nuclear symptoms do not change when treated with antidepressant medication.
In psychoanalysis, the term borderline was first used, once, by Wilhelm Reich in his 1925 monograph on the “impulse-ridden character.” With what would amount to a premonition of ideas to come, Reich noted the marked ambivalence, the primacy of pregenital aggression, the definite ego and superego defects, and the primitive narcissism that characterize these patients.
Though the concept of the borderline is present in the work of all the British object-relations theorists, from Melanie Klein to Fairbairn, Winnicott, Bion, and Balint, as a term and as a clinical diagnostic category, it is, in birth and development, purely American. It is this American context that my paper addresses. In an effort to be schematic, and at the risk of oversimplification, I would distinguish three periods of evolution of the borderline concept in the United States.
I would call the years from 1938 to 1953 the period of unexpected realizations. During this time psychoanalysts recognized with surprise that many of their office patients did not benefit from the classic technique; occasionally such patients even became worse, developing transient psychotic states. While some (A. Stern, H. Deutsch) contributed elaborate clinical descriptions, others (P. Greenacre) turned their attention to the pathogenesis of the disorder.
Most reviewers agree that the actual term borderline was established in the psychoanalytic literature by Adolphe Stern in his 1938 paper “Psychoanalytic Investigation and Therapy in the Borderline Group of Neuroses” (Stone 1986). Succinctly but in detail Stern described the borderline condition under the following ten headings: narcissism, psychic bleeding, inordinate hypersensitivity, psychic and bodily rigidity, negative therapeutic reaction, constitutionally based feelings of inferiority deeply embedded in the personality, masochism, a state of deep organic insecurity or anxiety, use of projective mechanisms, and difficulties in reality testing, particularly in the area of personal relationships.
Stern thought that, for the borderline, narcissism is the substrate from which defenses originate on the basis of needs, whereas for the psychoneurotic, anxiety arises in connection with psychosexual impulses. Three years later, in an effort to shed light on the pathogenesis of these conditions, Phyllis Greenacre (1941) proposed the idea of constitutional predisposition to anxiety, which she called basic or elemental anxiety:
I believe this organic stamp of suffering to consist of a genuine physiological sensitivity, a kind of increased indelibility of reaction, which heightens the anxiety potential and gives greater resonance to the anxieties of later life. The increase in early tension results first in an increase in narcissism and later in an insecure and easily slipping sense of reality. I referred (in an earlier paper) to an increase in the sense of omnipotence which may occur to overcome or balance the preanxiety tension state of the organism, and to an increased mirroring tendency arising partly from the imperfectly developed sense of reality. This tendency is the antecedent of the tendency towards overfacile identification. Libidinal attachments are urgent but shallow. The patient is not well individuated, with the libido quickly and urgently invested and withdrawn. [p. 618]
Helene Deutsch did not use the term borderline, but her 1942 description of the “as-if” personality foreshadows Winnicott’s concept of the false self, Erikson’s concept of ego diffusion, and many of Kernberg’s diagnostic criteria for borderline psychopathology (identity diffusion, poor sublimatory potential, emptiness, impulsivity, etc.). She called attention to the disturbance of object relationships; the lack of warmth; the excessive formality; the quality of always acting as though on stage but without any enthusiasm; the absence of conviction as to one’s true substance, leading to adaptation via superficial identification with others; and defectiveness in sublimatory activities, since no endeavor seems more real or engaging than any other.
The second period in the development of the borderline concept started in the early ’50s with the classic paper of Robert Knight on “Borderline States” (1953), and the subsequent two all-day panel discussions held by the American Psychoanalytic Association. I shall call this the period of the elusive consolidation, and I’ll explain why.
During this time, mainstream American psychoanalysis felt threatened by the innovations in psychoanalytic technique proposed by Franz Alexander and his Chicago colleagues; these were parameters of technique that came under the rubric of the “corrective emotional experience.” As a result, the early ’50s were characterized by the concerted effort to redefine psychoanalysis and to differentiate it from psychotherapy, so as to avoid what Freud (1919) had warned against as “[alloying] the pure gold of analysis freely with the copper of direct suggestion” (p. 168). This was a time of closing of the ranks. It was hoped that an overarching theoretical construction, based on the paradigm of ego psychology, would explain and contain all forms of psychopathology. Consequently all clinical observations were filtered through the model of the tripartite mind, and all therapeutic interventions were molded, elaborated, and verbalized in terms of id, ego, and superego. This represented a wished-for consolidation of clinical knowledge and theory in psychoanalysis.
Thus it is the basic assumptions of ego psychology that are reflected in Knight’s paper on “Borderline States,” in which he maintained that, as a result of constitution, traumatic events, and disturbed human relationships, the ego of the borderline patient is fragile and hence unable to maintain its functioning. If left to its own devices, the ego of the borderline patient gravitates toward autistic thinking, which is indicative of the schizophrenic tendencies underlying it.
Knight insisted on a careful diagnostic assessment of the ego functions of the prospective analysand, and he cautioned against interpreting the few defenses left to the ego of the patient. On the whole, he felt that psychoanalysis was contraindicated. Although Knight disapproved of the term “borderline,” his understanding and description were so resonant with the mandated ego-psychological postulates of the decade that his paper ended up promoting both the term and the category of the borderline as distinct from the psychoses.
Inspired by ego psychology, but feeling the need to move beyond the confines of the consulting room and the restrictive overdetermination of the first six years of life, Erik Erikson (1950) proposed a synthesis of sequential life tasks leading to integrity and emotional health. He amplified and enriched the concept of identity and its pathological counterpart, identity diffusion; years later, Otto Kernberg would assign to the latter the status of a cardinal and pathognomonic constellation of symptoms.
Otto Kernberg’s work ushers in the third period (1967–1975) of the development of the borderline concept. I will resort to a musical analogy and call this the period of the re-orchestration of the score. But before describing this third period, I would like to discuss some particular aspects of the nature of the score that was so much in need of re-orchestration. I am referring to the intellectual hegemony of ego psychology in the United States.
Every country creates the psychoanalysis it needs. Analysts do not function in a vacuum; they relate to the philosophical currents and struggle with the intellectual controversies of their times, and, wittingly or not, they participate in the conventions of the native culture. The psychosocial environment of the United States when the immigrant analysts arrived there in the ’30s and ’40s was very different from the one they were coming from. In the decade preceding World War II, American culture was permeated by a belief in a civilized morality and a coherent system regulating social and religious norms, defining behavior, prescribing a unique regime of sexual hygiene, and stipulating models of manhood and womanhood. At the same time, there existed a fierce passion for business, upward mobility, and acquisition of wealth.
The American world view presumed that willing something can render it actual (W. James); that truth is a consensus formed by a community of inquirers over time (C. Peirce); that nature is a plastic medium predisposed to human mastery (J. Dewey); that life progresses toward the better since its meaning, always contingent, awaits the future and shuns the past; and, finally, that life’s overarching purpose is endless action leading to progress (Diggins 1993).
The immigrating Jewish analysts were highly educated and cultured scholars and scientists who had been driven out of their homes by an ominous fanaticism. They were grieving for two losses: that of their home, family, and country, and, from 1939 on, that of their spiritual father, Freud. When we consider that the reaction to object loss consists of introjection and preservation of the object by psychological means, then we can understand why Freud’s last theoretical legacy (which, coincidentally, was being expanded by the only other living psychoanalyst bearing his name, Anna Freud) became all the more venerable and treasured. Among other tenets, this model, called the tripartite or structural model of the mind, entailed the idea of human adaptation involving autoplastic modification of the self and alloplastic modification of the environment, and the idea of a conflict-free sphere of the ego.
On the clinical level psychoanalysis accordingly defined itself as the means of freeing the ego from its crippling defenses and rendering it more adaptive to the environment. On the scholarly level psychoanalysis strove to become a psychology of normal development, with the ego being the field of empirical observation and hypothesis testing. Thus psychoanalysis, transferred across the ocean, could become a positive and positivistic science in the land of pragmatism.
In the period between the two world wars American academic psychiatrists lacked the educational background of their European colleagues, with its emphasis on the philosophical inquiries of phenomenology and existentialism. Nevertheless, most of them had been inoculated by the teachings of Adolph Meyer at Johns Hopkins University and were eager to have more systematic psychodynamic training. The newly arrived Middle European analysts promised to fulfill that desire and were put in a position of respectability and esteem. Thus externally derived expectations of authoritativeness were added to the inner psychological imperative of grieving for what had been lost. The price was considerable: two key elements of psychoanalysis, ambiguity and potential for subversiveness, suffered. It is interesting to note that intricacies involving language and (to a lesser extent) the unconscious did not become the focus of interest and passionate involvement as they did for European analysts. How could, for example, a foreign-born analyst who needed to maintain the authority attributed to him risk the ambiguity inherent in word play?
Kernberg emerged in American psychoanalysis in the late ’60s with a series of papers on character pathology (Kernberg 1967, 1975, 1984). He posited a specific, albeit broad, realm of psychological functioning that was flanked on the one side by the neurotic personality organization and on the other by that of psychosis. He considered this neither a nosological entity nor a temporary state wavering between neurosis and psychosis but a stable and specific psychological structure that he called borderline personality organization.
Kernberg established the line between neurotic and borderline conditions by contrasting the defense of splitting—characteristic of borderline personality organization—with the more advanced defense of repression, indicative of neurotic functioning; he distinguished borderlines from psychotics by the capacity of the former to test reality. Under borderline personality organization he included all the clinical manifestations of severe character pathology, that is, the antisocial personalities, the self-mutilators, the severe addicts, the polymorphous perverts, the “as-if” and the prepsychotic characters (schizoid, paranoid, cyclothymic), and so forth.
Kernberg distinguished three sets of criteria, or features of the borderline personality organization: descriptive, genetic-dynamic, and structural. The descriptive criteria are chronic, diffuse anxiety; polysymptomatic neurosis; polymorphous perverse sexuality; the “classical” prepsychotic personality structures; impulse neurosis; addictions; and “lower level” character disorders. Genetic-dynamic elements include specific condensation of genital and pregenital conflicts; a precocious development of oedipal conflicts from the second and third year on; and the contamination of oedipal object images by oral and anal rage. The structural criteria are nonspecific manifestations of ego weakness (poor tolerance of frustration and anxiety, lack of impulse control, inadequate sublimation); the predominance over repression of splitting and the allied archaic defense operations of idealization, projective identification, denial, omnipotence, and devaluation; pathological internalized object relations; and primary-process thinking in unstructured situations.
Devising a specific approach to assessment that he called the structural interview, Kernberg shifted the diagnostic emphasis from the evaluation of symptoms to that of the structural determinants of character. He reformulated the development of internalized object relationships as follows. In early development, the ego has two essential tasks to accomplish: the differentiation of self-images from object-images and the integration of both self- and object-images under the influence of libidinal drive derivatives. In the future borderline the first task is attained, but the second one, the integration of libidinally determined and aggressively determined self- and object-images, fails because of the pathological predominance of primitive pregenital oral aggression. The resulting lack of synthesis of contradictory self- and object-images interferes with the integration of the self-concept and the establishing of a total object relationship and object constancy.
The need to keep good self- and object-images from being contaminated by aggressive influences leads to an active process of keeping apart introjections and identifications of opposite quality—that is, to splitting. Splitting, then, is a fundamental cause of ego weakness. Since splitting also requires less countercathexis than repression, a weak ego easily falls back on splitting and a vicious circle is created whereby ego weakness and splitting reinforce each other. This leads to seriously compromised oedipal and postoedipal identifications and, eventually, to identity diffusion (Meissner 1984).
Kernberg based his special form of modified psychoanalytic psychotherapy on this reciprocal and mutually reinforcing relationship between splitting and ego weakness. He believed (as had Knight [1953] and Frosch [1964] earlier), that borderline patients do not tolerate the regression brought about by psychoanalysis, nor are they helped by supportive psychotherapy. The latter leads to a precarious dissociation of the negative transference, which never gets worked through but is instead enacted on other persons while the treatment relationship remains shallow.
According to Kernberg, the borderline develops a primitive transference that reflects numerous dissociated or split-off aspects of the self and distorted and fantastic dissociated or split-off object representations. He therefore advocated a series of technical modifications involving frequent (three times a week) face-to-face interviews as well as systematic working through of the negative transference in the here and now, without attempting its genetic reconstruction. He felt that a complete transference neurosis should not be permitted to develop, nor should the transference be resolved by interpretation alone. But if the negative transference, along with its parallel manifestations in the patient’s actual milieu, is systematically analyzed, then the splitting operations are weakened, the ego is strengthened, and a gradual integration of good and bad self- and object-representations takes place.
Kernberg’s theoretical integrations and technical suggestions fertilized the field and brought together empirical-descriptive approaches, developmental research, and clinical case studies to an unprecedented degree.3 This work culminated in the inclusion of the borderline concept as Borderline Personality Disorder in the official manual, DSM-III, of the American Psychiatric Association (Akhtar 1992). Thus in only 40 years this bland and nondescript word, “borderline,” had journeyed all the way from Stern’s account of a few difficult patients in his Manhattan practice to an established diagnostic category covering hundreds of thousands of outpatients in every private office, mental-health clinic, and training facility in the country.
By reformulating the borderline concept, Kernberg helped to bring about two changes in American psychoanalysis. First, ego psychology, which had drifted into claiming the status of a science of normal development and psychosocial adaptation, was invigorated and enriched by new insights into preoedipal development. Second, the Kleinian contribution to object-relations theory was introduced into the broad mainstream of psychoanalytic discourse. Although many psychoanalytic theorists in the United States had been aware of and involved in object-relations theory and its clinical applications, their work was in general seen as antithetical to the canons of ego psychology. While retaining the economic metapsychological hypothesis based on drive theory, Kernberg initiated an effort at integration of the two approaches by addressing himself to the diagnostic obscurities and the treatment impasses of the borderline patient, and by working ceaselessly on the theoretical interface—that is, on the structural derivatives of object relations.
By the early ’80s, however, a number of authors had become critical of Kemberg’s ideas. Some felt that his formulations have an unqualified specificity, whereas borderline patients represent such a heterogeneous group that no developmental defect, nuclear conflict, or structural abnormality can serve as the central point around which an explanatory theory can be woven. To aim at precision and systematization was thought to be counteranalytic. Other critics remarked that the structural interview, with its use of clarification, confrontation, and early interpretation of the transference, disturbs the establishment of a genuine therapeutic relationship from which a true psychoanalytic understanding can evolve (Weinshel 1988).
But the most frequent criticism came from everyday clinical experience. Many psychoanalysts and psychoanalytically oriented psychotherapists thought that most borderline patients could not tolerate the techniques embedded in Kernberg’s approach: strict limit setting, technical neutrality, and systematic analysis of the archaic, primarily negative transference and the defenses against envy. They maintained that instead of approaching these patients as possessing an archaic defensive structure designed to protect them from intrapsychic conflict, one could as well understand their current predicament and their past histories not as defenses, but as the outcome of developmental failures resulting from phase-specific vicissitudes in the experience of the mother–toddler relationship. According to these critics, the therapeutic relationship should have a more reparative role aimed at building, strengthening, and consolidating the missing structure.
Margaret Mahler made the most diligent effort to describe this structural defect by relating it to the separation-individuation process. She and her co-workers proposed that failure in the rapprochement subphase produces a relatively unassimilated bad introject around which the child’s inner experience is organized. Specifically, it is the upsurge of aggression at this time of the separation-individuation process, maintained by a lack of optimal emotional availability on the part of the mother, that provides the conditions for the organization of the borderline intrapsychic economy. Early ambivalence is not overcome, object constancy is not achieved, and the individual depends throughout life on external objects to provide himself with a sense of cohesive self-experience.
The hypothesis of a link between a circumscribed developmental failure and a specific form of psychopathology is always enticing but undoubtedly simplistic. Such a position ignores the fact that psychological development is subject to many elaborations of a progressive or a regressive nature. For example, we know that perplexities over separation and individuation are critically reworked in adolescence and early adulthood. Mahler (1971) herself eventually came to believe that there was no direct connection between the findings of child observation and subsequent borderline pathology.
Mahler’s emphasis on the centrality of the rapprochement crisis for the development of object constancy and whole self- and object-representations served as the basis for many variations. In one of these, Adler and Buie (1979) brought together Piaget’s findings on the achievement of evocative memory in the toddler, Fraiberg’s thoughts on mental representation, and Winnicott’s concept of the holding environment. Around the age of a year and a half the child’s mind gains the ability to evoke the memory of an object despite the absence of visual cues. From the inanimate object, he or she progresses to the memory of the human object, the mother. The ability of the child to evoke (remember) the image and the associated ministrations of the mother enables him or her to overcome the devastating sense of aloneness and panic that results from separations. The central deficit that leads to borderline pathology, then, is a deficit in positive soothing introjects stemming from the failure to achieve solid evocative memory, which in turn is a result of inadequate maternal holding experiences.
Adler and Buie’s hypothesis rests on a primary inner representational deficit; this calls into question Kernberg’s emphasis on splitting, since in their view there is no good object representation to begin with, one to be contrasted with the bad representation and actively kept apart from it through splitting. Their approach addresses itself to a subset of inner experiences and behaviors of the borderline centering on separation issues. It attempts to explain the craving that borderlines have for intense attachments and their tendency to react with fragmentation, terror of abandonment, and painful aloneness every time a disruption of an important relationship feels imminent. The goal of psychoanalytic psychotherapy, accordingly, is to provide a relationship of safety, security, and trust in order to enable the patient, through optimal disillusionment, to internalize the therapist as a positive introject and to develop the capacity for evocative memory, a developmental advance that will permit the patient to overcome the state of aloneness, panic, rage, and emptiness.
This general line of thought, which highlights the external world, puts the emphasis on the crucial role of the mother’s availability and her attunement to the child’s phase-appropriate needs. In its de-emphasis of instinctual aggression, unconscious fantasy, and primitive defenses, it came to be known as the theory of psychological deficit. It was supported and facilitated by the emerging presence on the American psychoanalytic scene of the psychology of the self, a new and controversial set of propositions regarding psychological development and its theorization.
Heinz Kohut originally developed his ideas in order to attend and respond to extremely sensitive patients suffering from serious problems of self-esteem regulation. He felt that these patients developed different types of transferences that demanded more empathy and less, if any, interpretation of defenses (Kohut 1971). As the psychology of the self gradually developed its own theoretical postulates and terminology and its own operational concepts, it expanded its theoretical and technical claims beyond the narcissistic characters to the realm of the borderline disorders and the neuroses (Kohut 1977).
The psychology of the self alters all previous psychoanalytic concepts: it drops the economic point of view and puts the emphasis on the structure of the self, including subjective conscious and preconscious experiences—continuing into adolescence and adult life—of selfhood and the self in relation to sustaining selfobjects. This self is not involved in instinctual expression but seeks relatedness. Drives, therefore, are seen as the breakdown products of an endangered or fragmenting self.
The psychology of the self recognizes the individual’s need to organize his or her psyche into a cohesive configuration, the self, as the most fundamental essence of human psychology. It also recognizes the need for the establishment of relationships between the self and its surroundings, relationships that evoke, maintain, and strengthen the structural coherence and vitality of the self and the harmony among its components. The primary means of data collection becomes sustained, empathic introspective immersion in the patient’s subjective world, as reflected in the transference: what Kohut calls listening from the patient’s frame of reference. The individual’s sense of compromise or failure is thought to stem from defects in the self brought about by unempathic responses, during childhood, on the part of selfobject figures.
To a certain extent the emergence of the psychology of the self and its distinctive role in American psychoanalysis can be understood in the context of the earlier intellectual monopoly of ego psychology. For almost thirty years object-relations theory—that spectrum of psychological explanations that are based on the premise that the self is composed of elements taken in from the outside, primarily aspects of the functioning of other persons—was not treated as an equal partner by the ego psychologists. No genuine cross-fertilization occurred. This was reflected in the teaching and the practice of psychoanalysis: defense interpretation and technical neutrality were put in the center of clinical inquiry, while the patient’s subjectivity was assigned a peripheral role.
Since the thinking behind object-relations theory was not permitted to become gradually integrated with ego psychology, it was inevitable that it would “march in” in the form of a fully articulated theory aimed at reversing and replacing all previous concepts. It is my hypothesis that, had the writings of Fairbairn, Winnicott, Guntrip, and Balint been read at the time they were published with the same open-mindedness and commitment that greeted the writings of Hartmann, there would not have been a psychology of the self, at least not in the form we know today. (Parenthetically, a frequent criticism raised against Kohut is that he did not sufficiently acknowledge his indebtedness to Fairbairn and Winnicott.)
According to the psychology of the self, borderline states involve serious, permanent, or protracted enfeeblement of, and damage to, the self, but complex defenses cover this basic deficit and protect the individual from close relationships that might activate the underlying fragmentation. The damage to the self is caused by severe and ongoing failures in parental empathic response to the selfobject needs of the child. Because of the constant threat of loss of cohesion, the child cannot undertake attachments to selfobjects; he or she remains with a chronic and overwhelming sense of dread which contributes to problems in self-regulation, self-control, self-soothing, and maintenance of self-esteem. In later life he or she may resort to compensatory stimulating activities such as drug abuse, indiscriminate sexuality, eating disorders, and so forth.
Kohut at first thought that the borderline individual could not form stable selfobject transferences (the expression of the need for idealizing, mirroring, or twinship in the therapeutic relationship). He later moderated this position and stated his belief that some borderlines with an unusually empathic, dedicated, and facilitative analyst could come to understand both their selfobject needs for affirmation, admiration, and soothing and their parents’ genetically pathogenic responses, and could be transformed into narcissistic personalities (Kohut 1984).
Kohut’s cautious position was modified to the extreme by Stolorow and colleagues (1987), who argued that the borderline condition is totally iatrogenic, brought about by insensitive therapeutic techniques. They claimed that the borderline structure (or the borderline personality organization) does not exist; primitive defensive operations (splitting, projective identification, etc.) are nothing but indications of the failure of archaic selfobject relations.
From this brief survey of the development of the borderline concept in American psychoanalysis, it is clear that no theory generates a satisfactory account of all aspects of the condition. Each approach tends to describe and explain best one or another aspect of the syndrome while failing to account equally well for other aspects. Each one starts out from a different reference point within the spectrum of psychoanalytic models, and its explanatory power varies accordingly.
It is possible that what we call borderline pathology is untreatable by psychoanalysis. The pathology is rooted in the pre-verbal world and involves deficits in symbolically encoded ideation, whereas psychoanalysis is essentially a verbally mediated communicative effort presupposing at least a minimal development of symbolic thought. What is more certain is that, as far as pathogenesis is concerned, all the divergent theoretical approaches, no matter how elaborate they are, need to anchor themselves on one or the other side of the radical nature-nurture dichotomy. Thus Kernberg’s construct assumes pregenital oral aggression; Kohut’s focuses on the selfobject.
There are several early French psychoanalytic contributions to the definition of the concept of the borderline.
In one of his late publications Charles Odier (1948), a pioneer of the Société Psychanalytique de Paris and the Swiss Psychoanalytic Society, described a central clinical characteristic of borderline pathology—fear of aloneness. The third chapter of his 1948 book L’angoisse et la pensée magique (Anxiety and Magical Thinking), is entitled “La névrose d’abandon” (“Abandonment Neurosis”). This was to become the title of a book by the Piaget-trained Swiss psychologist Germaine Guex (1950). In both publications, the extensive description and elaboration of the fear of abandonment is remarkable, considering that they predate by thirty years the work of Adler, Buie, and Kohut. As if to suggest a new nosological group, in the manner of hysterics or phobics, these patients are called les abandonniques; it is assumed that there is no evocative memory of the soothing introject, and technical recommendations stress the primacy of validation of the patient over insight.
In the work of that most creative French psychoanalyst, Maurice Bouvet (see especially his paper on “La relation d’objet” in Bouvet 1967), we encounter the description of the neurosis of depersonalization as contrasted with the depersonalization crises of neurotics, and the description of a pregenital object relationship distinct from those of the psychotic and the neurotic. Bouvet clearly anticipates later notions such as Kernberg’s borderline personality organization.
Finally, the work of Jean Bergeret is characterized by typically French psychostructural thinking. For him, the borderline state is a disorder of narcissism in which the object relationship remains anaclitic. Bergeret recognizes the existence of two authentic developmental lines of psychic structure, that, once in motion, lead to either a neurotic structure or a psychotic one; each is independent of the other and there is no intercommunication. Between these domains lie the borderline states, less rigid, less definitive, less solid, and much more mobile.
According to Bergeret (1993), when the ego, after having surpassed early frustrations and prepsychotic fixations, is ready to be engaged in normal oedipal evolution, it undergoes a significant psychological traumatization. This traumatization is experienced as a precocious, massive, and brutal encounter with the givens of the oedipal situation. In order to maintain its narcissistic equilibrium, the ego seeks to integrate the experience along the lines of early frustrations and threats. The child can no longer rely on the love of the father in order to feel supported against the threat of its hostile feelings toward the mother, and vice versa. The ego becomes obliged to resort to other defensive maneuvers such as denial, projective identification, primitive idealization, and the like. As a result of this first psychic disorganizer, libidinal evolution stops, and the ego precociously enters a stage of compromised pseudolatency that lasts up to adulthood. This common origin of compromised pseudolatency (tronc commun aménagé) is not an authentic psychic structure, since the usual clinical criteria of fixity, stability, and specificity do not apply.
From then on, development varies enormously. If a second psychic disorganizer (postpartum depression, marriage, loss, social upheaval, accident) occurs in the life of the individual, then the provisional compromise falls apart and the borderline adopts the neurotic, the psychotic, or the psychosomatic solution. Without a second disorganizer there is a gradual evolution either towards perversion or towards chronic characterological disturbance. Some borderlines remain in the pseudolatency stage for all their lives, decompensating in old age or unexpectedly resorting to suicide.
For reasons of space I can only mention by name some of the more recent French contributions to the subject of borderline pathology: André Green (1977), Janine Chasseguet-Smirgel (1978), and Joyce McDougall (1979).
Adler, G., and Buie, D. H. (1979). Aloneness and borderline psychopathology: developmental issues. International Journal of Psycho-Analysis 60:83–96.
Akhtar, S. (1992). Broken Structures: Severe Personality Disorders and Their Treatment. Northvale, NJ: Jason Aronson.
Bergeret, J. (1993). Les états limites et leurs aménagements. In Psychologie Pathologique, ed. J. Bergeret. Paris: Masson.
Bleuler, E. (1911). Dementia Praecox or the Group of Schizophrenias. New York: International Universities Press.
Bouvet, M. (1967). Oeuvres Psychanalytiques. Paris: Payot.
Chasseguet-Smirgel, J. (1978). Reflections on the connexions between perversion and sadism. International Journal of Psycho-Analysis 59:27–35.
Deutsch, H. (1942). Some forms of emotional disturbance and their relationship to schizophrenia. Psychoanalytic Quarterly 11:301–321.
Diggins, J. P. (1993). The Promise of Pragmatism. Chicago: University of Chicago Press.
Erikson, E. H. (1950). Childhood and Society. New York: Norton.
Falret, J. (1890). Etudes cliniques sur les maladies mentales. Paris: Bailliere.
Freud, S. (1919). On the teaching of psychoanalysis in universities. Standard Edition 17:169–174.
Frosch, J. (1964). The psychotic character: clinical psychiatric considerations. Psychoanalytic Quarterly 38:91–96.
Green, A. (1977). The borderline concept. In Borderline Personality Disorders, ed. P. Hartocollis, pp. 15–44. New York: International Universities Press.
Greenacre, P. (1941). The predisposition to anxiety. Part II. Psychoanalytic Quarterly 10:610–638.
Grinker, R., Werble, B., and Drye, R. (1968). The Borderline Syndrome: A Behavioral Study of Ego Functions. New York: Basic Books.
Guex, G. (1950). La névrose d’abandon. Paris: Presses Universitaires de France.
Gunderson, J. G., and Singer, M. (1975). Defining borderline patients. American Journal of Psychiatry 133:1–10.
Hoch, P., and Polatin, P. (1949). Pseudoneurotic forms of schizophrenia. Psychoanalytic Quarterly 23:248–276.
Jaspers, K. (1913/1923/1946). General Psychopathology. Manchester: University of Manchester Press, 1963.
Kernberg, O. (1967). Borderline personality organization. Journal of the American Psychoanalytic Association 15:641–685.
——— (1975). Borderline Conditions and Pathological Narcissism. New York: Jason Aronson.
——— (1984). Severe Personality Disorders. New Haven, CT: Yale University Press.
Kety, S. (1975). Mental illness in biological and adoptive families of adopted schizophrenics. In Genetic Research in Psychiatry, ed. R. R. Fieve et al., pp. 147–165. Baltimore: Johns Hopkins University Press.
Knight, R. (1953). Borderline states. Bulletin of the Menninger Clinic 17:1–12.
Kohut, H. (1971). The Analysis of the Self. New York: International Universities Press.
——— (1977). The Restoration of the Self. New York: International Universities Press.
——— (1984). How Does Analysis Cure? Chicago: University of Chicago Press.
Kraepelin, E. (1905). Einführung in die psychiatrische Klinik, 2nd. ed. Leipzig: Barth.
Kretschmer, E. (1925). Physique and Character. New York: Harcourt Brace.
Leichtman, M. (1989). Evolving concepts of borderline personality disorder. Bulletin of the Menninger Clinic 53:229–249.
Mahler. M. S. (1971). A study of the separation-individuation process and its possible application to borderline phenomena in the psychoanalytic situation. Psychoanalytic Study of the Child 26:402–424. New York: International Universities Press.
McDougall, J. (1979). Primitive communication in the use of countertransference. In Countertransference, ed. L. Epstein and A. Feiner, pp. 267–304. New York: Jason Aronson.
Meissner, W. W. (1984). The Borderline Spectrum. New York: Jason Aronson.
Odier, C. (1948). L’angoisse et la pensée magique. Neuchâtel: Delachaux & Niestle.
Perry, J. C., and Klerman, G. L. (1978). Clinical features of the borderline personality disorder. American Journal of Psychiatry 137:165–173.
Pinel, P. (1801). Abhandlung über Geistesverwirrungen oder Manie. Vienna: Carl Schaumburg.
Pritchard, J. C. (1835). Treatise on Insanity. London: Sherwood, Gilbert and Piper.
Reich, W. (1925). Der triebhafte Charakter. Leipzig: Internationaler Psychoanalytischer Verlag.
Schmideberg, M. (1947). The treatment of psychopathic and borderline patients. American Journal of Psychotherapy 1:45–71.
Schneider, K. (1923). Psychopathic Personalities. Springfield, IL: Charles C Thomas, 1958.
Spitzer, R. L., and Gibbon, M. (1979). Crossing the border into borderline personality and borderline schizophrenia: the development of criteria. Archives of General Psychiatry 36:17–24.
Stern, A. (1938). Psychoanalytic investigation and therapy in the borderline group of neuroses. Psychoanalytic Quarterly 7:467–489.
Stolorow, R., Brandchaft, B., and Atwood, G. (1987). Psychoanalytic Treatment. An lntersubjective Approach. Hillsdale, NJ: Analytic Press.
Stone, M. H. (1986). Essential Papers on Borderline Disorders. New York: New York University Press.
Weinshel, E. M. (1988). Borders of the borderline: modesty in psychoanalytic diagnosis. Psychoanalytic Inquiry 8:333–354.
Zilboorg, G. (1941). Ambulatory schizophrenia. Psychiatry 4:149–155.
1. There are two studies from the United States, one by Gregory Zilboorg (1941) on ambulatory schizophrenia and the other by Hoch and Polatin (1949) on pseudoneurotic schizophrenia. Both describe a diverse and unstable group of patients: some psychopaths or sexual perverts, others ineffective and unobtrusive people, with pananxiety, panneurosis and pansexuality; many were extremely sensitive to criticism and prone to rage. It is clear from the titles of their papers that both Zilboorg and Hoch and Polatin consider this diverse group to be a good-outcome variant of schizophrenia.
2. With the study of Roy Grinker and colleagues (1968) the empirical approach was given a new impetus and became more sophisticated; see Kety (1975), Gunderson and Singer (1975), Perry and Klerman (1978), Spitzer and Gibbon (1979).
3. At times, Kernberg’s comprehensiveness and articulateness can lead to the impression that his tightly woven conceptualizations are totally the product of his own inspiration. But the fact is that he has always expressed his indebtedness to Melanie Klein, Fairbairn, Bion, Mahler, and, above all, Edith Jacobson.