Memory, Abuse, and Science

The controversy concerning traumatic memories and recovered memory therapy has generated at least as much heat as light. For example, Kenneth Pope (September 1996) , one of psychology's most distinguished ethicists, verged close to ad hominem in his discussion of criticism of the trauma-memory argument and recovered memory therapy, effectively distracting readers from the salient scientific and clinical issues at stake. Early in the article, Pope (1996, p. 959) cited my definition of false memory syndrome, quoted a criticism of the term by Carstensen et al. (1993) as "non-psychological," and implied that my writing on this issue is intellectually dishonest. Unfortunately, Pope neglected to give a bibliographic reference citation to the chapter at issue, which was originally written and circulated in 1994 (Kihlstrom, in press) . As a result, readers will have difficulty knowing that I actually provided–on the page after that from which Pope took his quotation–a refutation of Carstensen et al.'s criticism. As Pope must have known, and Carstensen et al. should have realized, the word syndrome is not the exclusive property of the medical profession. Language exists for all to use. In fact, as far as I can determine, the earliest nonmedical usage of the word syndrome to refer to a pattern of behavior dates back almost 40 years (de Beauvoir, 1959) . If Pope is to join Carstensen et al. in questioning the intellectual honesty of those who utter the word syndrome without the benefit of a majority vote of the American Psychiatric Association, I hope he will include battered woman syndrome (Walker, 1984) , postincest syndrome (Blume, 1990) , and repressed memory syndrome (Frederickson, 1992) , among many other syndromes, as the targets of his criticism.

Later in the article, Pope (1996) took issue with my assertion that it is not permissible to infer a history of childhood sexual abuse from certain mental and behavioral symptoms (e.g., wearing loose clothing or having an inability to trust other people)–an apparently common clinical practice. Again, Pope gave inadequate references–two postings to an Internet mailing list (themselves improperly referenced; see the Publication Manual of the American Psychological Association, 4th ed., pp. 173—174; American Psychological Association [APA], 1994 ) and an article, published in a Dutch psychotherapy journal, that apparently had its origins in another Internet posting. As a result, the readers of Pope's article will be unlikely to realize that my remarks occurred in the context of a vigorous and informal Internet debate concerning the validity of symptom checklists of the sort proposed by Blume (1990) . Nor will the readers have had the opportunity to make up their own minds about this issue, because Pope's failure to properly cite my work effectively deprives them of independent access to what I actually wrote and the context in which I wrote it.

The fact is that the inference of a history of childhood sexual abuse, based solely on adult mental and behavioral symptoms of the sort listed in Blume's (1990) checklist–a practice seemingly endorsed by Walker (1994, pp. 113—114) in a book published by APA–is impermissible on both scientific and logical grounds ( Kihlstrom, 1996a , 1996b ). Scientifically, such an inference is impermissible because there is no empirical evidence for a specific association between child sexual abuse and any of these sorts of symptoms (for a comprehensive review, see Kendall-Tackett, Williams, & Finkelhor, 1993 ; for a detailed methodological discussion, see Dawes, 1993 ). But even if such associations were established empirically, the inference of childhood sexual abuse from adult psychological symptoms would remain logically impermissible because such an inference violates normative rules of deductive reasoning (it is, in fact, an example of the common logical error of affirming the consequent).

Pope (1996, p. 967) was right to be concerned about the implications of my argument for the mandated reporting of suspected child abuse. But the implication is not that those who suspect child abuse should not report it or that therapists should not pursue diagnostic leads. The implications are simply that intuitions and good intentions are not enough and that such suspicions and leads should be firmly based on what can be logically inferred from empirical science. Otherwise, clinical psychology forsakes its scientific base and lapses into solipsism.

The controversy concerning traumatic memories and recovered memory therapy threatens to tear apart organized psychology. Witness APA's own Working Group on Investigation of Memories of Childhood Abuse, which evidently failed to reach consensus on any important scientific or clinical issue and whose final report, as yet unpublished, was written in the form of a debate. The dispute over recovered memories and false memories is not between scientists and practitioners, as there are plenty of both on each side. Rather, the dispute has to do with the question of how closely clinical practice should be tied to basic science. Clinical psychology, as a profession, owes its status and autonomy to the assumption that it is firmly based on scientifically validated principles and techniques. The excesses of those who promote the trauma-memory argument and recovered memory therapy cast this assumption in doubt and thereby threaten the profession as a whole: If recovered memory therapy is discredited, it may take the rest of psychotherapy down with it. The best way to protect psychology as a profession and to promote psychology in the public interest is to insist on the centrality of psychology as a science. And that, for the scientists and practitioners whom Pope (1996) criticized, is what the false memory debate is all about.

Correspondence may be addressed to John F. Kihlstrom, Department of Psychology, University of California, Berkeley, CA, 94720.
Electronic mail may be sent to kihlstrm@cogsci. berkeley.edu