Gerhard Stemberger told me of your discussion about "projective identification". Your thoughts are very stimulating. All this is very interesting for me, because I will hold a lecture "Psychoanalytic concepts in the light of Gestalt theory" at the 10th Scientific Convention of the Society for Gestalt Theory and its Applications (GTA)in Vienna (see http://rdz.stjohns.edu/~gerhard/gta/tgeng.html), and "projective identification" will serve as an example, how this rather complex phenomenon can be explained in terms of Gestalt theory.
Stuart Stawman mentioned the Object relation theory. To me this theory remains somewhat vague and gives a "historical solution" to a present process. Kernberg (1978), for example, refers to very early identifications (first four month after birth), that are in the present situation re-activated in both, the projecting person and his partner. In a therapeutic relationship the therapist is forced into what Kernberg calls a "empathic regression". In the course of this process the therapist will overtake the projected aggressions and try to dominate the client.
Iīm not really satisfied by this explanation, because it does not take into account the dynamics of the present here and now-process. It still appears like a physicalistic simplification (?): aggression is transferred from one person to another like water from one bottle into another.
How can a phenomenon like this be explained in Gestalt-theoretical terms? Kurt KOFFKA (1935) proposed a theory explaining how one personīs feelings can be perceived by another person. He differentiates between a geometrical point-to-point representation (this is what happens in the physical world) and a dynamic organisation of the psychophysical field and thus the phenomenal world. The phenomenal world contains the "perceived world" with the perceiving person herself, the other persons and the environment. The perceiving person experiences in his/her phenomenal world a specific pattern of motion in another person that "contains" as Gestalt-qualities certain affective states. That means, like "sad" pictures or "delightful" melodies the affective state is immediately perceivable. There is no need for interpretation or association.
So it is possible that an empathic person (e.g. therapist)experiences in his/her phenomenal world disassociated affective states of another person, although it seems that there is actually nothing to be seen, to be heard, to be felt. The affective state as a Gestalt-quality (although dissasociated)is not lost but still located within the phenomal field of the perceiving person but cannot be connected to the perceived person. Instead, the feelings are transferred to the own person.
In this way the phenomenon "projective identification" can be explained systematically in Gestalt-theoretical terms without referring to "historical solutions". There can nevertheless be the need to explain which biographical events may have contributed to a person's disassociating of parts of her personality. But the Gestalt-theoretical frame seems sufficient to explain the dynamic process underlying a projective identification. The particularity of this defence mechanism seems to be that in empathic situations persons have difficulties in distinguishing aspects of the own self versus aspects of others within the phenomenal world, especially when one person has no access to certain parts of his/her phenomenal field. The psychoanalytic explanations emphasizing early childhood experiences remain hypothetical. We have to be cautious in connecting present pathological phenomena in adult persons with hypothesized early developments. At least some of the theoretical assumptions of FREUD and KLEIN have to be corrected in the light of empirical developmental psychology (e.g. STERN, 1985).