Discussion of the Paper by E. R. Shapiro, R. L. Shapiro, J. Zinner and D. A. Berkowitz on 'The Borderline Ego and the Working Alliance: Indications for Family and Individual Treatment in Adolescence'

This paper by E. R. Shapiro et al. (this issue) reports another segment of the authors' excellent clinical studies of disturbed adolescents and their families. I discern only one difficulty, and that is their use of the concept of splitting. A brief recapitulation of their thesis will prepare the way for clarifying my reservations.

The thesis of the paper is that the borderline adolescent is compelled by a variety of forces (striving for autonomy is one) towards separation from his family. This separation intensifies life and death dependency needs, and it exacerbates tension over similar dependency issues in the parents. The resulting distress is more than any of the troubled family can tolerate; they revert as a group to pervasive use of certain defences which are developed at an earlier stage in their life cycles. The authors identify these defences as splitting, projective identification and, by implication, introjection. The family employs these defences in such a way as to fixate the adolescent's pathological state; his reaction to the parents' use of projective identification is especially pernicious.

The borderline adolescent is very often intensely negativistic in impulses, feelings, and ideation towards his parents, the therapist and himself. Non-specific ego weakness and his ready resort to projective identification frequently make it difficult for him to perceive his therapist as really helpful. The working alliance is especially vulnerable to being capsized by hostilities emanating from family interactions. The authors therefore hold that the therapist should work both with the family as a group and with the adolescent individually. When the therapist clarifies and corrects pathological family interactions, his patient has less trouble to see him as a helpful person who is different from pathological introjects. The leaky working alliance can be plugged up in this way, and therapy can move ahead more securely.

The clinical data and thesis of the