More Questions About Recovered Memories

A recent article (1) purported to “provide further evidence supporting the occurrence of amnesia for childhood traumatic experiences and the subsequent recovery of memory” (p. 754). In a key but methodologically problematic finding, “[a] majority of participants were able to find strong corroboration of their recovered memories” (p. 749). This “strikingly high” corroboration rate, however, was based on self-reported information recalled by the participants and accepted as recounted. Among 19 participants claiming complete amnesia who had attempted to confirm memories of sexual abuse, 89% (N=17) provided “corroborations” consisting of their memory of “verbal validation” alone (p. 753).

Although the authors acknowledged that a “major methodological limitation” of the study was the fact that “retrospective…self-reports were potentially subject to distortion and inaccuracies” (p. 754), there was no assessment describing the nature and quality of the self-reported corroborations, which would appear crucial to drawing conclusions about the veridicality of the recovered memories. Since retrospective verbal self-reports might have included pseudocorroborations representing confirmation bias, suggestion and belief paradigms, situational demand characteristics, and source amnesia (2, 3), the high corroboration rates could bespeak pseudomemories or screen memories masking other trauma (4).

Furthermore, even if “grossly improper therapeutic practices” (1, p. 754) were not a significant factor in memory recovery, unintended suggestive influences within the study itself may have biased the findings. Participants were asked “if there was a period during which they ‘did not remember that this [traumatic] experience happened’” (p. 751). With this question alone, the actuality of the traumatic experience was inherently validated by the investigators, and the experience of not remembering it was implicitly suggested. The fact that participants were recruited from a unit specializing in the treatment of posttraumatic and dissociative disorders could mean that suggestive influences and affiliative needs swayed group answers (3, p. 58). Questions about the “circumstances of first recovered memory” (1, p. 751) may have elicited autosuggestive responses. There apparently were no control questions or conditions. Ordinarily, patients might be confused about whether their recall of early traumatic experience is veridical (2, 4), yet the report does not indicate if participants ever had any doubt whether the events of the recovered memories actually occurred as remembered.

That “the vast majority of participants…did not recall any overt suggestion before the first recovered memory” (1, p. 752) does not rule out direct or indirect suggestive influence, whether inside or outside therapy sessions (2). Reading popular books, viewing or reading media, or talking with others on the subject of recovered memory may have influenced recollection. The actual time of suggestive effect could have followed the recalled time of recovery, which unwittingly may have been temporally displaced for narrative consistency.

These comments do not dispute the possibility of amnesia for traumatic experience that is later recalled or the discovery of information that confirms the veridicality of the memory. Consistent with suitable clinical technique, the retrieval of independent data is essential for investigating the objective-versus-subjective truth of early memories (5). However, without corroborative detail for readers to trace the study’s conclusions, generalizations about recovered memories hinging solely on self-reported “actual independent confirmation” (p. 753) should be viewed with scientific skepticism.