Discussion of “The Merits and Problems with the Concept of Projective Identification” by Janet Finell*

I wish to thank Dr. Finell for asking me to discuss her excellent paper, for it gives me a chance to say a few words about a subject that has perplexed me for some time. Projective identification as a construct or metaphor has always been confusing, and in my mind, obfuscating. Dr. Finell's paper raises two basic questions: Does it describe anything that goes on clinically between patient and analyst that has not already been described in different and simpler language? Does the concept of projective identification in any way enhance our theoretical understanding of the patient or the patient-analyst interaction?

Although I seem to be in the minority these days, my answer to both questions is “no.” I find the concept confusing and misleading, and I agree with Meissner (1980) who believes that the term should be dropped from common usage. In fact, I feel that the concept of projective identification has tended to exonerate the analyst, that is, it has helped him or her avoid the real issue of countertransference. Instead, the patient can be blamed for our untoward reactions. The patient has put into us split-off aspects of the self which we take in and contain. The real emphasis in the concept of projective identification has been on what the patient does to us, not what we bring to the situation.

In reviewing Grotstein's (1981) book, Dorpat (1983) points out how the concept of projective identification has been used in an ambiguous and contradictory manner by Grotstein and by other writers on the subject. It is used to describe an unconscious fantasy of attributing