TFP is based on object relations theory (Kernberg, 1975, 1976) and is designed to achieve change in the patient's representational world through the interpretation of the transference relationship with particular emphasis on the here and now (Clarkin, Yeomans, & Kernberg, 1999). This intense focus on the transference functions to consolidate an attachment relationship to the therapist, to elicit and work through the insecure attachment representations that tend to characterize borderline pathology, and to improve the patients' capacity to represent the internal world of self and others. Fonagy and colleagues (Fonagy, Gergely, Jurist, & Target, 2002) have observed that the majority of treatments for BPD, regardless of theoretical underpinnings, are characterized by the dual goals of building an attachment relationship with the therapist and improving the patients' capacity to think about self and others in mental state terms, defined as interpreting the behavior of oneself and others in terms of intentional mental states, such as desires, feelings, beliefs, and motivations (Allen, 2003).
TFP fosters the building of an internal representation of minds of self and other (mentalization) by focusing on the differentiation and integration of representations of self and significant others, and by identifying the major affects linking such self-object dyads, thereby integrating dissociated or split-off affect states that are linked to these representations. The result is that affective experiences become enriched and modulated. In short, the major goal of TFP is to change the pathogenic object relations that lead to chronic affective, behavioral, and cognitive disturbance. Clinical researchers at the Personality Disorders Institute have described the tactics and techniques of TFP in three volumes of a treatment manual (Kernberg, Seizer, Koenigsberg, Carr, & Applebaum, 1989; Clarkin et al., 1998; Yeomans et al., 2002) and have delineated how complex treatment issues may be addressed in TFP as well as how it may be tailored to the needs of the individual patient (Koenigsberg et al., 2000). Recent outcome studies from the Personality Disorders Institute with 17 patients who completed 1 year of TFP indicated that borderline patients showed a significant reduction of suicide attempts and suicidal behaviors, a decrease in medical risk and severity of medical conditions following self-injurious behaviors, and a decrease in hospitalizations (Clarkin et al., 2001).
The integration of attachment concepts with TFP
Although the outcome data on TFP are promising, we know that not all borderline patients benefit equally from TFP. During the past 5 years, we have begun to investigate how characteristics such as the individual's state of mind with respect to early attachment relationships and capacity to reflect on mental states of self and others (Fonagy et al., 1995, 2002), might affect the patient's progression through the phases of TFP and reactions to the techniques of TFP. Bowlby (1988) fully intended that the concepts of attachment theory would illuminate our understanding of more severely disturbed, narcissistic, and borderline patients and their treatment. He hypothesized that just as the availability of a secure base in childhood facilitates the child's exploration of the external world, so does the therapist and the therapeutic situation serve as a secure base from which the patient can engage in self-exploration. Bowlby (1975) wrote that the chief role of the clinician was to "provide the patient with a temporary attachment figure" (p. 291). Similarly, Fonagy and colleagues (2002) have hypothesized that the capacity to think about the self and others in mental state terms (i.e., to attribute intention, beliefs, and attitudes to significant others) is anchored in secure attachment relationships.
The view that there may be an attachment substrate to the therapeutic relationship that is parallel to, but distinct from, the activation of libidinal or aggressive feelings both converges with and extends the object relations foundations of TFP (Diamond & Yeomans, 2003). Although those of us who have developed and elaborated on TFP have historically given more emphasis to dynamic conflicts and, particularly, the role of aggression as an impediment to the integration of the internal world and the formation and maintenance of secure attachment bonds, we have found it productive to explore the ways in which the attachment features of borderline pathology, particularly the insecure states of mind with respect to attachment that characterize borderline pathology, help to configure the transference and countertransference dynamics central to the techniques of TFP (Diamond et al., 1999, 2002; Koenigsberg et al., 2000). In addition, Bowlby's theories about the bidirectional nature of the attachment behavioral system has made our therapists and researchers more cognizant of the ways that attachment factors might influence the therapist as well as the patient. Like all attachment relationships, the therapeutic one was thought by Bowlby (1969/1982, 1973) to be inherently bi-directional, with attachment-seeking behaviors (proximity seeking, smiling, calling) tending to evoke corresponding adult attachment or caretaking behaviors (soothing, holding, protecting). In Bowlby's view (1978), the attachment behavioral system inevitably contributes to the configuration of transference and countertransference dynamics, for it is activated throughout the life cycle in situations where an individual who is ill and in distress seeks protection from or contact with someone deemed older or wiser.
Indeed, recent empirical and clinical investigations on the extension of attachment concepts into the clinical arena have shown that attachment status has been linked to transference-countertransference dynamics (Fonagy, 1991; Gunderson, 1998; Holmes, 1995, 1996; Szajnberg & Crittenden, 1997), the quality and nature of the therapeutic alliance (Dozier & Tyrrell, 1998; Dozier, Cue, & Barnett, 1993; Eagle, 2003; Mackie, 1981), and patterns of patient-therapist narrative discourse (Fonagy, 2001; Slade, 1999). Although a comprehensive description of these investigations is beyond the scope of this article, we will highlight several recent studies that are particularly relevant to the current study. Dozier and colleagues found that patients in supportive treatment who were classified as secure/autonomous with respect to attachment were more involved in and cooperative with treatment, whereas those with dismissing states of mind were more rejecting of help and superficial in their engagement with the therapist and those with unresolved/preoccupied states of mind required more crisis intervention (Dozier, 1990). Dozier and colleagues have also found that clinicians with secure states of mind are more likely to challenge patients' own strategies for relating interpersonally and to intervene in greater depth, whereas clinicians with insecure states of mind are more likely to mirror the patients' interpersonal dynamics (Tyrrell, Dozier, Teague, & Fallot, 1999). They suggest that the best treatment outcomes and overall ratings of treatment alliance occur when patients and therapists' attachment state of mind are complementary rather than concordant. In this context, the therapist is more likely to challenge the patient's characteristic ways of regulating affect and distress in interpersonal contexts, leading to better therapeutic outcomes.
Another recent study by Eagle and colleagues (Parish & Eagle, 2003; Parish, 2000) provides additional empirical support for the centrality of attachment constructs to the therapeutic relationship. Using the Components of Attachment Questionnaire (CAQ; Parrish, 2000), Parish and Eagle (2003) identified nine major components of an attachment relationship as being present in the therapeutic relationship: Proximity Seeking, Separation Protest, Secure Base, Turning To for Comfort, Support, and Reassurance, Stronger and Wiser, Available and Responsive, Strong Feelings, Particularity, and Evoking a Mental Representation. The components found to be the most strongly associated with the therapeutic relationship for patients were the components Stronger and Wiser and Available and Responsive. In addition, Eagle and colleagues found that the more intense the attachment to the therapist, the more positive the working alliance; they also found that the intensity of attachment varied among those with different states of mind, with dismissing patients having the lowest intensity of attachment to the therapist (Parish & Eagle, 2003).
The studies provide empirical affirmation for Bowlby's (1988) conceptualization that ( the therapist is a prototypical attachment figure in adulthood, and that patient's and therapist's attachment status may affect the treatment process and relationship. However, the majority of previous studies on attachment in the therapeutic relationship have been conducted with attachment measures that assess the state of mind with respect to attachment of patients and therapists vis-à-vis their own early attachment relationships and experiences, rather than quality of attachment within the patient-therapist relationship itself.
- Diamond, Diana, Kenneth Levy, John Clarkin, and Chase Stovall-McClough; Patient-therapist attachment in the treatments of borderline personality disorder; Bulletin of the Menninger Clinic; Summer 2003; Vol. 67 Issue 3
Reflection Exercise #10
The preceding section contained information
about Transference-Focused Psychotherapy for Borderline clients. Write
three case study examples regarding how you might use the content of this section
in your practice.
According to Diamond, to what three concepts has client-therapist attachment been linked? Record the letter of the correct answer