Cognitive-Behavioral Treatment of Borderline Personality Disorder
-Marsha M. Linehan

Chapter 2: Dialectical and Biosocial Underpinnings of Treatment


Every theory of personality functioning and of its disorders is based on some fundamental world view. Often this world view is left unspoken, and one has to read between the lines to figure it out. For example, Rogers's client-centered theory and therapy are based on the assumptions that people are fundamentally good and that they have an innate drive toward self-actualization. Freud assumed that individuals seek pleasure and avoid pain. He further assumed that all behavior is psychologically determined, and that there is no accidental behavior (behavior determined by accidental events of one's environment).

Similarly, DBT is based on a specific world view, that of dialectics. In this section, I provide an overview of what I mean by ``dialectics.'' I hope to show you that understanding this point of view is important and can enhance the ways of thinking about and interacting with borderline patients. I am not going to give a philosophical lecture on the meaning and history of the term, nor an in-depth coverage of current philosophical thinking in this area. Suffice it to say that dialectics is alive and well. Most people are aware of dialectics through the socioeconomic theory of Marx and Engels (1970). As a world view, however, dialectics also figures in theories of the development of science Kuhn, 1970), biological evolution (Levins & Lewontin, 1985), sexual relations (Firestone, 1970), and more recently the development of thinking in adults (Basseches, 1984). Wells (1972, cited in Kegan, 1982) has documented a shift toward dialectical approaches in almost every social and natural science during the last 150 years.

Why Dialectics?

The application of dialectics to my treatment approach began in the early 1980s with a series of therapy observations and discussions by my clinical research team. The team observed me in weekly therapy sessions while I attempted to apply to parasuicidal patients the cognitive--behavioral therapy I had learned at the State University of New York at Stony Brook under Gerald Davison and Marvin Goldfried. After each session, we would discuss both my behavior and that of the patient. At that time, the aims were to identify helpful techniques or, at a minimum, those that did not hamper therapeutic change and a positive working relationship. I was then to try to apply them in a consistent manner in future sessions. Subsequent discussions were aimed at keeping what was useful, discarding what was not, and developing behaviorally anchored descriptions of what exactly I as the therapist was doing.

A number of things happened during the course of treatment development. First, we verified that I could apply cognitive-behavioral therapy with this population; that was reassuring, since that was the primary intent of the project. However, as we observed what I was doing, it seemed that I was also applying a number of other procedures not traditionally associated with cognitive or behavior therapy. These techniques were things such as matter-of-fact exaggerations of the implications of events, similar to Whitaker's (1975, pp. 12-13); encouraging the acceptance rather than change of feelings and situations, in the tradition of Zen Buddhism (e.g., Watts, 1961); and double-bind statements such as those of the Bateson project directed at pathological behavior (Watzlawick, 1978). These techniques are more closely aligned with paradoxical therapy approaches than with standard cognitive and behavioral therapy. In addition, the pace of therapy seemed to include rapid changes in verbal style between, on the one hand, warm acceptance and empathetic reflection reminiscent of client-centered therapy, and, on the other hand, blunt, irreverent, confrontational comments. Movement and timing seemed as important as context and technique.

Although a colleague and I subsequently developed the relationship between DBT and paradoxical treatment strategies (Shearin & Linehan, 1989), when I was originally explicating the treatment I was reluctant to identify the approach with paradoxical procedures, because I was afraid that inexperienced therapists might overgeneralize from the ``paradoxical'' label and prescribe suicidal behavior itself; this was and is explicitly not done in the therapy. But I needed a label for the therapy. Clearly, it was not only standard cognitive--behavioral therapy. The emphasis at that time in cognitive therapy on rationality as the criterion of healthy thought seemed incompatible with my attention to intuitive and nonrational thought as equally advantageous. I was also becoming convinced that the problems of these patients did not result primarily from cognitive distortions of themselves and their environment, even though distortions seemed to play an important role in maintaining problems once they began. My focus in much of treatment on accepting painful emotional states and problematic environmental events seemed different from the usual cognitive-behavioral approach of trying to change or modify painful emotional states or act on environments to change them.

I began to think of ``dialectical'' as a descriptor of the therapy because of my intuitive experience in conducting therapy with this population of severely disturbed, chronically suicidal patients. The experience can best be described in terms of an image. It is as if the patient and I are on opposite ends of a teeter-totter; we are connected to each other by the board of the teeter-totter. Therapy is the process of going up and down, each of us sliding back and forth on the teeter-totter, trying to balance it so that we can get to the middle together and climb up to a higher level, so to speak. This higher level, representing growth and development, can be thought of as a synthesis of the preceding level. Then the process begins again. We are on a new teeter-totter trying to get to the middle in an effort to move to the next level, and so on. In the process, as the patient is continually moving back and forth on the teeter-totter, from the end toward the middle and from the middle back toward the end, I move also, trying to maintain a balance.

The difficulty in treating a suicidal borderline patient is that instead of on a teeter-totter, we are actually balanced on a bamboo pole perched precariously on a high wire stretched over the Grand Canyon. Thus, when the patient moves backward on the pole, if I move backward to gain balance, and then the patient moves backward again to regain balance, and so forth, we are in danger of falling into the canyon. (The pole is not infinitely long.) Thus, it seems that my task as the therapist is not only to maintain the balance, but to maintain it in such a way that both of us move to the middle rather than back off the ends of the pole. Very rapid movement and countermovement of the therapist seem to constitute a central part of the treatment.

The tensions that I experienced during therapy; the need to move to balance or synthesis with this patient population; and the treatment strategies reminiscent of paradoxical techniques that seemed a necessary adjunct to standard behavioral techniques--all these led me to the study of dialectical philosophy as a possible organizing theory or point of view.1 Dialectically speaking, the ends of the teeter-totter represent the opposites (``thesis'' and ``antithesis''); moving to the middle and up to the next level of the teeter-totter represents the integration or ``synthesis'' of these opposites, which immediately dissolves into opposites once again. This psychotherapeutic relationship between the opposites embodied in the term ``dialectics'' has been regularly pointed out since the early writings of Freud (Seltzer, 1986).

However serendipitous the original choice of a label was, the movement to a dialectical view subsequently guided the therapy development in a much broader fashion than would have been possible with just a paradoxical twist to techniques. Consequently, the treatment has evolved into its form of the past few years as an interaction between therapy process and dialectical theory. Over time, the term ``dialectics'' as applied to behavior therapy has come to imply two contexts of usage: that of the fundamental nature of reality and that of persuasive dialogue and relationship. As a world view or philosophical position, dialectics forms the basis of the therapeutic approach presented in this book. Alternatively, as a form of dialogue and relationship, dialectics refers to the treatment approach or strategies used by the therapist to effect change. Thus, central to DBT are a number of therapeutic dialectical strategies; these are described in Chapter 8.

Dialectical World View

A dialectical perspective on the nature of reality and human behavior has three primary characteristics.

The Principle of Interrelatedness and Wholeness

First, dialectics stresses interrelatedness and wholeness. Dialectics assumes a systems perspective on reality. The analysis of parts of a system is of limited value unless the analysis clearly relates the part to the whole. Thus. identity itself is relational, and boundaries between parts are temporary and exist only in relation to the whole; indeed, it is the whole that determines the boundaries. Levins and Lewontin (1985) state this well:
    Parts and wholes evolve in consequence of their relationship, and the relationship itself evolves. These are the properties of things that we call dialectical: that one thing cannot exist without the other, that one acquires its properties from its relation to the other, that the properties of both evolve as a consequence of their interpretation. (p. 3)
This holistic view is compatible with both feminist and contextual views of psychopathology. Such a perspective, when applied to treatment of BPD made me question the importance given to separation, differentiation, individuation, and independence in Western cultural thought. Notions of the individual as unitary and separate have only gradually emerged over the last several hundred years (Baumeister, 1987; Sampson, 1988). Since women receive the diagnosis of BPD much more frequently than men, the influence of gender on notions of self and appropriate interpersonal boundaries is of particular interest in our thinking about the disorder.

Both gender and social class significantly influence how one defines and experiences the self. Women, as well as other individuals with less social power, are more likely to have a relational or social self (a self that includes the group) as opposed to an individuated self (one that excludes the group) (McGuire & McGuire, 1982; Pratt, Pancer, Hunsberger, & Manchester, 1990). The importance of a relational or social self among women has been highlighted by many feminist writers, the best-known of whom is Gilligan (1982). Lykes (1985) has perhaps argued the feminist position most cogently in defining ``the self as an ensemble of social relations'' (p. 364). It is very important to note that Lykes and others do not speak simply of the value of interdependence among autonomous selves. Rather, they describe a social or relational self that is itself ``a coacting network of relationships embedded in an intricate system of social exchanges and obligations'' (Lykes, 1985, p. 362). When the self is defined as ``in relation,'' inclusive of others in its very definition, no fully separate self exists--that is, no self separated from the whole. Such a relational self, or ``ensembled individualism'' in Sampson's terms, characterizes the majority of societies, both historically and cross-culturally (Sampson, 1988).

Attention to these contextual factors is particularly essential when a cultural construct such as ``self'' is employed to explain and describe another cultural construct such as ``mental health.'' While the traditional definition of self may generally prove adaptive for some individuals in Western society, one must consider that our definitions and theories are not universal but are products of Western society, and thus may prove inappropriate for many individuals. As Heidi Heard and I have argued elsewhere (Heard & Linehan, 1993), and as I discuss later in this chapter and in Chapter 3, the problems encountered by the borderline individual may result in part from the collision of a relational self with a society that recognizes and rewards only the individuated self.

The Principle of Polarity

Second, reality is not static, but is comprised of internal opposing forces ``thesis'' and ``antithesis''), out of whose integration (``synthesis'') evolves a new set of opposing forces. Although dialectics focuses on the whole, it also emphasizes the complexity of any whole. Thus, within each one thing or system, no matter how small, there is polarity. In physics, for example, no matter how hard physicists try to find the single particle or element that is the basis of all existence, they always end up with an element that can be further reduced. In the single atom there is a negative and a positive charge; for each force, there is a counterforce; even the smallest element of matter is balanced by anti-matter.

A very important dialectical idea is that all propositions contain within them their own oppositions. Or, as Goldberg (1980) put it,
    I assume that truth is paradoxical, that each article of wisdom contains within it its own contradictions, that truths stand side by side. Contradictory truths do not necessarily cancel each other out or dominate each other, but stand side by side, inviting participation and experimentation. (pp. 295-296)
If you take this idea seriously, it can have a rather profound impact on your clinical practice. For example, in most descriptions of BPD, the emphasis is on identifying the pathology that sets the individual apart from others. Treatment is then designed to ferret out the pathology and create conditions for change. A dialectical perspective, however, suggests that within dysfunction there is also function; that within distortion there is accuracy; and that within destruction one can find construction. It was turning this idea around-"contradictions within wisdom" to "wisdom within contradictions"--that led me to a number of decisions about the form of DBT. Instead of searching for the validity of the patient's current behavior in the learning of the past, I began to search for and find it in the current moment. Thus, the idea took me a step beyond simply empathizing with the patient. Validation is now a crucial part of DBT.

The same idea led me to the construct of "wise mind," which is a focus on the inherent wisdom of patients. DBT assumes that each individual is capable of wisdom with respect to her own life, although this capability is not always obvious or even accessible. Thus, the DBT therapist trusts that the patient has within herself all of the potential that is necessary for change. The essential elements for growth are already present in the current situation. The acorn is the tree. Within the DBT case consultation team, the idea led to the emphases on finding the value in each person's point of view, rather than defending the value of one's own position.

Thesis, Antithesis, Synthesis:
The Principle of Continuous Change

Finally, the interconnected, oppositional, and nonreducible nature of reality leads to a wholeness continually in the process of change. It is the tension between the thesis and antithesis forces within each system (positive and negative, good and bad, children and parents, patient and therapist, person and environment, etc.) that produces change. The new state following change (the synthesis), however, is also comprised of polar forces; and, thus, change is continuous. The principle of dialectical change is important to keep in mind, even though I use these terms (``thesis,'' ``antithesis,'' ``synthesis'') rarely.

Change, then (or ``process,'' if you will), rather than structure or content, is the essential nature of life. Robert Kegan (1982) captures this point of view in his description of the evolution of self as a process of transformations over one's lifespan, generated by tensions between self-preservation and self-transformation within the person and within the person-environment system punctuated by temporary truces or developmental balances. He writes:
    As it is to understand the way the person creates the world, we must also understand the way the world creates the person. In considering where a person is in his or her evolutionary balancing we are looking not only at how meaning is made; we are looking too, at the possibility of the person losing this balance. We are looking, in each balance, at a new sense of what is ultimate and what is ultimately at stake. We are looking, in each new balance, at a new vulnerability. Each balance suggests how the person is composed, but each suggests, too, a new way for the person to lose her composure. (p. 114)
A dialectical point of view is quite compatible with psychodynamic theory, which stresses the inherent role of conflict and opposition in the process of growth and change. It is also compatible with a behavioral perspective which stresses the inherent wholeness of the environment and individual, and the interrelatedness of each in producing change. Dialectics as a theory of change is somewhat different from the self-actualizing notion of development assumed by client-centered therapy. In that perspective, each thing has within it a potentiality that will unfold naturally throughout its lifetime. ``Unfolding'' does not imply the tension inherent in dialectical growth. It is this tension that produces gradual change, punctuated by spurts of sudden shifts and dramatic movement.

In DBT, the therapist channels change in the patient, while at the same time recognizing that the change engendered is also transforming the therapy and the therapist. Thus, there is an ever-present dialectical tension within therapy itself between the process of change and the outcome of change. At each moment, there is a temporary balance between the patient's attempts to maintain herself as she is without changing, and her attempts to change herself regardless of the constraints of her history and current situation. The transition to each new temporary stability is often experienced as a painful crisis. "Any real resolution of the crisis must ultimately involve a new way of being in the world. Yet the resistance to doing so is great, and will not occur in the absence of repeated and varied encounters in natural experience" (Kegan, 1982, p. 41). The therapist helps the patient resolve crises by supporting simultaneously her attempts at self-preservation and at self-transformation. Control and direction channel the patient toward increased self-control and self-direction. Nurturing stands side by side with teaching the patient to care for herself.

Dialectical Persuasion

From the point of view of dialogue and relationship, "dialectics" refers to change by persuasion and by making use of the oppositions inherent in the therapeutic relationship, rather than by formal impersonal logic. Thus, unlike analytical thinking, dialectics is personal, taking into account and affecting the total person. It is an approach to engaging a person in dialogue so that movement can be made. Through the therapeutic opposition of contradictory positions, both patient and therapist can arrive at new meanings within old meanings, moving closer to the essence of the subject under consideration.

As noted above, the synthesis in a dialectic contains elements of both the thesis and antithesis, so that neither of the original positions can be regarded as ``absolutely true.'' The synthesis, however, always suggests a new antithesis and thus acts as a new thesis. Truth, therefore, is neither absolute nor relative; rather, it evolves, develops, and is constructed over time. From the dialectical perspective, nothing is self-evident, and nothing stands apart from anything else as unrelated knowledge. The spirit of a dialectical point of view is never to accept a final truth or an undisputable fact. Thus, the question addressed by both patient and therapist is "What is being left out of our understanding?"

I do not mean to imply that a sentence such as "It is raining and it is not raining" embodies a dialectic. Nor am I suggesting that a statement cannot be wrong, or not factual in a particular context. False dichotomies and false dialectics can occur. However, in these cases the thesis and/or antithesis has been misidentified, and thus one does not have a genuine antagonism. For example, a common statement during the Vietnam War, "Love it or leave it," was a classic case of a misidentification of the dialectic.

As I discuss in Chapters 4 and 13, dialectical dialogue is also very important in therapy team meetings. Perhaps more than any other factor, attention to dialectics can reduce the chances of staff splitting in treating borderline patients. Splitting among staff members almost always results from a conclusion by one or more factions within the staff that they (and sometimes they alone) have a ``lock'' on the truth about a particular patient or clinical problem.

Borderline Personality Disorder as Dialectical Failure

In some ways, borderline behaviors can be viewed as results of dialectical failures.

Borderline "Splitting"

As discussed in Chapter 1, borderline and suicidal individuals frequently vacillate between rigidly held yet contradictory points of view, and are unable to move forward to a synthesis of the two positions. They tend to see reality in polarized categories of ``either-or,'' rather than ``all,'' and within a very fixed frame of reference. For example, it is not uncommon for a such individuals to believe that the smallest fault makes it impossible for a person to be "good" inside. Their rigid cognitive style further limits their ability to entertain ideas of future change and transition, resulting in feelings of being in an interminable painful situation. Things once defined do not change. Once a person is "flawed," for instance, that person will remain flawed forever.

Such thinking among borderline individuals has been labeled "splitting" by psychoanalysts, and it forms an important part of psychoanalytic theory on BPD (Kernberg, 1984). Dichotomous thinking or splitting can be viewed as the tendency to get stuck in either the thesis or the antithesis, unable to move toward synthesis. An inability to believe that both a proposition (e.g., "I want to live") and its opposite ("I want to die") can be simultaneously true characterizes the suicidal and borderline individual. Splitting, from a psychodynamic point of view, is a product of the irresolvable conflict between intense negative and positive emotions.

From the dialectical perspective, however, conflict that is maintained is a dialectical failure. Instead of synthesis and transcendence, in the conflict typical of borderline individuals there is opposition between firmly rooted but contradictory positions, wishes, points of view, and so on. The resolution of conflict requires first the recognition of the polarities and then the ability to rise above them, so to speak, seeing the apparently paradoxical reality of both and neither. At the level of synthesis and integration that occurs when polarity is transcended, the seeming paradox resolves itself.

Difficulties with Self and Identity

Borderline individuals are frequently confused about their own identity, and tend to scan the environment for guidelines on how to be and what to think and feel. Such confusion can arise from a failure to experience their essential relatedness with other people, as well as the relationship of this moment to other moments in time. They are forever on the edge of the abyss, so to speak. Without these relational experiences, identity becomes defined in terms of each current moment and interaction experienced in isolation, and thus is variable and unpredictable rather than stable. In addition, there is no other moment in time to modulate the impact of the current moment. For a borderline patient, another person's anger at her in a particular interaction is not buffered by either other relationships where people are not angry or other points in time when this person is not angry at her. "You are angry at me" becomes infinite reality. The part becomes the whole. A number of other theorists have pointed out the important role of memory for affective events (Lumsden, 1991), especially interpersonal events (Adler, 1985), in the development and maintenance of BPD. Mark Williams (1991) has made a similar argument with respect to failures in autobiographical memory. Clearly, prior events and relationships must be available to memory if they are to buffer and be integrated within the present.

Interpersonal Isolation and Alienation

The dialectical perspective on unity presupposes that individuals are not separate from their environment. Isolation, alienation, feelings of being out of contact or not fitting in--all characteristic feelings of borderline individuals--are dialectical failures coming from the individuals setting up of a self-other opposition. Such an opposition can occur even in the absence of an adequate sense of self-identity. Often among borderline individuals, a sense of unity and integration is sought by suppression and/or nondevelopment of self-identity (beliefs, likes, desires, attitudes, independent skills, etc.), rather than by the dialectical strategy of synthesis and transcendence. The paradox that one can be different but at the same time part of the whole is not grasped. The opposition between person (part) and environment (whole) is maintained.

Case Conceptualization:
A Dialectical Cognitive Behavioral Approach

Case conceptualization in DBT is guided both by dialectics and by the assumptions of cognitive-behavioral theory. In this section, I review several characteristics of cognitive-behavioral theory that are important to DBT; I also suggest how a dialectical cognitive-behavioral approach differs somewhat from more traditional cognitive, behavioral, and biological theories. More specific theoretical points are reviewed as they relate to the specific DBT intervention strategies.

The Definition of "Behavior"

"Behavior, as used by cognitive-behavioral therapists, is a very broad term. It includes any activity, functioning, or reaction of the person--that is, "anything that an organism does involving action and response to stimulation" (Merriam-Webster Dictionary, 1977, p. 100). Physicists are using the term similarly when they speak of the behavior of a molecule; likewise systems analysts speak of the behavior of a system. Human behavior can be overt, (i.e., public and observable to others) or covert (i.e., private and observable only to the person behaving). In turn, covert behaviors may occur inside the person's body (e.g., stomach muscles tightening) or outside the body but nonetheless private, (e.g., behavior when a person is alone).

The Three Modes of Behavior

Contemporary cognitive-behavioral therapists typically categorize behavior into one of three modes: motoric, cognitive-verbal, and physiological. Motor behaviors are what most people think of as behavior; they include overt and covert actions and movements of the skeletal muscular system. Cognitive-verbal behavior includes such activities as thinking, problem solving, perceiving, imaging, speaking, writing, and gestural communication, as well as observational behavior (e.g., attending, orienting, recalling, and reviewing). Physiological behaviors include activities of the nervous system, glands, and smooth muscles. Although usually covert (e.g., heartbeat), physiological behaviors can also be overt (e.g., blushing and crying).

A number of things are important to note here. First, dividing behaviors into categories or modes is intrinsically arbitrary and is done for the convenience of the observer. Human functioning is continuous, and any response involves the total human system. Even partially independent behavioral subsystems share neural circuits and interconnecting neural pathways. However, behavioral systems that in nature do not occur separately are nonetheless often distinguished conceptually, because the distinction provides some increase in our ability to analyze the processes in question.

Emotions as Full-System Responses

Emotions, from the present perspective, are integrated responses of the total system. Generally, the form of the integration is automatic, either because of biological hard-wiring (the basic emotions) or because of repeated experiences (learned emotions). That is, an emotion typically comprises behaviors from each of the three subsystems. For example, basic researchers define emotions as comprised of phenomenological experience (cognitive system), biochemical changes (physiological system), and expressive and action tendencies (physiological plus motor systems). Complex emotions might also include one or more appraisal activities (cognitive system). Emotions, in turn, usually have important consequences for subsequent cognitive, physiological, and motor behavior. Thus, emotions not only are full-system behavioral responses, but themselves affect the full system. The complex, systemic nature of emotions makes it unlikely that any unique precursor of emotion dysregulation, either in general or with particular respect to BPD, will be found. There are many roads to Rome.

Intrinsic Equality of Behavioral Modes as Causes of Functioning

In contrast to biological psychiatry and cognitive psychology, the position taken here is that no mode of behavior is intrinsically more important than the others as a cause of human functioning. Thus, in contrast to cognitive theories (e.g., Beck, 1976, Beck et al., 1973, 1990), DBT does not view behavioral dysfunction, including emotion dysregulation, as necessarily resulting from dysfunctional cognitive processes. This is not to say that under some conditions cognitive activities do not influence motor and physiological behaviors, as well as the activation of emotional behaviors; in fact, a wealth of data suggests that the opposite is the case. Close to the topic of this book, for example, are the repeated findings of Aaron Beck and his colleagues (Beck, Brown, & Steer, 1989; Beck, Steer, Kovacs, & Garrison, 1985) that hopeless expectations about the future predict subsequent suicidal behaviors.

Moreover, in contrast to biological psychology and psychiatry, DBT does not view neurophysiological dysfunctions as intrinsically more important influences on behavior than other avenues of influence. Thus, from my perspective, although behavior-behavior or response system-response system relationships and causal pathways are important in human functioning, they are not more influential than any other pathways. The crucial question becomes this: Under what conditions does one behavior or behavioral pattern occur and influence another (Hayes, Kohlenberg, & Melancon, 1989)? Ultimately, however, from a dialectical framework, simple linear causal patterns of behavioral influence are not sought. Rather, the important question is more like that suggested by Manicas and Secord (1983): What is the nature of a given organism or process under prevailing circumstances? From this perspective, events, including behavioral events, are always the outcome of complex causal configurations at the same and at many different levels.

The Individual Environment System:
A Transactional Mode

A number of etiological models of psychopathology have been offered in the literature. Most current theories are based on some version of an interaction model, in which characteristics of the individual interact with characteristics of the environment to produce an effect--in this case, psychological disorder. The ``diathesis-stress model'' is by far the most general and ubiquitous interactive model. This model suggests that a psychological disorder is the result of a disorder-specific predisposition toward disease (the diathesis), which is expressed under conditions of general or specific environmental stress. The term ``diathesis'' generally refers to a constitutional or biological predisposition, but more modern usage includes any individual characteristic that increases a person's chance of developing a disorder. Given a certain amount of stress (i.e., noxious or unpleasant environmental stimuli), the individual develops the diathesis-linked disorder. The person is not equipped to cope with such stress, and thus behavioral functioning disintegrates.

In contrast, a dialectical or transactional model assumes that individual functioning and environmental conditions are mutually and continuously interactive, reciprocal, and interdependent. Within social learning theory, this is the principle of ``reciprocal determinism'': The environment and the individual adapt to and influence each other. Although the individual is surely affected by the environment, the environment is also affected by the individual. It is conceptually convenient to distinguish the environment from the individual person, but in reality they cannot be distinguished. The individual-environment is a whole system, defined by and defining the constituent parts. Because influence is reciprocal, it is transactional rather than interactional.

Chess and Thomas (1986) have written extensively about this pattern of reciprocal influence with respect to the effects of different temperamental characteristics of children on their family environments, and vice versa. Their notion of "poorness of fit" as an important factor in the etiology of psychological dysfunction has heavily influenced the theory proposed here. I discuss these ideas more fully later in the chapter.

Besides focusing on reciprocal influence, a transactional view also highlights the constant state of flux and change of the individual-environment system. Thomas and Chess (1985) have labeled such a model "homeodynamic," in contrast to interactive models that conceptualize the end state of individuals and environments as some sort of ``homeostatic'' equilibrium. A homeodynamic model is also dialectical. They quote from Sameroff (1975, p. 290), who makes this point very well:
    [The interactive model] is insufficient to facilitate our understanding of the actual mechanisms leading to later outcomes. The major reason behind the inadequacy of this model is that neither constitution nor environment are necessarily constant over time. At each moment, month, or year the characteristics of both the child and his [sic] environment change in important ways. Moreover, these differences are interdependent and change as a function of their mutual influence on one another.
Millon (1987a) has made much the same point in discussing the etiology of BPD and the futility of attempting to locate the "cause" of the disorder in any single event or time period.

A transactional model highlights a number of points that are easy to overlook in a diathesis-stress model. For example, people in a particular environment may act in a manner that is stressful to an individual in it only because the environment itself was exposed to the stress that this individual placed on it. Examples of such individuals include the child who, due to sickness, requires expenditure of much of the family's financial resources, or the psychiatric patient who uses up much of the inpatient nursing resources because of the need for constant suicide precautions. Both of these individuals' environments are stretched in their ability to respond well to further stress; other people in both environments may invalidate or temporarily blame the victim if any further demand on the system is made. Although the system (e.g., the child's family) may have been predisposed to respond dysfunctionally in any case, it may have avoided such responses if it had not been exposed to the stress of that particular individual.

A transactional model does not assume necessarily equal power of influence on both sides of the equation. For example, some genetic influences can be powerful enough to overwhelm a benign or even a healing environment. Current research suggests a much greater influence of genetic heritage on even normal adult personality characteristics than was previously believed (Scarr & McCartney, 1983; Tellegen et al., 1988). Nor can we discount the influence of a powerful situation on the behavior of most individuals exposed to the situation, despite large, pre-existing individual personality differences (Milgram, 1963, 1964). Any person, no matter how hardy, who is exposed repeatedly to violent sexual or physical abuse will be harmed.

A Visual Representation of an Environment Person System

A visual representation of an environment-person system is shown in Figure 2.1. I developed the particular model shown here a number of years ago to capture the data on suicidal and parasuicidal behavior. To the left is a box representing the environmental subsystem. Although in this scheme the environment is represented as four-cornered, this is done only for theoretical purposes relevant to suicidal behavior. Depending on the particular environmental factors believed to be important in an event or behavior pattern under study, one could represent the environment with as many sides as there are factors in the theory.

The person is subdivided into two separate subsystems. The behavioral subsystem is a triangle representing the three modes of behavior described above. The circular arrows at each point of the triangle indicate that responses within each behavioral mode are self-regulatory, in that changes in one response effect changes in another. Interestingly, although this aspect of behavior is well studied for physiological responses, corresponding attention has not been paid to how the motor behavioral and the cognitive verbal response modes self-regulate.

The second triangle represents stable, organismic characteristics of the person that are not typically influenced by either the individual's behavior or the environment. These stable characteristics can, however, have important influences on both the environment and the behavior of the individual. In the model represented here, the triangular points represent gender, race, and age. As with the environmental square, however, these points are simply conceptually convenient. Gender, race, and age are related in important ways to suicidal behaviors. Other disorders will require representation of different organismic variables. For example, in the study of schizophrenia, one might want an organismic point representing genetic makeup.

Biosocial Theory: A Dialectical Theory of Borderline Personality Disorder Development


DBT is based on a biosocial theory of personality functioning. The major premise is that BPD is primarily a dysfunction of the emotion regulation system; it results from biological irregularities combined with certain dysfunctional environments, as well as from their interaction and transaction over time. The characteristics associated with BPD (see Chapter 1, especially Tables 1.2 and 1.5) are sequelae of, and thus secondary to, this fundamental emotion dysregulation. Moreover, these same patterns cause further deregulation. Invalidating environments during childhood contribute to the development of emotion dysregulation; they also fail to teach the child how to label and regulate arousal, how to tolerate emotional distress, and when to trust her own emotional responses as reflections of valid interpretations of events. As adults, borderline individuals adopt the characteristics of the invalidating environment. Thus, they tend to invalidate their own emotional experiences, look to others for accurate reflections of external reality, and oversimplify the ease of solving life's problems. This oversimplification leads inevitably to unrealistic goals, an inability to use reward instead of punishment for small steps toward final goals, and self-hate following failure to achieve these goals. The shame reaction--a characteristic response to uncontrollable and negative emotions among borderline individuals--is a natural result of a social environment that ``shames'' those who express emotional vulnerability.

As noted in Chapter 1 in a slightly different context, the formulation proposed here is similar to that of Grotstein et al. (1987), who have proposed that BPD is a disorder of self-regulation. By this they mean that the disorder represents a primary breakdown of the regulation of states of self, such as arousal, attention, sleep, wakefulness, self-esteem, affects, and needs, together with the secondary sequelae of such a breakdown. As Grotstein et al. have noted, few theories of BPD have integrated biological and psychological factors into a coherent theory. To date, most theories have been either squarely psychological, whether psychoanalytic (e.g., Adler, 1985; Masterson, 1972, 1976; Kernberg, 1975, 1976; Rinsley, 1980a, 1980b; Meissner, 1984) or cognitive behavioral (e.g., Beck et al., 1990; Young, 1987; Pretzer, in press); or they have been products of biological psychiatry (e.g., Klein, 1977; Cowdry & Gardner, 1988; Akiskal, 1981, 1983; Wender & Klein, 1981). Grotstein's (1987) formulation is a wedding of biological psychiatry and psychoanalytically informed psychological theory. Stone (1987) has suggested a similar integration. He nicely describes the difficulty of becoming well versed in the two broad areas of psychology and biology and integrating them into a theoretical position on BPD as approximating "in complexity the task of translating a text composed, perversely, of Arabic words alternating with Chinese" (pp. 253 254).

The biosocial formulation presented here is based primarily on the experimental literature in psychology. What I have found in perusing this literature is that there is a wealth of basic empirical data on such diverse topics as personality and behavioral functioning, genetic and physiological bases of behavior and personality, temperament, basic emotional functioning, and environmental effects on behavior; however, with only a few exceptions (e.g., Costa and McCrae, 1986), there has been little attempt to apply this basic research literature in psychology to the understanding of personality disorders. This state of affairs probably exists because, until very recently, the empirical study of personality disorders has been done primarily by psychiatrists, whereas the empirical study of behavior per se (including the study of biological bases of behavior) has been the domain of psychologists. The gulf between these two fields has been large, with members of neither reading much of the literature in the other. Empirically based clinical psychology, which one could consider the natural bridge between the two disciplines, has until recently shown little or no interest in personality disorders.

Borderline Personality Disorder and Emotion Dysregulation

As I stated above, the biosocial theory is that BPD is primarily a disorder of the emotion regulation system. Emotion dysregulation, in turn, is due to high emotional vulnerability plus an inability to regulate emotions.4 The more emotionally vulnerable the individual is, the greater the need for emotion modulation. The thesis here is that borderline individuals are emotionally vulnerable as well as deficient in emotion modulation skills, and that these difficulties have their roots in biological predispositions, which are exacerbated by specific environmental experiences.

The premise of excessive emotional vulnerability fits empirical descriptions, developed in entirely separate research traditions, of both parasuicidal and borderline populations. I have reviewed this literature in Chapter 1. In summary, the emotional picture of both parasuicidal and borderline individuals is one of chronic, aversive affective experiences. Failures to inhibit maladaptive, mood-dependent actions are by definition part of the borderline syndrome. Discussions of affect dysregulation with respect to BPD usually concentrate on the depression mania continuum (e.g., Gunderson & Zanarini, 1989). In contrast, I am using ``affect'' here in a much more global sense, and suggest that borderline individuals have regulation difficulties across several (if not all) emotional response systems. Although it is likely that emotion dysregulation is most pronounced in negative emotions, borderline individuals also seem to have difficulty regulating positive emotions and their sequelae.

Emotional Vulnerability

Characteristics of emotional vulnerability include high sensitivity to emotional stimuli, emotional intensity, and slow return to emotional baseline. ``High sensitivity'' means that the individual reacts quickly and has a low threshold for an emotional reaction; that is, it does not take much to provoke an emotional reaction. Events that might not bother many people are likely to bother the emotionally vulnerable person. The sensitive child reacts emotionally to even slight frustration or disapproval. At the adult level, the therapist's leaving town for the weekend may elicit an emotional response from the borderline patient, but not from most other patients. The implications for psychotherapy are, I suspect, obvious. The feeling, noted frequently by therapists and families of borderline individuals, of having to "walk on eggs" is a result of this sensitivity.

"Emotional intensity" means that emotional reactions are extreme. Emotionally intense individuals are the dramatic people of the world. On the negative side, partings may precipitate very intense and painful grief; what would cause slight embarrassment for another may cause deep humiliation; annoyance may turn to rage; shame may develop from slight guilt; apprehension may escalate to a panic attack or incapaciting terror. On the positive side, emotionally intense individuals may be idealistic and likely to fall in love at the drop of a hat. They may experience joy more easily, and thus may also be more susceptible to spiritual experiences.

A number of investigators have found that increases in emotional arousal and intensity narrow attention, so that emotion-relevant stimuli become more salient and are more closely attended to (Easterbrook, 1959; Bahrick, Fitts, & Rankin, 1952; Bursill, 1958; Callaway & Stone, 1960; Cornsweet, 1969; McNamara & Fisch, 1964). The stronger the arousal and the greater the intensity, the narrower the attention becomes. Clinically, these phenomena seem exceptionally characteristic of borderline individuals. It is an important point to keep in mind, however, that these tendencies are not pathological per se; they are characteristic of any individual during extreme emotional arousal. The relative paucity of theory and research examining the emotions as antecedents of cognitions, compared to the large amount on cognitions as precursors to emotion, may be the consequence of our Western view of individual behavior as a product of the rational mind (Lewis, Wolan-Sullivan, & Michalson, 1984).

"Slow return to emotional baseline" means that reactions are long-asting. It is important to note here, however, that all emotions are relatively brief, lasting from seconds to minutes. What makes an emotion feel long-lasting is that emotional arousal, or mood, tends to have a pervasive effect on a number of cognitive processes, which in turn are related to the activation and reactivation of emotional states. Bower and his colleagues (Bower, 1981; Gilligan & Bower, 1984) have reviewed a large number of research studies indicating that emotional states (1) selectively bias the recall of affectively toned material, resulting in superior memory when the emotional state at recall matches the learning state; (2) enhance the learning of mood-congruent material; and (3) can bias interpretations, fantasies, projections, free associations, personal forecasts, and social judgments in a fashion congruent with current mood. Emotions may also be more self-perpetuating among borderline individuals because of the greater intensity of their emotional responses, as suggested above. With high emotional arousal, the environment (including the therapist's behavior) can be selectively attended to, so that actions and events consistent with the current primary mood are attended to and other aspects are neglected. The effect of mood on cognitive processes makes sense in view of the theory that emotions are full-system responses. A current emotion integrates the entire system in its favor. In some senses, it is rather surprising that any emotion ever ends, since emotions, once started, are repeatedly refired. A slow return to emotional baseline exacerbates this reactivating effect; it also contributes to high sensitivity to the next emotional stimulus. This characteristic can be very important in treatment. It is not unusual for a borderline patient to say that it takes several days to recover from a psychotherapy session.

Emotion Modulation

The research on emotional behavior suggests that emotion regulation requires two somewhat paradoxical strategies. The individual must first learn to experience and label the discrete emotions that are hard-wired into the neurophysiological, behavioral expressive, and sensory feeling systems. Then the individual must learn to reduce emotionally relevant stimuli that serve either to reactivate and augment ongoing negative emotions or to set off secondary dysfunctional emotional responses. Once an intense emotion is activated, the individual must be able to inhibit or interfere with the activation of mood-congruent afterimages, afterthoughts, afterappraisals, afterexpectations, and afteractions, so to speak.

Basic emotions are fleeting and generally adaptive (Ekman, Friesen, & Ellsworth, 1972; Buck, 1984). Constant inhibition or truncating of negative emotions seems to have a number of dysfunctional consequences. First, inhibition can lead to neglect of the problem situation instigating the emotion. An individual who never experiences anger in the face of injustice is less likely to remember unjust situations. Situations that are truly dangerous may not be avoided if fear is never experienced. Apologies may never be given and relationships may be left unrepaired when guilt or shame is always cut off before it can affect a person's behavior within a relationship. Second, the inhibition or truncating of negative emotions serves to increase emotional avoidance. If the individual has learned a secondary emotional reaction to negative emotions, the inhibition of the original emotion removes any chance of relearning. The paradigm is similar to the escape-learning paradigm. Animals taught to escape from a chamber by having their feet shocked whenever they enter the chamber will cease to enter the chamber; if the shock apparatus is subsequently turned off, the animals will never learn the new contingencies. They must enter the chamber for new learning to occur. The invalidating family (which I describe later) is much like the shock apparatus in the escape-learning paradigm. Borderline individuals learn to avoid negative emotional cues; they become negative-emotion-phobic. Without experiencing the negative emotions, however, the individual fails to learn that she can tolerate the emotions and that punishment will not follow their expression.

Third, we simply do not know the outcomes of emotional inhibition and truncation over the long run. Research is desperately needed here. There is some evidence that emotional experiencing and catharsis lead to less stressful negative emotional states. There is also evidence that emotional catharsis increases emotionality rather than reduces it (see Bandura, 1973, for a review of this research). Under what conditions emotional experiencing enhances versus interferes with therapeutic progress is an important question that has not been adequately addressed.

John Gottman and Lynn Katz (1990) have outlined four emotion modulation activities or abilities. These include the abilities to (1) inhibit inappropriate behavior related to strong negative or positive affect, (2) self-regulate physiological arousal associated with affect, (3) refocus attention in the presence of strong affect, and (4) organize one self for coordinated action in the service of an external, non-mood-dependent goal. The principle of changing or modulating emotional experiences by changing or resisting emotion-linked behavior is one of the important principles underlying behavior therapy exposure techniques. Besides increasing emotionality directly, inappropriate, mood-dependent behavior usually leads to consequences that elicit other unwanted emotions. Coordinated action in the service of an external goal serves to keep life progressing forward. Thus, such behavior has the long-term potential to enhance positive emotions, decrease stress and thereby reduce vulnerability to emotionality. In addition, such action is the opposite of mood-dependent behavior, and thus is an instance of acting oneself into feeling different. I discuss these principles in some detail in Chapter 11.

Changing emotions by changing physiological arousal is the principle behind a number of therapeutic emotion change strategies, such as relaxation therapies (including desensitization), some medications, and breathing training in the treatment of panic. The ability to modify physiological arousal associated with affect means that the individual is able not only to reduce

the high arousal associated with some emotions, such as anger and fear (i.e., to calm down), but to increase the low arousal associated with other emotions, such as sadness and depression (i.e.. to ``rev up,'' so to speak). Usually, this will require the ability to force activity, even when the person is not in the mood. For example, one of the basic techniques in cognitive therapy of depression is activity scheduling.

The important role of controlling attention as a way to regulate contact with emotional stimuli has been pointed out by many (e.g., Derryberry & Rothbart, 1984, 1988). Shifting attention toward a positive stimulus can enhance or maintain ongoing positive arousal and emotion; shifting it away from a negative stimulus may attenuate or contain negative arousal and emotion. Thus, individuals with control over attention focusing and attention shifting two related but distinct processes (Posner, Walker, Friedrich, & Rafal, 1984) have an advantage in regulation of emotional responses. In turn, individual differences in attention control are evident from the earliest years of life (Rothbart & Derryberry, 1981) and appear as stable temperamental characteristics in adults (Keele & Hawkins, 1982; Derryberry, 1987; MacLeod, Mathews, & Tata, 1986). This point is particularly interesting, in light of data reviewed by Nolen-Hoeksema (1987) suggesting gender differences in attentional response sets under stress. She concludes that, at least when depressed, women have a more ruminative response set than men. Rumination about one's current depressed mood, in turn, generates depressing explanations that increase depression further and lead to greater helplessness on future tasks (Diener & Dweck, 1978). In contrast, men are more likely to engage in distracting behaviors that dampen depressed mood. It seems reasonable to hypothesize that an inability to distract oneself from negative, emotionally sensitive stimuli may be an important part of the emotion dysregulation found among borderline individuals.

Biological Underpinnings

The mechanisms of emotion dysregulation in BPD are unclear, but difficulties in limbic system reactivity and attention control may be important. The emotion regulation system is a complex one, and there is no a priori reason to expect that the dysfunction will be the result of a common factor in all borderline individuals. Biological causes could conceivably range from genetic influences to disadvantageous intrauterine events to early childhood environmental effects on development of the brain and nervous system.

Cowdry et al. (1985) report data suggesting that some borderline individuals may have a low threshold for activation of limbic structures, the brain system associated with emotion regulation. In particular, they note the overlap among symptoms of complex partial seizures, episodic dyscontrol, and BPD. Positive benefits among borderline individuals for ananticonvulsant (carbamazepine) whose neurophysiological effects are known to be located in the limbic area lends further support to this notion (Gardner & Cowdry, 1986, 1988).

Other investigators have reported that patients with BPD have significantly more electroencephalographic (EEG) dysrhythmias than their depressed control patients (Snyder & Pitts, 1984; Cowdry et al., 1985). Andrulonis and his colleagues (Andrulonis et al., 1981; Akiskal et al., 1985a, 1985b) have attempted to link neurologically based dysfunctions to BPD. However, they did not employ comparison groups, and thus it is difficult to interpret their findings. In contrast, Cornelius et al. (1989) reviewed a number of studies in which borderline patients were compared with patients exhibiting various other psychiatric disorders. Generally, they reported no EEG differences; no differences in familial mental retardation, epilepsy, or neurological disorders; no differences on a broad battery of tests assessing major areas of cognitive functioning; and no differences in overall neurodevelopmental histories. Interestingly, Cornelius et al. did report data indicating the early onset of borderline-type behavior patterns among borderline patients. For example, childhood temper tantrums and persistent rocking or head banging were more frequent among children later diagnosed as having BPD than among those later diagnosed as depressed or schizophrenic.

Still another research strategy attempting to locate biological influences on behavior is the comparison of various behavioral dysfunctions in family members of the population of interest. Studies of first-degree relatives of borderline patients have found higher prevalences of affective disorder (Akiskal, 1981; Andrulonis et al., 1981; Baron, Gruen, Asnis & Lord, 1985; Loranger, Oldham, & Tulis, 1982; Pope et al., 1983; Schulz et al., 1986; Soloff & Millward, 1983; Stone, 1981), of closely related personality traits such as histrionic and antisocial characteristics (Links, Steiner, & Huxley, 1988; Loranger et al., 1982; Pope et al., 1983; Silverman et al., 1987), and of borderline personality disorder (Zanarini, Gunderson, Marino, Schwartz & Frankenburg, 1988) than among relatives of control groups. However, many other investigators have failed to find similar associations when all relevant characteristics have been controlled (see Dahl, 1990, for a review of this literature). A twin study by Torgersen (1984) supports a psychosocial over a genetic model of transmission. There has been little or no research attempting to link temperamental characteristics of borderline individuals to data on the genetic and biological etiology of those particular temperamental attributes. Such research is sorely needed.

Factors other than genes, however, may be equally important in determining neurophysiological functioning, especially in the emotion regulation system. We know, for example, that characteristics of the intrauterine environment can be crucial in the development of the fetus. Furthermore, these characteristics influence later behavioral patterns of the individual. Just a few examples will make my point here. Fetal alcohol syndrome, characterized by mental retardation and hyperactivity, impulsiveness, distractibility, irritability, delayed development, and sleep disorders, is caused by maternal ingestion of excessive alcohol (Abel, 1981, 1982). Similar dysfunctions are regularly noted in babies of drug-addicted mothers (Howard, 1989). There is accumulating evidence that environmental stress experienced by the mother during pregnancy can have deleterious effects on the later development of the child (Davids & Devault, 1962; Newton, 1988).

Postnatal experiences can also have important biological consequences. It has been well established that radical environmental events and conditions can modify neural structures (Dennenberg, 1981; Greenough, 1977). There is little reason to doubt that neural structures and functions related to emotional behaviors are similarly affected by experiences with the environment (see Malatesta & Izard, 1984 for a review). The relationship of environmental trauma to emotion regulation is particularly salient in the case of BPD given the prevalence of childhood sexual abuse among this population--a topic I discuss later in this chapter.

Borderline Personality Disorder and Invalidating Environments

The temperamental picture of the borderline adult is quite similar to that of the "difficult child" described by Thomas and Chess (1985). From their studies of temperamental characteristics of infants, they identified difficult children as the ``group with irregularity in biological functions, negative withdrawal responses to new stimuli, non-adaptability or slow adaptability to change, and intense mood expressions that are frequently negative'' (p. 219). In their research, this group comprised approximately 10% of their sample. Clearly, however, not all children with a difficult temperament grow up to meet criteria for BPD. Although the majority (70%) of difficult children studied by Chess and Thomas (1986) had behavior disorders during childhood, most of these children improved or recovered by adolescence. In addition, as Chess and Thomas point out, children who originally do not have a difficult temperament may acquire one as they develop.

Thomas and Chess have suggested that the "goodness of fit" or "poorness of fit" of the child with the environment is crucial for understanding later behavioral functioning. Goodness of fit results when the properties of the child's environment and its expectations and demands are in accord with the individual's own capacities, characteristics, and style of behavior. Optimal development and behavioral functioning are the results. In contrast, poorness of fit results when there are discrepancies and dissonances between environmental opportunities and demands and the capacities and characteristics of the child. In these instances, distorted development and maladaptive functions result (Thomas & Chess, 1977, Chess & Thomas, 1986). It is this notion of ``poorness of fit'' that I propose as crucial for understanding the development of BPD. But what kind of environment would constitute a "poor fit" leading to this particular disorder? I propose that an "invalidating environment" is most likely to facilitate development of BPD.

Characteristics of Invalidating Environments

An invalidating environment is one in which communication of private experiences is met by erratic, inappropriate, and extreme responses. In other words, the expression of private experiences is not validated; instead, it is often punished, and/or trivialized. The experience of painful emotions, as well as the factors that to the emotional person seem causally related to the emotional distress, are disregarded. The individual's interpretations of her own behavior, including the experience of the intents and motivations associated with behavior, are dismissed.

Invalidation has two primary characteristics. First, it tells the individual that she is wrong in both her description and her analyses of her own experiences, particularly in her views of what is causing her own emotions, be-liefs, and actions. Second, it attributes her experiences to socially unacceptable characteristics or personality traits. The environment may insist that the individual feels what she says she does not (``You are angry, but you just won't admit it''), likes or prefers what she says she does not (the proverbial ``When she says no, she means yes''), or has done what she said she did not. Negative emotional expressions may be attributed to traits such as overreactivity, oversensitivity, paranoia, a distorted view of events, or failure to adopt a positive attitude. Behaviors that have unintended negative or painful consequences for others may be attributed to hostile or manipulative motives. Failure, or any deviation from socially defined success, is labeled as resulting from lack of motivation, lack of discipline, not trying hard enough, or the like. Positive emotional expressions, beliefs, and action plans may be similarly invalidated by being attributed to lack of discrimination, naivete, overidealization, or immaturity. In any case, the individual's private experiences and emotional expressions are not viewed as valid responses to events.

Emotionally invalidating environments are generally intolerant of displays of negative affect, at least when such displays are not accompanied by public events supporting the emotion. The attitude communicated is similar to the ``you can pull yourself up by the bootstraps'' approach; it is the belief that any individual who tries hard enough can make it. Individual mastery and achievement are highly valued, at least with respect to controlling emotional expressiveness and limiting demands on the environment. Invalidating members of such environments are often vigorous in promulgating their point of view and actively communicate frustration with an individual's inability to adhere to a similar point of view. Great value is attached to being happy, or at least grinning in the face of adversity; to believing in one's capacity to achieve any objective, or at least never ``giving in'' to hopelessness; and, most of all, to the power of a ``positive mental attitude'' in overcoming any problem. Failures to live up to these expectations lead to disapproval, criticism, and attempts on the part of others to bring about or force a change of attitude. Demands that a person can place on these environments are usually very restricted.

This pattern is very similar to the pattern of high ``expressed emotion,'' found in the families of both depressives and schizophrenics with high relapse rates (Leff & Vaughn, 1985). The work with expressed emotion suggests that such a family constellation can be extremely powerful with the vulnerable individual. ``Expressed emotion,'' in that literature, refers to criticism and over involvement. The notion here includes those two aspects, but in addition stresses a nonrecognition of the actual state of the individual. The consequence is that the behaviors of others, including caregivers, in the individual's environment are not only invalidating of the individual's experiences but also nonresponsive to the needs of the individual.

A few clinical examples may provide a better idea of what I mean here. During a family session with a borderline woman who had a history of alcoholism and frequent serious suicide attempts, her son commented that he just didn't understand why she couldn't let problems ``roll off her back'' as he, his brother, and his father did. A substantial number of patients in my research project were actively dissuaded from going into psychotherapy by their parents. One 18-year-old patient who had been hospitalized several times, had a history of numerous attempts to harm herself, was hyperactive and dyslexic, and was heavily involved in the drug culture was told weekly by her parents after her group therapy sessions that she did not need therapy and that she could just straighten up on her own if she really wanted to. ``Talking about problems just makes problems worse,'' her father said. Another patient was told while growing up that if she cried when she got hurt playing, her mother would give her a ``real'' reason to cry: If the tears continued, her mother would hit her.

Consequences of Invalidating Environments

The consequences of invalidating environments are as follows. First, by failing to validate emotional expression, an invalidating environment does not teach the child to label private experiences, including emotions, in a manner normative in her larger social community for the same or similar experiences. Nor is the child taught to modulate emotional arousal. Because the problems of the emotionally vulnerable child are not recognized, little effort goes into attempts to solve the problems. The child is told to control her emotions, rather than being taught exactly how to do that. It is a bit like telling a child with no legs to walk without providing artificial legs for her to walk on. The nonacceptance or oversimplification of the original problems precludes the type of attention, support, and diligent training such an individual needs. Thus, the child does not learn to adequately label or control emotional reactions.

Second, by oversimplifying the ease of solving life's problems, the environment does not teach the child to tolerate distress or to form realistic goals and expectations.

Third, within an invalidating environment, extreme emotional displays and/or extreme problems are often necessary to provoke a helpful environmental response. Thus, the social contingencies favor the development of extreme emotional reactions. By erratically punishing communication of negative emotions and intermittently reinforcing displays of extreme or escalated emotions, the environment teaches the child to oscillate between emotional inhibition on the one hand, and extreme emotional states on the other.

Finally, such an environment fails to teach the child when to trust her own emotional and cognitive responses as reflections of valid interpretations of individual and situational events. Instead, the invalidating environment teaches the child to actively invalidate her own experiences and to search her social environment for cues about how to think, feel, and act. A person's ability to trust herself, at least minimally, is crucial; she at least has to trust her decision not to trust herself. Thus, invalidation is ordinarily experienced as aversive. People who are invalidated will usually either leave the invalidating environment, attempt to change their behavior so that it meets the expectations of their environment, or try to prove themselves valid and thereby to reduce the environment's invalidation. The borderline dilemma arises when the individual cannot leave the environment and is unsuccessful at changing either the environment or her own behavior to meet the environment's demands.

It might perhaps seems that such an environment would produce an adult with dependent personality disorder instead of BPD. I suspect that such an outcome would be likely with a less emotionally vulnerable child. But with an emotionally intense child, the invalidating information coming in from the environment is almost always competing with an equally strong message from the child's emotional responses: ``You may be telling me that what you did was an act of love, but my hurt feelings, terror, and rage tell me that it wasn't loving. You may be telling me that I can do it; and it's no big deal, but my panic is saying that I cannot and it is.''

The emotionally vulnerable, invalidated individual is in a bind similar to that of the overweight individual in our society. The culture (including daily weight reduction ads on TV and radio) and thin family members repeatedly tell the obese person that losing weight is easy; and keeping it off requires just a little will power. A body weight over the cultural ideal is thought to be the mark of a gluttonous, lazy, or undisciplined person. A thousand diets, intense hunger while dieting, herculean efforts to get and stay thin, and a body, that regains weight at the drop of a calorie say otherwise. How does the heavy person respond to this double message? Usually by alternating between dieting and extreme discipline on the one hand, and giving in, relaxing, and refusing to diet on the other. The yo-yo syndrome among dieters is similar to the emotional oscillation among borderline individuals. Neither source of information can be comfortably ignored.

Varieties of Sexism: Prototypic Invalidating Experiences

The prevalence of BPD among women requires that we examine the possible role of sexism in its etiology. Certainly, sexism is an important source of invalidation for all women in our culture; just as certainly, all women do not become borderline. Nor do all women with vulnerable temperaments become borderline, even though all women are exposed to sexism in one form or another. I suspect that the influence of sexism in the etiology of BPD depends on other characteristics of the vulnerable child, as well as on the circumstances of sexism in the family raising the child.

Sexual Abuse. The most extreme form of sexism is, of course, sexual abuse. The risk for sexual abuse is approximately two to three times greater for females than for males (Finkelhor, 1979). The prevalence of childhood sexual abuse in the histories of women meeting criteria for BPD is such that it simply cannot be ignored as an important factor in the etiology of the disorder. Of 12 hospitalized borderline patients assessed by Stone (1981), 9, or 75%, reported a history of incest. Childhood sexual abuse was reported by 86% of borderline inpatients compared, to 34% of other psychiatric inpatients, in a study by Bryer, Nelson, Miller, and Krol (1987). Among borderline outpatients, from 67% to 76% report childhood sexual abuse (Herman, Perry, & van der Kolk, 1989; Wagner, Linehan, & Wasson, 1989), in contrast to a 26% rate among nonborderline patients (Herman et al., 1989). Ogata, Silk, Goodrich, Lohr, and Westen (1989) found that 71% of borderline patients reported a history of sexual abuse, compared to 22% of major depressive control patients.

Although in epidemiological data girls are at no higher risk for physical abuse than boys are, one study found rates of reported childhood physical abuse to be higher among borderline patients (71%) than among nonborderline patients (38%) (Herman et al., 1989). Furthermore, there is a positive association between physical and sexual abuse (Westen, Ludolph, Misle, Ruffin, & Block, 1990), suggesting that those at risk for sexual abuse are at higher risk for physical abuse also. Bryer et al. (1987), however, found that whereas early sexual abuse predicted the diagnosis of BPD, the combination of sexual and physical abuse did not. Ogata et al. (1989) also reported similar rates of physical abuse in borderline and depressed patients. Thus, it may be that sexual abuse, in contrast to other types of abuse, is uniquely associated with BPD. Much more research is needed here to clarify the relationships.

A very similar connection has been found between childhood sexual abuse and suicidal (including parasuicidal) behaviors. Victims of such abuse have higher rates of subsequent suicide attempts than nonvictims do (Edwall, Hoffmann, & Harrison, 1989; Herman & Hirschman, 1981; Briere & Runtz,

1986; Briere, 1988); up to 55% of these victims go on to attempt suicide. Furthermore, sexually abused women engage in more medically serious parasuicidal behavior (Wagner et al., 1989). Bryer et al. (1987) found that childhood abuse (both sexual and physical) predicted adult suicidal behavior. Individuals with suicide ideation or parasuicide were three times more likely to have been abused in childhood than were patients without such behaviors.

Although it is generally viewed as a social stressor, child abuse may play a less obvious role as a cause of physiological vulnerability to emotion dysregulation. Abuse may not only be pathogenic for individuals with vulnerable temperaments; it may ``create'' emotional vulnerability by affecting changes in the central nervous system. Shearer, Peters, Quaytman, and Ogden (1990) suggest that perpetual trauma may physiologically alter the limbic system. Thus, severe, chronic stress may have permanent adverse effects on arousal, emotional sensitivity, and other factors of temperament.

Sexual abuse, as it occurs in our culture, is perhaps one of the clearest examples of extreme invalidation during childhood. In the typical case scenario of sexual abuse, the victim is told that the molestation or intercourse is ``OK'' but that she must not tell anyone else. The abuse is seldom acknowledged by other family members, and if the child reports the abuse she risks being disbelieved or blamed (Tsai & Wagner, 1978). It is difficult to imagine a more invalidating experience for the typical child. Similarly, physical abuse is often presented to the child as an act of love or is otherwise normalized by the abusive adult. Some clinicians have suggested that the secrecy of sexual abuse may be the factor most related to subsequent BPD. Jacobson and Herald (1990) reported that of 18 psychiatric inpatients with histories of major childhood sexual abuse, 44 had never revealed the experience to anyone. Feelings of shame are common among sexual abuse victims (Edwall et al., 1989) and may account for this failure to disclose the abuse. We cannot exclude the invalidating component of sexual abuse as contributory to the BPD.

Parental Imitation of Infants. Parents' tendencies to imitate an infant's emotionally expressive behaviors constitute an important factor in optimal emotional development (Malatesta & Haviland, 1982). Failure to imitate or noncongruent imitation the former of which is failure to validate, and the latter of which is invalidation are related to less optimal development. Interestingly, with respect to gender differences in the incidence of BPD, mothers tend to show more contingent responding to sons' smiles than to daughters' smiles and imitate sons' expressions more often than daughters' (Malatesta & Haviland, 1982).5

Dependence and Independence: Invalidating (and Impossible) Cultural Ideals for Women. The research data are overwhelming in confirming large differences between male and female interpersonal relationship styles. Flaherty and Richman (1989) have reviewed extensive data in the areas of primate behavior and evolution, developmental studies, parenthood, and adult social support and mental health. They conclude that various socialization experiences, beginning at infancy, render women more affectively connected and perceptive in the interpersonal sphere than men. The relationship between receiving social support from others and personal well-being, and, conversely, the relationship between social support distress and somatic complaints, depression, and anxiety are stronger for females than for males. That is, whereas the degree of social support received is not closely related to emotional functioning among men, it is highly correlated to emotional well-being among women. In particular, Flaherty and Richman (1989) found that the intimacy component of social support is most closely associated with well-being among women. In reviewing research on assertion and women, Kelly Egan and I concluded that women's behavior in groups or dyads is consistent with an emphasis on maintaining relationships almost to the exclusion of achieving task objectives, such as solving problems or persuading others (Linehan & Egan, 1979).

Given the prevalence of interpersonal bonding and social support as important (indeed, crucial), dimensions for well-adjusted women, one can ask this question: What happens to women who either are not given the social support they need or are taught that their very need for social support is itself unhealthy? Just such situations seem to exist. Almost without exception, interpersonal independence for both males and females is extolled as the ideal of ``healthy'' behavior. Feminine characteristics such as interpersonal dependence and relying on others--which, as noted above, are positively related to women's mental health--are generally perceived as mentally ``unhealthy'' (Widiger & Settle, 1987). We so value independence that we apparently cannot conceive of the possibility that a person could have too much independence. For example, although there is a ``dependent personality disorder'' in the DSM-IV, there is no in ``independent personality disorder.''

This emphasis on individual independence as normative behavior is unique to, and pervasive in, Western culture (Miller, 1984; see Sampson, 1977, for a review of this literature). In fact, one can conclude that normative feminine behavior, at least that part having to do with interpersonal relationships, is in a collision with current Western cultural values. It is no wonder that many women come to experience conflict over issues of independence and dependence. Indeed, it appears that there is a ``poorness of fit'' between women's interpersonal style and Western socialization and cultural values for adult behavior. It is interesting, however, that the pathology is laid on the doorstep of the conflicted women, rather than on that of a society that seems to be moving further and further away from valuing community and interpersonal dependency.

Femininity and Bias. Sexism can be a special problem for those female children whose talents are those generally rewarded in men but often ignored or invalidated in women. For example, mechanical ability, sports achievements, interest in math and science, and logical, task-oriented thinking are valued more in men than in women. Any sense of pride or accomplishment can easily be invalidated in women with such characteristics. An even worse situation occurs when these talents valued in men are not matched by talents and interests valued in women (e.g., interest in appearing attractive, home-oriented skills). In such a situation, the female child is not rewarded for the talents that she does have, and in addition is punished for emitting ``unfeminine behaviors'' or failing to emit ``feminine'' behaviors. When the child's behaviors are tied to temperamental characteristics, she is in further trouble. For example, gentleness, softness, affection, responsiveness to others, empathy, nurturance and soothing, and similar characteristics are highly valued ``feminine'' associated characteristics (Widiger & Settle, 1987; Flaherty & Richman, 1989); however, they are not the characteristics associated with a difficult temperament.

For the female child punished for having characteristics that interfere with her meeting the cultural ideal for women, life must be particularly difficult when she has brothers who are not punished for identical behaviors or sisters who effortlessly meet standards for femininity. The injustice is not to be missed in these situations. The environment outside the home does little in these cases to ameliorate the problem, since the same values are held across the culture. It is difficult to imagine how such a child could not grow up believing that there must be something wrong with her.

In my clinical experience, just this state of affairs seems to be common among borderline patients. We have been struck in our clinic with the number of patients who are talented in areas valued highly in men but little in women, such as mechanical and intellectual pursuits. Our borderline group therapy is entirely female, and a frequent topic of discussion is the difficulties the patients experienced as children because their interests and talents appeared more masculine than feminine. Another common experience seems to have been growing up in families that valued the boys more than the girls, or at least gave them more leeway, more privileges, and less punishment for the behaviors that led the girls to grief. Although sexism is clearly a fact, its relationship to BPD as I have described here is just as clearly speculative. We simply need more research data on this point.

Types of Invalidating Families

My colleagues and I have observed three types of invalidating families among patients in our clinic: the ``chaotic'' family, the ``perfect'' family, and, less commonly, the ``typical'' family.

Chaotic Families. In the chaotic family, there may be problems with substance abuse, financial problems, or parents who are out of the home much of the time; in any case, little time or attention is given to the children. For example, the parents of one of my patients spent almost every afternoon and evening at a local tavern. The children came home from school each day to an empty house and were left to fend for themselves for dinner and structure in the evenings. Often they wandered over to a grandmother's for dinner. When the parents were home, they were volatile; the father was often drunk; and they could tolerate few demands from the children. Needs of the children in such a family are disregarded and consequently invalidated. Millon (1987a) has suggested that the increase in chaotic families may be responsible for the increase in BPD.

Perfect Families. In the ``perfect'' family, the parents for one reason or another cannot tolerate negative emotional displays from their children. Such a stance may be the result of a number of factors, including other demands on the parents (such as a large number of children or stressful jobs), an inability to tolerate negative affect, self-centeredness, or naive fears of spoiling a child with a difficult temperament. In my experience, when members of such a family are asked directly about their feeling toward the borderline family member, they express a great deal of sympathy. However, without meaning to, these other members often express consistent invalidating attitudes for example, expressing surprise that the borderline individual can't just ``control her feeling.'' One such family member suggested that his daughter's very serious problems would be cured if she just prayed more.

Typical Families. When I originally observed the invalidating environmental style, I called it the ``American way syndrome,'' since it is so prevalent in American culture. However, when I gave a lecture in Germany, my German colleagues informed me that I could have called it the ``German way syndrome.'' It is most likely a product of Western culture in general. A number of emotion theorists have commented on the tendency in Western societies to emphasize cognitive control of emotions and to focus on achievement and mastery as criteria of success. The individuated self in Western culture is defined by sharp boundaries between self and others. In cultures with this view, the behavior of mature persons is assumed to be controlled by internal rather than external forces. ``Self-control,'' in this context, refers to the people's ability to control their own behavior by utilizing internal cues and resources. To define oneself differently for example, to define the self in relation to others, or to be field-dependent is labeled as immature and pathological, or at least inimical to good health and smooth societal functioning (Perloff, 1987).

(Although this conception of the individual self pervades Western culture, it is universal neither cross-culturally nor even within Western culture itself.)

A key point must be kept in mind about the invalidating family. Within limits, an invalidating cognitive style is not detrimental for everyone or in all contexts. The emotion control strategies used by such a family may even be useful at times to the person who is temperamentally suited to them and who can learn attitude and emotional control. For example, research by Miller and associates (Efran, Chorney, Ascher, & Lukens, 1981; Lamping, Molinaro, & Stevenson, 1985; Miller, 1979; Miller & Managan, 1983; Phipps & Zinn, 1986) indicates that individuals who tend to psychologically ``blunt'' threat-relevant cues when faced with the prospect of uncontrollable aversive events show lower and less sustained physiological, subjective, and behavioral arousal than individuals who tend to monitor or attend to such cues. Knussen and Cunningham (1988) have reviewed research indicating that belief in one's own behavioral control over negative outcomes, instead of blaming others (a key belief in the invalidating family), is related to more favorable future outcomes in a variety of areas. Thus, cognitive control of emotion can be quite effective in certain circumstances. Indeed, this approach got the railroad across the United States, built the bomb, got many of us through school, and put up skyscrapers in big cities!

The only problem here is that the approach ``only works when it works.'' That is, telling persons who are capable of affect self-regulation to control their emotions is quite a different proposition from telling this to an individual who does not have this capability. For example, one mother I was working with who had a 14-year-old daughter with a ``difficult'' temperament and a 5-year-old daughter with an ``easy'' temperament. The older daughter had difficulty with anger, especially when her little sister was teasing her. I was trying to teach the mother to validate this daughter's emotional reactions. After the 5-year-old pushed a complex puzzle of the 14-year old's onto the floor, the older child screamed at her sister and stormed out of the room,

leaving the sister in tears. The mother happily reported that she had ``validated'' the older daughter's emotions by saying, ``Mary, I can understand why you got angry. But in the future, you have got to control your explosions!'' It was difficult for the mother to see how she had invalidated the daughter's difficulties in controlling her emotions. In the cases of emotionally reactive and vulnerable persons, invalidating environments vastly oversimplify these person's problems. What other people succeed in doing controlling emotions and emotional expression the borderline individual can often succeed at only sporadically.

Emotion Dysregulation and Invalidating Environments: A Transactional Vicious Cycle

A transactional analysis suggests that a system that may originally have consisted of a slightly vulnerable child within a slightly invalidating family can, over time, evolve into one in which the individual and the family environment are highly sensitive to, vulnerable to, and invalidating of each other. Chess and Thomas (1986) describe a number of ways in which the temperamental child, the slow-to-warm-up child, the distractible child, and the persistent child can overwhelm, threaten, and disorganize otherwise nurturing parents. Patterson (1976; Patterson & Stouthamer-Loeber, 1984) has also written extensively on the interactive behaviors of child and family that lead to mutually coercive behavior patterns on the part of all parties in the system. Over time, children and caregivers shape and reinforce extreme and coercive behaviors in each other. In turn, these coercive behaviors further exacerbate the invalidating and coercive system, leading to more, not fewer, dysfunctional behaviors within the entire system. One is reminded of a Biblical quotation: ``...for anyone who has will be given more; from anyone who has not, even what he thinks he has will be taken away'' (Luke 8:18; The Jerusalem Bible, 1966).

There is no question that an emotionally vulnerable child puts demands on the environment. Parents or other caregivers have to be more vigilant, more patient, more understanding and flexible, and more willing to put their own wishes for the child on temporary hold when these wishes exceed the child's capabilities. Unfortunately, what often happens is that the child's response to invalidation actually reinforces the family's invalidating behavior. Telling a child that her feelings are stupid or unwarranted does at times quiet the child down. Many people, including those with emotional vulnerability, sometimes withdraw and appear to feel better when their emotions are made light of. Invalidation is aversive, and thus suppresses the behavior it follows.

The ``controlling'' environment described by Chess and Thomas (1986) is a variation or extreme example of the invalidating environment described here. The controlling environment constantly shapes the child's behavior to fit the family's preferences and convenience rather than the child's short- and long-term needs. In that situation, of course, the validity of the child's behavior as it exists is not recognized. As the child matures, power struggles are inevitable, with the environment sometimes appeasing and giving in and at other times rigidly holding the line. Depending on the child's initial temperament, the eventual result of appeasement is a child tyrant, a child with negative passivity, or both. The manner of this development is described over and over again in manuals on parenting.

In essence, the error in such a family is twofold. First, the caregivers make an error in shaping. That is, they expect more or different behaviors than the child is capable of emitting. Excessive punishment and insufficient modeling, instructing, coaching, cheerleading, and reinforcement follow. Such a pattern creates an aversive environment for the child, in which needed help is

not forthcoming and unavoidable punishment occurs. As a result, the child's negative emotional behaviors increase, including the expressive behaviors that are associated with the emotions. These behaviors function to terminate punishment, usually by creating such aversive consequences for the caregivers that they stop attempts at control.

And here caregivers make the second error: They reinforce the functional value of extreme expressive behaviors, and extinguish the functional value of moderate expressive behaviors. Such a pattern of appeasement following extreme emotional displays can unwittingly create the pattern of behaviors associated with BPD in the adult. When appeasement from others does not oc-cur, or occurs unpredictably, the unavoidability of aversive conditions mimicks the learned helplessness paradigm: Passive, helpless behaviors can be expected to increase. If passive or helpless behaviors are in turn punished, the person is faced with an unwinnable dilemma and will probably vacillate between extreme emotionally expressive behaviors and equally extreme passive and helpless behaviors. Such a state of affairs can, without too much difficulty, account for the emergence of many borderline characteristics as the child matures.

Emotion Dysregulation and Borderline Behaviors

Very little in human behavior is not affected by emotional arousal and mood states. Such diverse phenomena as concepts of the self, self-attributions, perceptions of control, learning of tasks and performance, patterns of self-reward, and delay of gratification are affected by emotional states (see Izard, Kagan, & Zajonc, 1984, and Garber & Dodge, 1991, for reviews). The thesis here is that most borderline behaviors are either attempts on the part of the individual to regulate intense affect or outcomes of emotion dysregulation. Emotion dysregulation is both the problem the individual is trying to solve and the source of additional problems. The relationship between borderline behavior patterns and emotion dysregulation is depicted in Figure 2.2.

Emotion Dysregulation and Impulsive Behaviors

Suicidal and other impulsive, dysfunctional behaviors are usually maladaptive solution behaviors to the problem of overwhelming, uncontrollable, intensely painful negative affect. Suicide, of course, is the ultimate way to change one's affective state (we presume). Other, less lethal (e.g., parasuicidal) behaviors, however, can also be quite effective. Overdosing, for example, usually leads to long periods of sleep; sleep, in turn, has an important influence on regulating emotional vulnerability. Cutting and burning the body also seem to have important affect-regulating properties. The exact mechanism here is unclear, but it is common for borderline individuals to report substantial relief from anxiety and a variety of other intense negative affective states following cutting themselves (Leinbenluft, Gardner, & Cowdry, 1987).

Suicidal behavior, including suicide threats and parsuicide, is also very effective in eliciting helping behaviors from the environment--help that may be effective in reducing the emotional pain. In many instances, in fact, such behavior is the only way an individual can get others to pay attention to and try to ameliorate her emotional pain. For example, suicidal behavior is a most effective way for a nonpsychotic individual to be admitted to an inpatient psychiatric unit. Many therapists tell their patients that they can or should phone them if they are feeling suicidal. The staff at a psychiatric inpatient unit in my area used to tell one of our patients that she could come right back in if she got ``command voices'' telling her to commit suicide. In our clinical population of parasuicidal borderline women, a majority report that the intent to change their environment is part of at least one instance of parasuicidal behavior.

Unfortunately, the instrumental character of suicide threats and parasuicide is frequently the most salient one for therapists and theorists working with borderline individuals. Thus, suicide attempts and other intentional self-injurious behaviors are often referred to as ``manipulative.'' The basis of this reference is usually a therapist's own feeling of being manipulated. As I have discussed in Chapter 1, however, it is a logical error to assume that if a behavior has a particular effect, the actor has therefore engaged in the behavior in order to bring about the effect. The labeling of suicidal behavior as manipulative, in the absence of an assessment of the actual intent of the behavior, can have extremely deleterious effects. This issue is discussed further in the Chapter 15 where I describe treatment strategies for suicidal behaviors.

Emotion Dysregulation and Identity Disturbance

Generally, people form a sense of self-identity through their own observations of themselves as well as through others' reactions to them. Emotional consistency and predictability across time and similar situations are prerequisites to this development of identity. All emotions involve some element of preference or approach-avoidance. A sense of identity, among other things, is contingent on preferring or liking something consistently. For example, a person who always enjoys drawing and painting may develop an image of herself that includes aspects of an artist's identity. Others observing this same preference may react to the person as an artist, further developing her image of herself. Unpredictable emotional lability, however, leads to unpredictable behavior and cognitive inconsistency; thus a stable self-concept, or sense of identity, fails to develop.

A tendency of borderline patients to inhibit, or attempt to inhibit, emotional responses may also contribute to an absence of a strong sense of identity. The numbness associated with inhibited affect is often experienced as emptiness, further contributing to an inadequate (and at times completely absent) sense of self. Similarly, if an individual's own sense of events is never ``correct'' or is unpredictably ``correct''--the situation in the invalidating family--then one would expect the individual to develop an overdependence upon others. This overdependence, especially when the dependence relates to preferences, ideas, and opinions, simply exacerbates problems with identity, and a vicious cycle is once again started.

Emotion Dysregulation and Interpersonal Chaos

Effective interpersonal relations are enormously benefited by both a stable sense of self and a capacity for spontaneity in emotional expression. Successful relationships also require a capacity to self-regulate emotions in appropriate ways, to control impulsive behavior, and to tolerate stimuli that produces pain to a certain degree. Without such capabilities, it is understandable that borderline individuals develop chaotic relationships. Difficulties with anger and anger expression, in particular, preclude the maintenance of stable relationships.

In addition, as I discuss further in Chapter 3, the combination of emotional vulnerability with an invalidating environment leads to the development of more intense and more persistent expressions of negative emotions. Essentially, the invalidating environment usually places the individual on an intermittent reinforcement schedule, in which expressions of intensely negative affect or demands for help are reinforced sporadically. Such a schedule is known to create very persistent behavior. When people currently involved with the borderline person also fall into the trap of inconsistently appeasing her--sometimes giving in to and reinforcing high-rate, high-intensity aversive emotional expressions and other times not doing so--they are recreating conditions for the person's learning of relationship-destructive behaviors.

Implications of the Biosocial Theory for Therapy with Borderline Patients

General Aims and Skills Taught

Recognition of these emotion regulation difficulties, originating in both biological makeup and inadequate learning experiences, suggests that treatment should focus on the twin tasks of teaching the borderline patient (1) to modulate extreme emotionality and reduce maladaptive mood-dependent behaviors, and (2) to trust and validate her own emotions, thoughts, and activities. The therapy should focus on skills training and behavior change, as well as on validation of the patient's current capabilities and behaviors.

A major portion of DBT is devoted to teaching just such skills. The skills are broken down into four types: (1) those that increase interpersonal effectiveness in conflict situations, and thus show promise in decreasing environmental stimuli associated with negative emotions; (2) strategies culled from the behavioral treatment literature on affective disorders (depression, anxiety, fear, anger) and posttraumatic stress, which increase self-regulation of unwanted emotions in the face of actual or perceived negative emotional stimuli; (3) skills for tolerating emotional distress until changes are forthcoming; and (4) skills adapted from Eastern (Zen) meditation techniques, such as mindfulness practice, which increase the ability to experience emotions and avoid emotional inhibition.

Avoiding ``Blaming the Victim''

The successful extinction of maladaptive, extreme emotional displays is contingent on a number of factors. Most importantly, a validating environment must be created that allows the therapist to extinguish maladaptive behaviors while at the same time soothing, comforting, and cajoling the patient through the experience. The process is tricky and requires an enormous amount of therapist tolerance, willingness to experience emotional pain, and flexibility. Often, however, in conducting therapy, therapists may apply to borderline patients the same expectations as those placed on other patients. When the borderline patients cannot meet these expectations, the therapists may be tolerant for a period. But as the patients' display of negative emotions increases, the therapists' patience or willingness to tolerate the pain they themselves are experiencing runs out, and they then appease, punish, or terminate therapy with these patients. Clinicians experienced in working with borderline patients have perhaps recognized themselves in the earlier descriptions of invalidating, controlling environments and of the families who get caught in the vicious cycle of appeasing and punishing these patients. Such an environment, when recapitulated in therapy, is simply a continuation of the invalidating environment that the patients have experienced throughout their lives.

A most typical form of punishment of borderline patients consists of behaviors that, in sum, are both invalidating of the patients and ``blaming the victims.'' Research in social psychology suggests that a number of factors are important in determining whether observers will blame victims of misfortune for their own misfortune. Relevant to the present topic are findings that in general, females are blamed more for misfortunes than are males in comparable situations (Howard, 1984). In the same research, Howard also found that when a victim is female, observers attribute blame to her character. However, when a victim is male, observers attribute blame to the male's behavior in the situation, not to his character. Other variables are also important: The observer has to care about the misfortune of the victim; the consequences have to be severe (Walster, 1966); and the observer has to feel helpless in controlling the outcome (Sacks & Bugental, 1987). Thus, when people care about what happens to others, they do not want these others to suffer, but they cannot keep misfortune or suffering from happening; they are likely to blame the victims for their own misfortune and suffering.

This is exactly the situation of therapy with most borderline patients. First, the ``victims'' are primarily women. Usually, their therapists care whether they are suffering. And certainly, few therapies to date have been shown to be particularly effective in stopping that suffering. Even if therapists believe that a particular treatment will be effective in the long run, because it has worked with other patients, helplessness in the face of the borderlines' intense suffering--suffering that causes the therapists reciprocal pain--is the repeated, day-to-day experience of working with these individuals. In the face of this helplessness the therapists may redouble their efforts. When the patients still do not improve, the therapists may begin to say that they are causing their own distress. The patients don't want to improve or change. They are resisting therapy. (After all, it works with almost everyone else.) They are playing games. They are too needy. In short, the therapists make a very fundamental but quite predictable cognitive error: They observe the consequence of behavior (e.g., emotional suffering for the patients or themselves) and attribute that consequence to internal motives on the part of the patients. I refer to this error repeatedly in further discussions of treatment of borderline patients.

"Blaming the victim" has important iatrogenic effects. First, it invalidates an individual's experience of her own problems. What the individual experiences as attempts to end pain are mislabeled as attempts to maintain the pain, to resist improving, or to do something else that the individual is not aware of. Thus, the individual learns to mistrust her own experience of herself. After some time, it is not unusual for the person to learn the point of view of the therapist, both because she does not trust her own self-observations and because doing so leads to more reinforcing outcomes. I once had a patient who was having immense trouble managing her homework practice; either she would not practice, or her practice attempts would not be successful. Simultaneously, she was repeatedly entreating me and my group coleader to help her feel better. One week, when I asked her what had interfered with her practicing her homework, she said with great conviction that she obviously did not want to be happy. If she did, she would have practiced her homework.

A key component of DBT is its insistence that the therapist refrain from blaming the victim for her own problems. This is not a position based on simple naivet%pge, although I have been accused of that. First, the caregiver's blaming of the victim usually leads to emotional distancing, negative emotions directed at the patient, decreased willingness to help, and punishment of the patient. Thus, the very help that is needed is more difficult to give. The caregiver becomes frustrated and often, but usually very subtly, strikes out at the patient. Because the punishment is not aimed at the actual source of the problem, it simply increases the patient's negative emotionality. A power struggle ensues--one that neither the patient nor the therapist can win.

Concluding Comments

It is important to keep in mind that the dialectical position presented here is a philosophical position. Thus, it can be neither proved nor disproved. For many, however, it is a difficult position to grasp. You may not see the need for it at first. Certainly, you can adopt some of DBT without necessarily embracing (or understanding) dialectics. If you are like me and my students, however, the idea will become more appealing over time and will subtly change your conceptualization of therapy issues. For me, it has had a profound effect on the way I conduct psychotherapy and the way I organize my treatment unit. DBT has been growing and changing continuously; the emerging implications of a dialectical perspective have been a source of much of the growth.

The biosocial theory I am presenting here is speculative. There has been little prospective research to document the application of this approach to the etiology of BPD. Although the theory is in accord with the known literature on BPD, no research has been mounted so far to test the theory prospectively. Thus, the reader should keep in mind that the logic of the biosocial formulation of BPD described in this chapter is based largely on clinical observation and speculation rather than on firm empirical experimentation. Caution is recommended.

Copyright © 1993, Guilford Publications, Inc. This excerpt is posted with permission of Guilford Publications, Inc. and is subject to copyright law and restricted from further use. No part of this excerpt may be reproduced, stored in a retrieval system or transmitted in any form or by any means (electronic, photocopying, recording or otherwise) without prior written permission of the publisher. To obtain permission please contact Guilford Publications, Inc. at the address below or e-mail: