Keywords

The history of the recovered memory debate has led to a number of contentious issues, some of which there has been movement toward reconciliation (Belli, 2012, this volume), and others which continue to be in dispute. Based on the contributions to this volume, which present a comprehensive picture of the continuing views of notable scholars who continue to explore the nature of recovered experiences, I have settled on four contentious issues that seem most profound as barriers to a full reconciliation of the pertinent issues. None of these issues are new to the debate, although each has been impacted by the most recent relevant evidence.

One of these issues concerns the extent to which laboratory research can be generalized to the real world, and hence, the extent to which laboratory findings on false memories and motivated forgetting can be generalized to whether child sexual abuse (CSA) events can be falsely remembered and forgotten. A related issue concerns the impact of emotion on memory, and whether the emotional experience associated with CSA victimization at its inception and during its recovery leads to qualitatively different memory processing in comparison to events that are not as emotionally charged. A third issue, as noted in my volume introduction (Belli, 2012), pertains to the prevalence of false recoveries, with some stating that they are all too commonplace whereas others assert that they have only rarely occurred. As a fourth issue, I will explore different views of the impact of the sociopolitical context on the debate and how the intensity of the debate can be traced to a conflict between social and political ideals.

The Generalizability of Laboratory Research

Much of what has been debated in the so-called memory wars and its aftermath is the appropriate interpretation of the relevance of laboratory based research in providing insight on the development of true or false recoveries in the real world. At the height of the debate, concern was raised on whether laboratory work in false memories was relevant to the potential generation of false memories of CSA. Freyd and Gleaves (1996), for example, argued against generalizing laboratory demonstrations of false memories in the Deese-Roediger-McDermott (DRM) paradigm (Roediger & McDermott, 1995). Freyd and Gleaves pointed to two reasons for a lack of relevance: (1) the units of analysis in the DRM (words) differ from those in the real world (events), and (2) because childhood sexual abuse is implausible, and the related lures in the DRM are plausible, the DRM does not capture the level of relatedness between true and false memories that exists in the real world. More recently, Pezdek and Lam (2007) reiterated the notion that the DRM lacks relevance, as do other types of false memory paradigms, because these paradigms do not lead to the creation of false memories of entire events (see also DePrince, Allard, Oh, & Freyd, 2004).

Research by Geraerts and colleagues (Geraerts, 2012, this volume; Geraerts et al., 2009) challenges views that the DRM is not relevant to the creation of false recoveries of CSA. Geraerts and colleagues have shown that those who recovered memories via suggestive therapy were most susceptible to producing false memories in a DRM task in comparison to other groups. Although Geraerts interprets these results as being consistent with the likelihood that some individuals have a heightened propensity to develop false memories in both the laboratory and the real world, DePrince et al. (2012, this volume) question this level of generalizability. Instead of an increased susceptibility to creating false memories of entire CSA events, DePrince et al. consider the DRM results of Geraerts et al. as pointing to a heightened susceptibility in misremembering details of events, and point to the possibility that the suggestive therapy participants overall had true whole event or gist recoveries but may have remembered the abuse details as more positive than they were.

It is not only the generalization of laboratory-based false memories that have been questioned. Garry and Loftus (2005) questioned the generalization of laboratory-based research on motivated forgetting (see Anderson & Huddleston, 2012, this volume) on the forgetting of CSA events. Indeed, Anderson and Huddleston are sensitive not to generalize too readily; they recognize that their laboratory work cannot replicate the complexity, emotional content, and personal relevance of real CSA. Yet, Anderson and Huddleston also express optimism as their work is based on a model in which unwanted memories can be suppressed in the presence of constant cues, mirroring the tenet of BTT (Freyd, 1996; see also DePrince et al.,2012, this volume) that incestuous abuse will likely have more forgetting in comparison to stranger abuse despite the constant opportunity for remembering the CSA in the former case due to being in continual presence of the perpetrator.

In dealing with questions about the generalizability of laboratory-based research, it must be emphasized that experimental psychology is founded on the principle that well controlled laboratory studies provide a theoretical understanding of the operation of fundamental cognitive processes, and that these theories based on fundamentals are generalizable to the real world (Banaji & Crowder, 1989; Gallo, 2010; Wade et al., 2007). As revealed by the contributions of Johnson, Raye, Mitchell, and Ankudowich, (2012, this volume) and Anderson and Huddleston (2012, this volume), an understanding of fundamental cognitive and neural processes explains the development of false memories and the inhibition of unwanted memories, respectively. Pertaining to the DRM task directly, Gallo (2010) observes that a theoretical understanding of the fundamental cognitive and neural processes that lead to false memory generation in the DRM will be able to shed insight on the observation that individuals who have developed false memories in the real world–such as remembering a past life (Meyersburg, Bogdan, Gallo, & McNally, 2009)—are also more susceptible to the DRM illusion. Similarly, seeking a more thorough theoretical understanding of the operation of fundamental processes is likely to be the best arbiter in determining whether one should draw distinctions between generating false memories to whole events in comparison to developing false memories for the details of events.

The Impact of Emotion on Memory

Controversy regarding the impact of emotion on memory cannot be better illustrated by the differing perspectives of the contributors regarding the ability to remember traumatic experiences. For McNally (2012, this volume), traumatic experiences are never forgotten, and hence, true recoveries exist precisely because the abuse events, when experienced, were not traumatic (see also McNally & Geraerts, 2009). Although Brewin (2012, this volume) accepts that some experiences of child sexual abuse may not be traumatic, he shares the views of DePrince et al. (2012, this volume) that victimization is often associated with high levels of negative emotion including fright, embarrassment, betrayal, a sense of powerlessness, and concern for one’s well-being that are appropriately characterized as trauma. Whereas some of this difference in opinion can be accounted for by different meanings assigned to the term trauma—for McNally the term trauma is restricted to events that are life threatening whereas for Brewin and DePrince et al. their definition of trauma is broader—there are nevertheless clear distinctions in these points of view on how trauma impacts memory.

Brewin’s model considers that experiences with trauma that often accompany CSA lead to severe psychological consequences that directly impact the nature of cognitive processing. Trauma induces a fragmentation of the self so that the abuse experiences are often dissociated from the usual working self that interacts with daily life. Although involuntary remembering of abuse events governed by internal or external cues have the potential of bringing abuse events to mind, the fragmentation of self may actively inhibit their awareness entirely, or mute the awareness so as to not threaten one’s sense of well-being. Only when these traumatic experiences are able to become better integrated with the self that a full-blown recovery occurs, often characterized in a manner identical to the intrusive nature of PTSD flashbacks. In Brewin’s model, trauma’s impact is to promote structural abnormalities in the autobiographical knowledge base that promote extraordinary forgetting. In contrast to a special forgetting mechanism as proposed by Brewin, McNally considers that although CSA victims may not think about abusive events for many years, it is precisely because the abuse is not traumatic (in the sense of being life threatening) that it is open to the same level of lack of attention that would characterize other ordinary events that children experience which are confusing and unpleasant.

Experimental psychologists often prefer to explain all memory processes via an appeal to ordinary mechanisms as they are more parsimonious and introduce less skepticism than introducing special mechanisms (Lindsay, 1998). There is also historical precedent to a preference for ordinary processing, even when emotion is involved. At one time it was widely held that flashbulb memories that accompany emotionally provoking culturally-shared tragedies such as the assassination of John Kennedy or the space shuttle Challenger explosion were remarkably accurate for an extended period of time because of special encoding mechanisms that imprinted the events into memory (Brown & Kulik, 1977). More recently, Talarico and Rubin (2003; see also Neisser & Harsch, 1992) have shown that flashbulb memories suffer from the same errors of omission and commission as memories of ordinary events. What is remarkable about flashbulbs is that people believe in their accuracy, which can be traced to a heightened sense of vividness in the details of what is remembered (even when these details are wrong) as predicted by the Source Monitoring Framework (see Johnson et al., 2012, this volume, on the role of ­emotion in memory errors).

Of course, the example of flashbulb memories may not characterize all emotional experiences. But whenever there are differences between the manner in which emotional and nonemotional experiences are remembered and forgotten, the key question is whether emotion impacts memory in a qualitatively different way, as suggested by an appeal to special memory mechanisms, or in a merely quantitative way by exaggerating how ordinary nonemotional processes operate (e.g., by adding to the vividness of details). Complicating the picture is that conjectured processes that have often denoted special memory mechanisms, such as repression and dissociation, are also viewed as having parallels in ordinary cognitive processing (Brewin, 1997). The dissociation that Brewin (2012, this volume) implicates in contributing to the fragmented self in someone exposed to trauma is an exaggeration of the multiple selves that all people maintain and which will constrain what information from the structure of autobiographical knowledge is most accessible given whatever aspect of the working self is most activated at any point in time (Conway, 2005). What is unclear is whether this exaggeration is a qualitative change, or one that can be viewed as an extension of ordinary cognitive processes.

Research reviewed by Anderson and Huddleston (2012, this volume) on retrieval inhibition provides some insights. Although retrieval inhibition can occur without emotion, in comparisons between emotional and neutral stimuli, some studies reveal greater retrieval inhibition for the emotional items (e.g., Depue, Banich, & Curran, 2006). Because a quantitative explanation would be hard pressed to find an ordinary mechanism that would lead to increased forgetting for stimuli that are more vivid or distinctive, a special memory—or forgetting—mechanism is suggested. In finding retrieval inhibition for negative but not for positive stimuli, Lambert, Good, and Kirk (2010) considered their results as supporting a “repression hypothesis.” However, as there are equivocal results in the research that has explored retrieval inhibition for emotional items (see Anderson & Huddleston), any conclusions regarding the potential presence of a special retrieval inhibition forgetting mechanism for negative items are premature.

The Prevalence of False Recoveries

Among the volume contributors, the contribution of DePrince et al. (2012, this volume) is the only one to explicitly challenge the notion that a substantive proportion of recovered memory experiences are false. They point to two issues. First, they consider that any application of suggestive techniques in therapy has been implemented by ill trained therapists and hence, the prevalence of the use of these techniques is quite low. Second, and especially in their assessment of the finding of Geraerts and colleagues (see Geraerts, 2012, this volume), they reason that there is no solid evidence that the use of suggestive techniques will lead to false memories of CSA. Their views are in direct contrast to Geraerts and also to Johnson et al. (2012, this volume) as these contributors consider that suggestive techniques have been used all too often, and that their use can lead to false recoveries.

A number of pieces of evidence point to suggestive therapeutic techniques as leading to false memories of CSA. During the late 1980s and early 1990s there were a number of publications in the professional literature (e.g., Claridge, 1992; Courtois, 1988, 1992; Dolan, 1991; Ellenson, 1985; McCann & Pearlman, 1990) and self-help books targeted to lay audiences (e.g., Bass & Davis, 1988; Blume, 1990; Engel, 1989; Fredrickson, 1992) that had advocated the use of memory recovery ­techniques. Anecdotes, including some from court cases, emerged during this time in which memories that were recovered with encouragement of therapy invited skepticism for a number of reasons. Some of these memories depicted events at very early ages—such as before the age of 1½ years—in which the abuse was alleged to have occurred, some portrayed an amnesia so dense that repeated brutalizations across decades and into young adulthood had been forgotten, and some involved countless perpetrators in conspiracy including satanic cults and infanticide that have never been documented (Ganaway, 1989; Loftus, 1993, 1997; Loftus & Ketchum, 1994; Ofshe & Watters, 1994; Wagenaar, 1996). Considerable experimental work has shown that false memories of holistic childhood events can be created in controlled laboratory conditions that mirror the kinds of memory recovery techniques that were being illustrated in the clinical literature and self-help books (e.g., Hyman, Husband, & Billings, 1995; Lindsay, Hagen, Read, Wade, & Garry, 2004; Mazzoni & Memon, 2003). In the latter half of the 1990s, trauma therapists, some of whom had once condoned the use of memory recovery techniques, noted a period of time in which “clinical excesses and errors” (Courtois, 1997, p. 342) had occurred, or acknowledged that a “minority of therapists have used questionable ‘memory recovery’ techniques” (Briere, 1997, p. 26), and hence put forward a set of guidelines designed to minimize the occurrence of false memories during trauma therapy (see especially Courtois, 1999; Lindsay & Briere, 1997).

Surveys of licensed therapy practitioners in North America have also revealed that a substantial minority have used suggestive techniques among clients who had been suspected of being victims of child sexual abuse (Legault & Laurence, 2007; Polusny & Follette, 1996; Poole, Lindsay, Memon, & Bull, 1995). Supporting the notion that using suggestive techniques leads to false memories, Legault and Laurence and Poole et al. did find modest correlations between the number of techniques used and rates of recovered memories of CSA during therapy. In terms of estimates of the prevalence of recovered memories in therapy, these surveys provide estimates of between 20% and 40% of clinicians who had at least one client recover a memory of CSA during the past year, and Legault and Laurence found that their respondents had reported a mean of 4.3% of clients with recovered memories during the past 2 years. In a survey of a U.S. national probability sample of women, Wilsnack, Wonderlich, Kristjanson, Volgentanz-Holm, and Wilsnack (2002) found that among respondents reporting a recovery experience of CSA, approximately 8% (unweighted) recovered their memory during the course of professional treatment (see also DePrince et al.), with the remainder having recovered spontaneously (recovery on one’s own). Hence, at this point in time, the available surveys of clinicians and the general public do not provide a consistent picture of the extent to which memories are recovered in the context of therapy. Differences among surveys are likely a function of estimation errors including question wording, respondent characteristics, and considerable sampling error when few data points are available (see Groves, 1989, for a review of estimation errors in surveys). Further, although the surveys of practitioners indicate that suggestive memory recovery techniques are used surprisingly often, and that there is an association between their use and the occurrence of recoveries during therapy, these surveys provide no direct evidence of memory recovery techniques leading to false memories.

With regard to question wording, the survey of Wilsnack et al. (2002) is potentially problematic in not adequately accounting for persons who believe that they were victims of CSA, but who have no explicit memories for the abuse (see McNally, 2012, this volume, on uncovering research participants who fit this description). Such beliefs may be false. After screening respondents with a question on whether they had felt they had been a victim of CSA with a family member as a perpetrator, respondents in Wilsnack et al. were then asked to categorize their abuse, in a mutually exclusive fashion, as having always been remembered, or having been recovered on one’s own or with the help of a professional (among other categories that none of the respondents had endorsed). It appears likely that any persons who had a belief that they were abused with no explicit memories would have endorsed the category indicative of a spontaneous recovery.

Within this context of uncertainty regarding the prevalence of false recoveries of CSA, several points are deserving of attention. First, even if the prevalence of false memories that result from suggestive therapy is very low, given the large numbers of persons in the population, the total number of persons with false memories would still be quite large (Lindsay, 1997 ). Second, the available evidence supports the conclusion that a considerable majority of persons who are victimized by CSA have some level of continuity in remembering their abuse, and hence, fully recovered memories of CSA events—whether true or false—constitute a minority of cases. The reasons for fully recovered memories being rarer than continuous memories may be a function of individual differences; only certain individuals may have the necessary cognitive control to forget abuse, or to be highly susceptible to suggestions, as indicated by Anderson and Huddleston (2012, this volume) and Geraerts (2012, this volume), respectively. Finally, as noted above, there is consensus, at least among the contributors to this volume, that any use of suggestive techniques in therapy is an inappropriate practice and that memory recovery should not be a goal of trauma therapy.

The Impact of the Sociopolitical Context on the Debate

As I had noted in my introduction to this volume (Belli, 2012), the recovered memory debate is a topic most apt for the Nebraska Symposium on Motivation not only because of the critical role of motivation in underlying the cognitive mechanisms responsible for both true and false recoveries, but also because the evidence and arguments that practitioners and scientists have offered to the debate are fueled by motivational and ethical concerns. DePrince et al. (2012, this volume) are very explicit about the motivations that guide their orientation. In contrasting privileged versus marginal voices, DePrince et al. portray the tragedy of CSA as a continuing vestige of a patriarchal culture in which dominance still largely resides among adult males. Victims of CSA, primarily but not always girls, have marginal voices to which perpetrators merely ask that we do nothing, that we keep the voices of their victims quiet. Extending the notion of the sociopolitical context as one in which certain groups are more privileged than others, DePrince et al. observe that scientific voices are also privileged ones, and that the authoritativeness that derives from these voices can either help to empower those voices in the margins, or can further discourage their being heard.

A controversial implication of DePrince et al.’s analysis regarding the sociopolitical context, and the role of scientific thought in either legitimizing or diminishing those victimized by CSA, is that those who have offered evidence and arguments in favor of false recoveries are engaging in practices that further the injustices that penetrate our culture. This implication is not new, and in the history of the recovered memory debate, advocates of the false recovery position have been sensitive to this critique and have offered their own perspectives regarding it. It must be emphasized that this implication cannot be effectively countered by the denunciation of CSA—indeed, the vast majority of those who have weighed in on the side of the false memory position do acknowledge CSA as being disturbingly common, tragic, and morally reprehensible—as what is implied is that advocating the likelihood of false memories is to provide an excuse for some to discredit reports of CSA more generally. Similarly, despite the recovered memory debate as not involving any attempt by researchers to question the fundamental accuracy of the reports of those who have continuously remembered being victimized by CSA, there continues to be concern that skepticism targeted to the veracity of any proportion of recovered memories can lead to some people to doubt the accuracy of continuously remembered abuse as well.

As one countering theme, Belli and Loftus (1994) argued that any skepticism that arises in the veracity of victims’ reports has its source in the extraordinary and seemingly impossible abusive events that had been recovered with the assistance of therapy. Hence, the onus of diminishing the voices of those victimized by CSA is not on those who warned about the dangers of suggestive therapy, but on those who engaged in suggestive therapeutic techniques. Another common theme was to assert that both CSA and false recoveries of abuse were tragic; with regard to the latter, the tragedy resides in both the needless suffering among those who had falsely recovered and in the needless endangering of the health of family relationships (e.g., Belli & Loftus; Lindsay & Read, 1994; Yapko, 1994). And yet another countering approach has been for advocates of the false recovery position to emphasize that the debate is one involving the properties of memory. Loftus (1997), for example, has asserted that the debate is not one “about the reality or the horror of sexual abuse, incest, and violence against children,” rather it is a “debate about memory” (p. 176).

Although each of these countering positions has merit, they do not directly address the very reasonable argument expressed by DePrince et al. that the sociopolitical context does impact psychological science in profound ways (see also Freyd & Quina, 2000; Pezdek & Lam, 2007). Foremost, it should be acknowledged that social and political concerns drive which observations in the real world are deserving of scientific attention in the pursuit to uncover fundamental psychological processes. As noted by Wade et al. (2007):

…Let us not forget that psychological scientists study false memories because we have looked in the real world and see what happens. Psychological science tries to understand behavior out there by bringing it into the laboratory, not the other way around. This reductionism is, of course, typical for other scientific disciplines as well. For example, it would be nonsensical to argue that the research molecular biologists carry out on HIV is irrelevant to AIDS in real patients (p. 26, emphasis in the original).

One of the real world concerns that have led to considerable scientific research into false memories has been to expose the dangers of eyewitness unreliability with the hope of stemming the injustice that follows from wrongful convictions (McMurtrie, 2007). In the context of the recovered memory debate, any false accusations of CSA are troubling in and of themselves. Yet, they become even more damaging when introduced as evidence into a criminal investigation or entered as evidence into a court of law. There can be no doubt that the recovered memory debate would not have become so heated if it weren’t because criminal accusations were being made on the basis of events that had not been remembered for many years, and which were apparently stimulated in the context of therapy (Loftus, 1993; Read & Lindsay, 1994). One of the principal threats to realizing the ideals of a free society are the dangers of false criminal accusations, or even more dramatically, the power of the state to mistakenly confine (or execute) citizens who are innocent of any wrong doing.

With regard to the impact of the sociopolitical context on the recovered memory debate, its influence led to two sources of injustice coming into conflict, one that emphasized inequalities between genders and groups, and one that observed that safeguards to the ideals of a free society were being challenged by the yet unforeseen overreliance on the reliability of eyewitness memory. Although DePrince at al. are correct regarding the injustices that have existed, and continue to exist, within a culture that has been dominated by adult males, and how scientific investigation cannot escape from this culture, gender or group power differences are not the only source of injustice that pervades our society. Within the sociopolitical context, both CSA and wrongful criminal accusations (and convictions) go beyond individual tragedies. They are both social tragedies as the occurrence of either threatens social and political ideals, ideals that are worth defending in the name of justice.

The Current Status of the Debate

As illustrated by the contributions to this volume, since the height of the so-called memory wars there has been considerable research that has provided valuable information relevant to understanding recovery experiences. There is movement toward a reconciliation of points of view as seen by a consensus—at least among the volume contributors—that a substantive proportion of recovered memories of CSA are authentic representations of actual abuse. Although there is continuing disagreement on the prevalence of false recoveries, there is consensus that suggestive ­therapeutic techniques are dangerous in having the potential to promote false ­memories and that memory recovery should not be a goal of trauma therapy.

Yet, as also seen by the contributions to this volume, there are continuing points of contention that reflect many of the same arguments which have been made throughout the history of the debate. It also must be recognized that the call for civility and the need to move toward a middle ground are not new (see especially Lindsay & Briere, 1997). The reasons for resistance to a fuller reconciliation largely reside in social ideals that have conflicting perspectives regarding how the debate should be framed, what research questions should be pursued, and how research findings should be interpreted.

With regard to social ideals, one may be tempted to directly compare the blight of CSA against the prevalence of therapy-induced false recoveries as to which is more problematic. With this direct comparison, there is no doubt that CSA is a far more egregious problem; abuse has had a longer history and considerably more people have been affected by it (Lindsay & Read, 1994). Yet, such a comparison ignores the larger issues that are raised by those social ideals that encompass the debate. Both CSA and therapy-induced illusory memories are symptomatic of much broader social and political concerns, with the former, equality among groups, and with the latter, due process protections from false accusations. From the perspective of this broader sociopolitical context, one can see that issues of inequality and imperfections in due process are both ancient, and both have adversely affected countless persons.

Because the recovered memory debate is embroiled in frustrations that people are experiencing in realizing worthy social ideals that have happened to come into conflict, achieving a complete reconciliation will be difficult, but not impossible. Hopefully, continuing progress toward reconciliation will occur, but the future may also witness a retrenchment into more divisive positions. For reconciliation to continue to move forward, all must be able to provide a fair assessment of the available evidence, appreciate the reasons that underlie different points of view, and project a civility in tone.