THE SCIENCE OF MEMORY:
APPLY WITH CAUTION

Experimental psychology has much to offer in the current debate about memories for childhood abuse. However, laboratory scientists, with their enormous cognitive authority to define reality for the rest of the population, must be especially conservative when arguing that laboratory results on memory generalize to contested memories of abuse.

In an article published in the July 1995 issue of the Journal of Experimental Psychology: Learning Memory and Cognition (volume 21, pp 803-814) Henry Roedgier and Kathleen McDermott reported that when subjects studied a list of words with a common, but not presented, associate, subjects frequently falsely reported remembering the never presented associated word as part of the list. The article began with these words: "False memories -- either remembering events that never happened, or remembering them quite differently from the way they happened -- have recently captured the attention of both psychologists and the public at large." It was David Gleaves' and my impression that Roediger and McDermott presented their laboratory results as "dramatic evidence of false memories" in such a way that some readers might well understand this to mean dramatic evidence for the concept of false memories of abuse.

In a commentary published in the May 1996 issue of the Journal of Experimental Psychology: Learning Memory and Cognition (volume 22, pp 811-813), Gleaves and I urged caution in making such a generalization, arguing that there are critical differences between Roediger and McDermott's findings and contested memories of abuse. One such difference buried in the definition of "events" is the unit of analysis (individual words versus episodes). An even more important difference is the purported similarity of the false item to the true items. In Roediger and McDermott's study, participants memorized words like "shoe, hand, toe, kick, sandals" and their false memories were of highly related words like "foot." In most cases of contested memories, one side argues that the memories are not just false in detail, but false in essence. Thus, if someone who was exposed as a child to anal rape, and forced to watch pornography of vaginal rape, later incorrectly remembered she was vaginally raped (arguably analogous to the false items in the word experiment), we would presumably not count this as a "false memory" in the popular sense of the term.

A new extension of this laboratory research (in which participants falsely remember words highly related to words on a list they studied) using neuroimaging technology is, as of this writing, receiving widespread media attention, and the new research is being presented as relevant to the contested memory debate. In an article in Newsweek (July 15, 1996) reporting on this research it is stated "when someone imagines a pseudo-event over and over, she often implants sensory data about it in the mind. She can actually see or hear or feel an event that never occurred." That Newsweek leapt from reporting PET results showing various patterns of brain activation for remembering true and very similar words on a list to making such a claim is clearly worrisome. Trauma and memory researchers must take the time and trouble to counter such unjustified leaps of imagination, just as they engage in interesting work on memory and trauma using behavioral, neuroimaging, and field methodologies.

Problems of application are also common when research is applied to individual cases of contested memories, such as the infamous Ingram case,. In 1988 Ingram, a deputy sheriff in Olympia, Washington, confessed to sexually abusing his two daughters. Later, after extended and repeated questioning, Ingram claimed to remember committing increasingly bizarre and horrific crimes. Charges based on these later confessions (which have been criticized widely for being inappropriately obtained) were eventually dropped. Ingram plead guilty to the original charges and he was sentenced to prison. However, he later recanted all of his confessions. Appeal courts have ruled that the initial confession of sexual abuse was properly obtained. In a recent newspaper article about the Ingram case, psychologist Elizabeth Loftus was quoted as saying "it is entirely possible to take individuals and create wholly false memories in their minds." The newspaper article (Brad Shannon, "Ingram's son claims he was abused" The Olympian, Saturday June 8, 1996) reported that "Loftus said a scholarly journal this month reported on an experiment in which 90 percent of subjects would confess if you tell them someone else saw them do it." Loftus was apparently referring to an article by Saul Kassin and Katherine Kiechel published in the May 1996 issue of Psychological Science (volume 7, pp 125-128). The journal article reported a single experiment in which participants were accused of damaging a computer by pressing the wrong key. Although all participants were innocent, some participants came to confess to this "crime" after being given false incriminating information by a confederate. Reporting this finding in an article about Paul Ingram, would seem to equate accidentally and fleetingly hitting the wrong key while typing (something all typists have done) with a father repeatedly and intentionally raping his daughter over several years.

Researchers must make an effort to untangle the appropriate from inappropriate application of research results to this debate. While fabricated memories and recovered memories are both real phenomena, many people tend to tangle the two sets of issues into a hopeless snarl, so that evidence in support of memory distortions and errors is used to invalidate a particular recovered memory, while evidence in support of memory tenacity is used to validate a particular recovered memory. Since we know that an essentially true recovered memory is possible and that an essentially false memory is possible, then logically individual contested memories can only be adjudicated on an individual basis. A crucial untangling strategy for future research on general phenomena involves taking care to pose questions separately. When the research is disseminated, its relevance and its limitations must be carefully communicated.

See also:

• J Freyd's publications on Trauma and Memory

Last update Feb 1998 jjf@dynamic.uoregon.edu

Return to Jennifer J. Freyd home page