KEKNBEKG AND KOHUT: A TEST OF COMPETING PSYCHOANALYTIC MODELS OF NARCISSISM

Over the past teii jears the problem of validating psjclioaiial)tic theory has received increasing citteiilioti from pgclioanal)sts atid Iiorip.~clioaiialy.s~s alike. Yet, very lillle research responsive to tltk questioti of ualidation Iias been produced. During this same period oiie of the most fertile controversies iti conternforaty {qdiounalysis reached its Ieclk-tlie Kernberg-Kohl debale coiiceniing the nature of narcissism. Hourever, for the lack of a valitlation strategy, this debate has becorne iticreasiiigly sterile, degenerating lo a llieoretical staletriate at titis poitit. This paper addresses the problem of validation by using causal morleliiig a tnecliaiiism far einpirically testing many of the claims of these cotripetirig models of tiarcksism. Allliougli lire data provide empirical supyort for both theories. strategic camparisom ten fatively suggest that Kolitds s e y pgcliology is more #arsiinoiiiomly ex-plained as a special case of Keniberg’s ego psjcliology-object rela-tioils theory. N 111s i w i s I D m * m u. AUDKESS TO T H E American Psychoanal- I ytic Association, Kaplan (198 1) enjoined his colleagues to move psychoanalytic theory “from discovery to validation. ” He stated: “No matter how we have arrived at our psychoanalytic hypotheses and theories, basic scientific method requires that techniques be devised to substantiate these discoveries. Discov-ery to be followed by validation is as essential for psychoanalysis as it is for any other science.... This, I feel, is the fundamental challenge to psychoanalysis ” (pp. 5, 23). Indeed, the problem of validation is implicated, directly and indirectly, in the intense Independent statistical and research consultant, New York City. Accepted for publication December 17, 1986. With deep appreciatio