Projective Identification

Recently, at TBIPS’ Relational Study Group, a paper we were discussing brought up projective identification, a concept that has baffled clinicians for decades. I would like to share my thoughts on projective identification. When Melanie Klein coined the term, it was to describe an intrapsychic phenomenon: how the unwanted parts of the self (often aggression or helplessness) were projected into the other, and were then seen as under the control of the self, and, consequently, so too was the other seen as under the control of the self. Note that s/he who did the projecting was also doing the identifying with the unwanted part.

Projective identification became redefined as a phenomenon between two people: what was projected by the self was identified with by the other such that the other began to behave commensurate with the projection. E.g. a patient projects sadistic impulses into the therapist and the therapist inadvertently becomes sadistic, or feels sadistic. The self could behave in such a way as to make the other feel what it felt like to be the one doing the projecting. Projective identification, then, began to include the countertransferential use of the projection giving the therapist information about the patient’s experience.

Even when projective identification is seen as emanating from the paranoid position, it nonetheless is an attempt to communicate how one is feeling, as well as a wish to be understood. Because patients are sometimes bereft that the therapist will ever know what it feels like for the patient to experience such rage or helplessness or envy or despair, they are nonetheless sometimes fortunate enough to utilize projective identification to help move the therapist’s understanding along. Otherwise, patients might feel untenably isolated and unconnected to us. Despite how uncomfortable (causing the therapist to disparage the defense and pathologize the projecting patient) it may be to be in the throws of the experience, if not bound by the projection, the therapist might sometimes have the wherewithal to inquire whether the patient has somehow managed to come upon a way to aid the therapist to better know what the patient’s experience felt like by inadvertently helping the therapist to feel that way too.

Since its original definition was conceived within a one-person psychology, contemporary Relational theorists, if they use the term at all, have further redefined projective identification. Relational theory questions the unidirectionality of projective identification. It does not conceive the projection to originate entirely from the patient’s psyche but instead recognizes how unlikely the therapist would be to ‘identify’ with a ‘projection’ were it not to resonate with something already within the experience of the therapist. The therapist, too, contributes to what is projected in addition to resonating with the projection.

I like to say that it is one thing to wear the attributions of a patient and explore what, for both the patient and oneself, it would be like to do so, but it is something else to be what the patient attributes to us. I suggest to less experienced colleagues that we be brave about discussing patients’ (e.g. aggressive) feelings toward us. When patients have ‘split’ us into the bad or hateful object, it sometimes helps restore personal equanimity to consider how and why it became necessary for the patient to operate, at this moment, from the paranoid position. Harder is to consider how we might have inadvertently contributed to this shift.