Alternative Perspectives on Psychodynamic Psychotherapy of Borderline Personality Disorder: The Case of “Ellen”

This report describes an intensive psychodynamic psychotherapy that the author conducted with a patient with borderline personality disorder named “Ellen.” Dr. Bateman, one of the founders of mentalization-based treatment (1) , and Dr. Kernberg, the founder of transference-based psychotherapy (2) , comment on the treatment, emphasizing the overlapping and distinctive aspects of the two forms of therapy. Each was asked to comment independently and then asked again to offer additional comments on issues that the other had brought up. As such, Ellen’s case illustrates alternative perspectives about psychotherapy with patients with borderline personality disorder.

This report offers vignettes derived from six time points in Ellen’s therapy: the time of referral, 3 months later, 11 months later, 4 years later (when the therapy effectively ended), and from follow-up at year 7.

Case Presentation

Vignette 1: The Start of Therapy

Dr. Bateman

I suspect that Ellen had no idea whether either of your suggestions about the causes of her self-harm were accurate or not. In mentalization-based treatment, we would focus on the patient’s mental state in the current situation. Ellen may have been wondering about who this “expert” was or about losing Dr. A because he could not cope with her. At the moment, your visit may have confirmed how sick she really was. Second, we would explore her state of mind long before she made the phone call to Dr. A. Although the call itself might have upset her, the call was a behavior arising in the context of failing mentalizing—being unable to process an emotional experience. We would trace Ellen’s states of mind, especially her affective experience, to find the point at which this occurred. The aim would not be to give her insight into the underlying causes. It would be to engage Ellen in a process of attending to her mental states.

Dr. Kernberg

Your dynamic formulation, i.e., that she was defensively refusing to address a topic she knew about, is right. However, I would inquire about her understanding of her cutting. I would not ask her to connect it to Dr. A. She was still too defensive.

Dr. Gunderson

For her part, in our initial meeting, Ellen wanted answers to two linked questions: did I think she had borderline personality disorder, and did I think she should discontinue her medications. With respect to both questions, I responded by saying that I did not know, but that answers would probably become evident if she decided to use me to help her understand herself. She found this response frustrating and complained, “You’re a doctor; you should know.” I also let her know that Dr. A would continue to manage her pharmacotherapy.

Although I had been told enough about Ellen to feel quite confident that she had borderline personality disorder, and I think disclosing the diagnosis of borderline personality disorder is usually a valuable first step in framing the goals of therapy and establishing an alliance, with Ellen, I was alert to the potential for splitting that could result from answering her two questions in ways that would contradict Dr. A’s viewpoints—she might feel that by seeing me she would betray Dr. A.

Dr. Bateman

Ellen wanted to know if she had borderline personality disorder, i.e., who she was. You “mirrored” that by saying you also did not know, which I think confused and terrified her. It provoked a momentary loss of mentalizing, and she reverted quickly to a schematic representation to restabilize mentalization—“You’re a doctor; you should know.” I do not think this means that you have to “know” and should have said “yes.” That would not have stimulated mentalizing. In mentalization-based treatment, we would comment on our own state of uncertainty, as you did, but then focus on the state of mind that this left her in (“I guess it is hard to talk to someone who is supposed to be an expert and seems uncertain about something so basic”). Your reassurance about Dr. A continuing medication management was important because it spoke to what in mentalization-based treatment we believe is a prementalizing way of thinking, in which “Shall I stop taking medication?” equals “Shall I stop seeing Dr. A?”

Dr. Kernberg

Although I too often talk openly about a patient’s diagnosis, I try not to push diagnostic issues beyond what the patient already knows. In this case, it was OK to say you did not know. I would, in addition, then interpret the query along the line that “If I were to say you have borderline personality disorder, it would attack Dr. A, and you would feel that you would be forced to lose one or the other of us.”

Dr. Gunderson

Despite the conventional wisdom that multiple treaters, i.e., split treatments, invite splitting, they can have significant advantages. The inability of the patient with borderline personality disorder to manage anger toward needed caretakers helps account for the high frequency of dropouts, noncompliance with medication use, use of suicide gestures as “calls for help,” and the resulting high burden on therapists. When two caretakers are present, there is a readily available opportunity to express anger toward either without fear that therapists’ expected withdrawal or retaliation would leave them abandoned. Split treatments—when practiced well—will diminish the likelihood of dropouts and improve the alliance with treatment goals as long as the treaters have an appreciation and respect for each other and for what each is providing.

Split treatment (individual + groups + communication) is a basic infrastructure for mentalization-based treatment (1) and dialectical behavioral therapy (3) that is also applicable to combining other modalities. In dialectical behavioral therapy, the process of going to talk with the “bad object” is called “repairing the relationship.” When Ellen took complaints about me to Dr. A, he appropriately encouraged her to talk to me about her complaints.

Dr. Bateman

We are in agreement about split treatments. Under circumstances in which treaters have “appreciation and respect for each other,” the structure of split treatments is excellent for promoting mentalization as long as the different therapists both focus on mentalizing the split, e.g., asking the patient to consider how he or she is with the other therapist and how he or she might address any complaint or other feelings about the other practitioner, etc.

Dr. Kernberg

While “split treatments” can be useful if they help patients integrate their anger, this is not usually the case. In my experience, if the therapist is comfortable with the patient’s anger, the patient will not act out on it. Therefore, I do not recommend split treatment as a standard component for transference-based psychotherapy.

Vignette 2: 3 Months

Dr. Bateman

We agree very much with your commenting about a patient’s appearance as a useful way of starting when there is something obvious going on. Also at this point, you keep things “current,” which is something that mentalization-based treatment emphasizes.

Dr. Kernberg

Although transference-based psychotherapy focuses on here-and-now interactions, in this situation, I would start by inviting her to say what’s on her mind and then, if needed, comment on her appearance. However, I would focus on the fact that she looked irritated, first asking her to own up to it, then exploring its source. The term “waif” might lend itself to misinterpretation, e.g., as derogatory.

Vignette 2 Continued

Dr. Bateman

You were giving her a way of explaining things that was developed by your mind rather than stimulating a process in her that was concerned with understanding her own mind; you may have been encouraging “pseudomentalizing.” Only after a joint attentional process was established would we interpretively mentalize the transference (“You got quite angry with me on Friday.”). We use the transference to show patients how the same behavior may be experienced differently and can be thought about differently by different minds—not to give insight.

Dr. Kernberg

Transference-based psychotherapy therapists would explore the source of her depression and inquire about why suicide is an option. Our focus would again be on her fear of telling you about her anger—anger at being pushed out of groups, anger at feeling asked to take care of her mother, and anger at you for misunderstanding her. We might also have noted at the outset that she had kept her appointment despite having been angry with you in her last appointment, letting her know that you recognized this as progress. We would note that there was a split-off part of her with a more positive affect that she was not fully aware of and point out how these represent contradictory states of self over time. Mentalization-based treatment neglects the need for interpretation of the alternating and currently irreconcilable feeling states (e.g., of hate and love ) or self-image (e.g., of good or bad ).

Dr. Gunderson

Ellen appeared to hear what I said but remained disengaged in response. Had she disagreed, I would have felt as if she were engaged with me and with her issues. Having failed, I then retreated into offering more supportive case-manager-type interventions. In dialectical behavioral therapy, this shift would have parallels with expecting too much change, and then following it with a reparatory effort to increase acceptance.

I think that Ellen was ashamed because she believed leaving the group therapy should not be a loss and she truly believed she should not be taking care of herself rather than her mother. So I did not accept her denials as necessarily reflecting what she knew or thought any more than I accepted her initial statements that she saw no connection between her arm slashing and Dr. A’s recent marriage. I did not ask Ellen to “confess.” I simply tried to normalize such reactions as if, to me, they were expectable and predictable.

Dr. Bateman

These are excellent mentalizing interventions. For some patients, it could be too long and complicated, but in this case, you reached out to Ellen’s vulnerability and stood alongside her in a way that gave her real support. Hence, she responded very positively as the session moved on.

Dr. Kernberg

In transference-based psychotherapy, we would not move to support such tools as offering to talk to the group therapist or advising Ellen to call home. We would explore why she was resistant, most notably, why she was fearful about getting angry at you and the possibility that she was fearful of communicating openly because of anticipating anger that she may have projected on you. We try to maintain technical neutrality as a foundation for the use of interpretations.

Dr. Gunderson

To my last comment concerning her mother, Ellen finally responded. She reported that she had already called and asked her mother whether her coming home was needed. She reported that her mother had said, “I’d feel terrible if I was responsible for disrupting your treatment.”

Dr. Bateman

Excellent mentalizing intervention: to the point, affected focus, captured her trapped state.

Dr. Kernberg

Excellent. You identified her internal conflict in the here and now, and you validated her ability to tolerate ambivalence.

Dr. Gunderson

I first tried to interpret her depression as a defensive response to two stressors, and then (because of her unresponsiveness to these efforts) I reverted to supportive interventions. My move to supportive interventions enacted her borderline transference; i.e., by becoming her caregiver, I was indirectly—but consciously—giving her assurance of her being a “good” person. Only after this did she become responsive, revealing an alert intelligence at work with respect to her mother’s ambiguous meaning. I think this demonstrated how sensitive to the immediate interpersonal context are the depressive symptoms of patients with borderline personality disorder, their capacity to be reflective (i.e., to mentalize), and their readiness to be self-disclosing. Depressions appear when patients with borderline personality disorder feel abandoned and remit quickly when they feel held, i.e., taken care of. That is why their depressions improve after the borderline psychopathology remits and rarely after antidepressant treatment (4) .

Dr. Bateman

In mentalization-based treatment we reverse the order; we use interpretations only after supportive work has been done. We would not formulate her subsequent responsiveness to your support as owing to what you call a “transference enactment.” Rather, we would suggest that your interventions had engendered a mentalizing process in which she was able not only to manage her internal state but also to become curious about the ambiguity of her mother’s response. Unlike the earlier session, when she initially could not tell you about her “affair” with Dr. A, Ellen realized that you did understand how real her mental states were for her and so she could disclose without being destabilized.

We very much agree that depressions should not be treated simply as comorbid disorders. The experience of depression in patients with borderline personality disorder is terrifying because self-related cognitions and affects are experienced as excessively real: “feeling bad” becomes “I am bad.” In this mental state, challenge or debate about the reality of the negative states is fruitless, possibly harmful, and simply makes the patient feel misunderstood.

Dr. Kernberg

I think your comments about the sensitivity of the depression of patients with borderline personality disorder to feeling held did not give adequate attention to the role of their guilt about anger and aggression. Depressions relate not simply to interpersonal abandonment experiences but to the patients’—in this case, Ellen’s—abandonment of their anger. She may have become engaged as a response to your support, but your supports validated her perceptions of herself as a victim.

Dr. Gunderson

Central to this account is my ongoing judgments about Ellen’s immediate state of responsiveness—whether Ellen was relating to me and was engaged by what I was saying or not. Selections from the hierarchy of therapeutic interventions developed by Horwitz et al. (5) and others should be guided by the patient’s immediate mental state. Patients with borderline personality disorder fluctuate dramatically within sessions in their level of engagement, in their alliance with the objectives of the therapy, and in their responsiveness to interpretations (6) .

Dr. Bateman

Therapists may not adequately note the rapid fluctuations in their patients’ mental states that occur within sessions. At one moment, the patient can be able to reflect, but then a therapist can unwittingly do something apparently minor that destroys the reflective process. Therapist intervention should match the patient’s mentalizing capacity by using attachment-stimulating supportive interventions when mentalizing is at its lowest and by mentalizing the transference only when mentalizing is at its most robust.

Vignette 3: 11 Months

Dr. Kernberg

I would remain silent here; wait and see what she says. It was her role to communicate to you what was going on. If she did not, then I would inquire what experience in the moment was keeping her from speaking.

Vignette 3 Continued

Dr. Bateman

Here are examples of what we consider to be good mentalization-based treatment interactions: “What’s going on?” (a “not knowing” inquiring mentalizing stance) and the patient saying, “I don’t know.” You then express some surprise; you do not insist that you know what is going on, but you highlight a possible aspect. You then highlight the possibility that what you have been doing might not have correctly understood her; i.e., you “mark” your response as yours and do not suggest it is hers. This encourages her to consider it.

Dr. Kernberg

She dares to disagree, to be more aggressive. This is a good sign of progress! I would point this out. In transference-based psychotherapy, therapists encourage patients to feel comfortable with criticisms, attacks, anger, and even hatred.

Dr. Gunderson

Unlike the last episode, at this time (8 months later), Ellen was more responsive to my efforts to make causal connections between her depressive mental state and events, and she seemed more able to acknowledge her anger at me and my significance to her.

Still, I worry that Ellen’s continued failure to validate my causal connections may reflect the futility of my interpretive efforts, possibly even making it more difficult for her to make such connections in the future. I think Linehan (dialectical behavioral therapy) might join Bateman and Fonagy (mentalization-based treatment) in considering my interpretive efforts harmful—failing to include the methodical and pains-taking inquiry about all the intermediary states of mind, i.e., “chain” or “functional” analyses needed to meaningfully link feelings and events.

Dr. Bateman

It is not that your interpretive efforts were futile. You were certainly trying to do “chain” or “functional” analyses, and you seemed fully aware that you might stimulate pseudomentalization. My caution again is to beware of burdening the patient with your own mentalizing.

Dr. Kernberg

Ellen had clearly progressed in mentalizing, i.e., she had become more aware of her own and others’ mental states. Your interpretations had helped. But this was only a start. She still feared being angry when someone hurt or disappointed her and feared rejection if she expressed her anger. Your gentle laughter at her “Bug off” comment was an indirect effort to make her expectation of rejection dystonic. She nevertheless experienced you as rejecting.

Vignette 4: 3 to 4 Years

Dr. Bateman

In mentalization-based treatment, we do not assign anger a central role in either the development or therapy of borderline personality disorder. Within therapy, it is too easy to interpret thoughtlessness, desperation, and terror as anger. Anger is commonly conflated with motivation and given inappropriate meaning, e.g., “He did this as an attack on me,” etc. From a mentalization perspective, anger is usually seen as a response to an action or comment from the therapist. Rather than taking up the purported meaning and centrality of the expressed anger, the mentalization-based treatment therapist will want to consider what he has done to evoke the anger.

In mentalization-based treatment, we would refrain from advising Ellen to focus on work rather than love. It is too easy to misjudge a patient’s capacities. The fact that she was able to start and commit to a relationship was a positive outcome. Although she continued to show vulnerability, I think her overall outcome was good.

Dr. Kernberg

Your encouragement to rely more on work is interesting, but transference-based psychotherapy would not do that. Although I agree with the conclusion that she had not integrated her aggression satisfactorily, the possibility of achieving this remained. Her engagement to an older man named John raised questions as to whether the sexual part of her life and of her transference had been fully explored in this therapy, just as these issues had not been adequately addressed with Dr. A.

Vignette 5: 7 Years

Dr. Bateman

I would not push the patient to resume therapy or necessarily suggest that it should be mentalization-based treatment if she decided to do so. Ellen had demonstrated a capacity to manage the powerful emotional event of her husband’s sudden death. Her current ability to tolerate and to process emotional states was commendable. Her concerns about her isolation and that she would never be a happy person were painful recognitions. Still, I had the impression that these were things that she accepted, albeit with regret.

Dr. Kernberg

At the time of this “refueling,” I would probably have encouraged an extended consultation in which discussions of the persisting limitations in her life might have provided motivation for further therapy.

Discussion

Ellen’s overall course of change reflected a generic sequence of what can be expected from long-term treatments: first alleviation of subjective distress, then behavioral change, then improved interpersonal relationships, and finally intrapsychic changes. Ellen was recurrently depressed for much of the first year, with gradual lessening of its severity. She did achieve behavioral change, but the most prototypic behavioral pathology, cutting, had ceased years before she began with me, a product, I think, of Dr. A’s support and of her aging. Six months after starting therapy, she had begun to work part-time; achieving stable community-based supports and vocational activity should be expected in the first year. She also had begun to depend on me as someone who she knew cared about her and was reliably attentive to her best interests without her expecting me to be a rescuer. Achieving a positive dependency is a nonspecific corrective attachment experience. The unexpectedly frequent remissions of borderline personality disorder observed in longitudinal studies (7 , 8) is a testimonial that this can be achieved—and usually is—with people other than therapists.

The most fundamental disagreements about techniques between the two discussants involved the importance assigned to interpretations in transference-based psychotherapy and to supportive interventions in mentalization-based treatment. This disagreement in turn was related to differences in how transference-based psychotherapy and mentalization-based treatment conceptualized borderline personality disorder psychopathology, and most specifically, the patient’s problems with anger. Transference-based psychotherapy sees unacknowledged or unintegrated anger as the core problem. The focus is on integrating this anger and the derivative hostile/punitive or helpless/victim part objects into a whole and stable self. To not do this, Dr. Kernberg suggests, perpetuates an identification as a victim. My inconsistent focus on anger accompanied by my readiness to be supportive was the reason why Dr. Kernberg felt the therapy was incomplete and that a return for a transference-based psychotherapy might still be in Ellen’s interest.

The mentalization-based treatment model considers anger a mental state that a therapist will identify to help the otherwise unknowing patients with borderline personality disorder label their experience and to help them learn about its role in causing behaviors, etc. But the mentalization-based treatment therapist would be wary of interpreting anger, especially in the therapy relationship (what transference-based psychotherapy would call transference), because even if accurate, such interpretations, unless “robust mentalizing” capacities are available, destabilize borderline patients rather than aid self-integration. Mentalization-based treatment would focus on the patient’s current mental state and mental functioning. By giving it attention and thereby underscoring the importance of these mental states while assuming a “not knowing” inquisitive stance, mentalization-based treatment therapists help patients become more introspective (“reflective”) and develop more of a sense of self and self-agency. Insight per se is eschewed. It seems notable that although both therapies aim to establish a more coherent and stable sense of self, their theories and techniques about how therapy facilitates this are radically different. To my mind, Ellen’s more stable sense of self and her improved ability to mentalize that allowed her to process the loss of her romantic partner were partly the result of a corrective relationship that made her become more accepting of herself and partly a result of my interpretations, although their traction depended upon the use of supportive interventions.

Ellen’s commitment to finding a love relationship at the expense of finding a vocation is not unusual for many people with borderline personality disorder. Although this led me to suggest she give primacy to work, neither of the consultants agreed with this. Ellen’s subsequent decision to give up on her hope for “love” is not an unusual adaptation for aging people with borderline personality disorder; usually this occurs in the 30s after concluding that such hopes have only brought heartache. For many, love gets replaced by broad and nonintensive sources of support that can come from churches, organizations, communal living, etc. Ellen will, I think, always be somewhat bitter and alone. If Ellen seeks psychiatric help, I think it will probably be resuming medications and not psychotherapy. Her life truly was unfair.

Beyond the interesting contrasts between mentalization-based therapy, transference-based psychotherapy, and my own perspectives, these alternatives all recognize that to be effective with patients with borderline personality disorder requires extensive modifications from the technical neutrality and lack of structure that characterize traditional psychoanalytic therapy. Much of the extensive early literature on psychoanalytic treatments for borderline personality disorder provides lessons in how inadequate structure, hostile or rescuing countertransferences, and a failure to be an active participant in here-and-now interactions led to rages, suicidal threats or gestures, therapeutic regressions, noncompliance, excessive intersession demands, and frequent dropouts. Patients such as Ellen should now be able to expect that the current generation of psychodynamic therapists will have learned these lessons.