There is extensive evidence that memory is extremely susceptible to the influence of suggestion, postevent information, and source confusion (Belli & Loftus, 1996; Brainerd & Reyna, 1996; Conway, Collins, Gathercole & Anderson, 1996; Hyman & Pentland, 1996; McDermott, 1996; Mitchell & Zaragoza, 1996; Payne, Elie, Blackwell & Neuschatz, 1996; Read, 1996; Roediger, Jacoby & McDermott, 1996; Zaragoza & Mitchell, 1996). For instance, Zaragoza & Mitchell (1996) showed people a video of a burglary and then asked questions containing misleading suggestions, some of which were repeated; then, they tested memory for the source of the suggestions. Zaragoza & Mitchell (1996) found that, in comparison to subjects exposed to the suggestion only once, those exposed repeatedly were more likely to confidently remember the suggested events from the video and to claim they could recall witnessing the suggested events.
Despite the consistency and reliability of this type of finding, laboratory research has been rejected by some as irrelevant to the debate about recovered memory. For instance, Freyd & Gleaves (1996; Kristiansen, Felton & Hovdestad, 1996; van der Kolk, 1994; but see Roediger & McDermott, 1996) argued that laboratory research on memories of benign, artificially constructed stimuli tells us nothing about processes involved in memory for severely traumatic events. However, recognizing that memory is malleable does not mean recovered memories of childhood abuse are necessarily inaccurate; it simply means they are not necessarily accurate. Whereas there is clinical observation and personal anecdote that individuals may avoid or be unaware of threatening memories (Martinez-Taboas, 1996), such memories can be influenced by various cognitive and social events (Loftus, 1993). In other words, there is no strong reason or evidence why memory for traumatic events should follow entirely different psychological principles from those followed by memory for nontraumatic events (Kihlstrom, 1994, 1995); equally, memory should follow the same principles whether or not hypnosis is involved (Kihlstrom & Barnhardt, 1993). Notably though, at a biological level, the release of neural hormones during trauma may enhance the consolidation and storage of memory for that event (Cahill, Prins, Weber & McGaugh, 1994; but see Bremner, Krystal, Charney & Southwick, 1996). Also, the experience of trauma may be more likely to lead to the occurrence of recurrent and intrusive memories than to the forgetting of that experience (Frankel, 1994; LeDoux, 1991; LeDoux, Romanski & Zagoraris, 1989).
Discussion about the recovery of memory for traumatic events typically shows various views about repression (McConkey, 1997). In general, repression can be said to involve the motivated forgetting of information that is threatening to the self. Repression and related constructs such as dissociation, however, have been conceptualized in a variety of ways and that variation is one of the problems of the debate on recovered memory (Bowers & Meichenbaum, 1984; Lynn & Rhue, 1994; Singer, 1990; Spiegel, 1994). Notably, analyses of the original concept of repression in the writings of Freud have highlighted its internal inconsistencies and its limited value beyond a very general description of assumed processes (Holmes, 1974, 1990; Macmillan, 1997; for a summary of empirical work, see Pope & Hudson, 1995). The relative impreciseness of the concept of repression and the difficulty of testing it led Holmes (1990) to say that 'despite over sixty years of research ... there is no controlled laboratory evidence supporting the concept of repression' (p. 96). Further, he suggested that those who use the notion should warn that 'the concept of repression has not been validated with experimental research and its use may be hazardous to the accurate interpretation of clinical behavior' (Holmes, 1990, p. 97). Notwithstanding this view, clinical observation suggests that thoughts about and memories of important personal events can be set aside from normal awareness, and concepts such as repression or dissociation may be of heuristic value in helping to understand that process (Bower, 1990; Davis, 1990; Erdelyi, 1993; Nemiah, 1984; Weinberger, 1990). Although such concepts may have some value, this does not mean that if 'repressed' or 'dissociated' thoughts and memories are reported, then they are necessarily accurate representations. That is, accepting a notion of repression does not necessarily mean that recovered memories are historically accurate. Even if a memory has been forgotten for a time, this does not mean that it is not influenced by the constructive and reconstructive features of memories in general (Bowers & Farvolden, 1996).
We should not assume that freshly reported material indicates the lifting of repression that is linked to traumas of childhood. Rather, the fact that people sometimes remember events they had forgotten does not mean those events were traumatic, nor does it mean those particular memories were repressed. In other words, much of the nonreporting of such events may occur because of normal forgetting, embarrassment over reporting the events, the consequences of reporting the events, or various other reasons that relate to factors other than repression. In this respect, it does not seem to be possible to distinguish between people who do not report abuse and those who do not remember it; among the latter, it does not seem possible to distinguish forgetting that reflects repression, dissociation, other pathological processes, and benign processes (Kihlstrom, 1995). Nevertheless, when clinicians are faced with clients who experience themselves remembering a previously forgotten trauma, they must recognize the clinical relevance of this; equally, however, clinicians need to recognize that memories are affected by factors like suggestion, transference, personal values, social interactions, and fantasies associated with the event and its remembering (Nash, 1994).
Whatever their nature, it is clear that memories and the meaning placed on them change during therapy in various ways. For instance, Foa, Molnar, and Cashman
(1995) examined the memory reports of female rape victims during therapy, and found that their length increased across treatment, the percentage of reported actions and dialogue decreased, and the percentage of thoughts and feelings increased. There was an increase in the number of thoughts that attempted to structure the memory of rape. Thus, their narratives changed with the imaginal reliving of the trauma, and the victims tried to restructure their memory to provide a sense of coherence. That coherence may give a strong feeling of narrative truth and may feel right for both the client and the clinician, but it may not be an indication of the historical truth of the event. The fact that narrative and historical truth (Spence, 1982, 1994) may not coincide is nonproblematic and manageable by clinicians with relevant knowledge and skill. However, it may be problematic in nonclinical settings, such as the courtroom, in which the processes, goals, and demands are very different from the clinical setting. As Spiegel & Scheflin (1994) noted, it is possible to convince oneself of a false belief, and memory alone cannot be trusted in the absence of independent corroboration.
Questions about the trust that can be placed in recovered memory and the utility of such memory in clinical and court settings have led to research on whether memories of childhood abuse can be recovered. While recognizing that childhood sexual abuse can cause significant physical and emotional harm (Janoff-Bulman, 1992; Kendall-Tacket, Williams & Finkelhor, 1993; Nash, Hulsey, Sexton, Harral-son & Lambert, 1993; Romans, Martin, Anderson, O'Shea & Mullen, 1995), recovered memories of abuse cannot be seen as self-validating. Rather, the nature and accuracy of memories recovered during therapy need to be determined independently instead of being assumed by the client, the clinician, or others; this is especially the case when dealing with those therapies that may strongly bias the creation of illusory memories (Lindsay & Read, 1994). As Bowers & Farvolden
(1996) noted, however, the situation becomes complicated if clinicians accept abuse memories at face value; sometimes clinicians do this because they feel they must serve the client by confirming each of his or her ideas, memories, and beliefs. This tendency by some clinicians is unfortunate not only because it may lead clients to assume the validity of memories that may not be accurate, but also because it conveys that the clinician knows the truth about the client. As Bowers & Farvolden
(1996) argued, however, neither the clinician nor the client has definite knowledge of the reasons for a particular problem. Of course, the clinician can have more or less plausible theories regarding that problem, but these theories should not be confused with absolute truth, however compelling they may seem. In other words, clinicians need to deal with recovered memories in terms of their clinical utility, without focusing on the truth or falsity of those memories (Fowler, 1994).
Was this article helpful?