False Memories Turned against the Self

Saul M. Kassin
Department of Psychology
Williams College

Memory construction is a phenomenon that has be-
come intuitive to the experimental psychologist. In
recent years, researchers have found that misleading
postevent information can alter actual or reported
memories of observed visual events ( Loftus, Miller, & Burns, 1978; McCloskey & Zaragoza, 1985), particu-
larly among young children ( Ceci & Bruck, 1993) and
adults under hypnosis ( McConkey & Sheehan, 1995).
Recent studies suggest that it is possible as well to
implant false recollections of words in a list ( Roediger & McDermott, 1995) and isolated childhood experi-
ences--such as being lost in a shopping mall--that
supposedly had been forgotten ( Loftus, 1993 ).

Despite the apparent ease with which experimenters
have been able to tinker with minor recollections of
their subjects, this research did not, and indeed could
not, prepare us for the kinds of wholesale manipulations
of autobiographical memory de Rivera describes. The
case studies he presents--of four former psychotherapy
patients who recovered "memories" of prolonged child
abuse only later to retract these memories--seem in-
credible, as do the mind-control and narrative models
he offers to explain these cases. As de Rivera himself
admits:

It is one thing to replace the image of a stop sign with
that of a yield sign . . . yet quite another to have a person
replace a purportedly happy childhood with a belief
that he or she was systematically sexually abused by a
previously adored parent.

Surprising or not, a sufficient number of false mem-
ory syndrome (FMS) cases have been reported in recent
years to shake even clinically minded skeptics prone to
trust self-reports of abuse but distrust the retractions of
these self-reports. Assuming that some unknowable
percentage of FMS cases prove to be legitimate (i.e.,
where traumas initially reported but later retracted did
not occur), psychologists should seek not only to vali-
date each diagnosis, but to identify the social influence
processes by which these extreme, heart-wrenching,
and self-destructive memories were ever constructed in
the first place. De Rivera's case studies and conceptual
analysis provide valuable insights toward this end.

De Rivera proposes two explanations, not mutually
exclusive, for the therapy-induced creation of false
memories. One is a "mind-control" model in which the
therapist overwhelms his or her patient by making an
abuse diagnosis and then supports that assessment by
manipulating the patient's informational and emotional
state. This heavy-handed type of influence closely re-
sembles the thought-reform or "brainwashing" tech-
niques previously seen in Korean War prison camps and
certain religious cults. In the alternative "narrative
model," the patient primarily leads the therapist, creat-
ing a trauma story from the past as a way to understand
or lay blame for his or her unhappy current state.

To evaluate these two models, de Rivera interviewed
four FMS victims, or retractors, concerning their back-
grounds, the memory-induction process they under-
went, and their retractions. Using the method of con-
ceptual encounter, de Rivera described the two possible
models to each respondent, and together they tried to
conceptualize her experience within these frameworks.
On a methodological level, this study is flawed in some
important ways. First, the sample is small, and we have
no basis for accepting de Rivera's belief that "it appears
representative." Second, de Rivera himself conducted
the interviews, a procedure that paves the way for the
intrusion of experimenter expectancy effects. Third, all
the data concerning the critical induction process were
obtained from the retractors themselves--"research
partners" with a prior record of deception and with
self-justificatory motives that might systematically
have corrupted their self-reports. There are two sides to
every story, of course, and studies have shown that
actors and observers clearly differ in the causal attribu-
tions they make ( Jones & Nisbett, 1971; Watson, 1982).
Thus, one can only speculate about the way the thera-
pists involved in these cases would have depicted the
same events.

In light of the foregoing limitations, a result favoring
one model over the other should be accepted with
caution. In fact, however, de Rivera finds that neither
model completely fits the experience of all four retrac-
tors and that both perspectives on FMS are necessary
for understanding the processes at work. Clearly, as-
pects of the stories told by Ann, Cath, and Doris con-
form to the mind-control model. Yet just as clearly, the
stories told by Beth and Doris suggest a more narrative
account. Thus:

Relatively normal persons from relatively functional
families may develop ...