Recovered Memories in the Courtroom

Chapter in Truth in Memory (Hardcover) (Stephen Jay Lynn and Kevin M. McConkey, Editors)

Ralph Underwager and Hollida Wakefield

The rules of the courtroom have changed.

Faced with the proffer of expert scientific testimony, the trial judge must determine at the outset, pursuant to Rule 104 (a), whether the expert is proposing to testify to (1) scientific knowledge that (2) will assist the trier of fact to understand or determine the fact at issue. This entails a preliminary assessment of whether the reasoning or methodology underlying the testimony is scientifically valid and of whether that reasoning or methodology properly can he applied so the facts in issue. We are confident that federal judges possess the capacity to undertake that review. (Daubert v. Merrell Dow Pharmaceuticals, Inc., 1993)

The ruling has two concepts. The first supersedes the Frye (Frye V. United Slates, 1923) standard by Federal Rules of Evidence. Rule 702 states that testimony is admissible if the scientific knowledge of the expert will assist the trier of' fact. However, the second concept of the Daubert ruling is that the substance of the testimony must qualify as "scientific knowledge." The ruling then goes on to give requirements for determining what is scientific knowledge. This is most crucial to understanding the way claims of recovered memory are likely to be treated in the courtroom. Now, under Rule 702, the judge's task is to screen out testimony that cannot be described as scientific knowledge. Then a methodological definition of scientific knowledge given, rather than any particular body of propositions (Imwinkelried, 1996). The trial judge has this responsibility. "This entails a preliminary assessment of whether the reasoning or methodology underlying the testimony is scientifically valid and of whether that reasoning or methodology properly can be applied to the facts in issue" (Daubert, 1993, p. 12).

This Supreme Court decision constitutes a paradigm shift in what the justice system accepts as scientific evidence (Underwager & Wakefield, 1993). The decision adopts the position of Sir KarI Popper's philosophy of science and, thus, repudiates logical positivism and the naive view that science proves what is true. Instead, as the first consideration, it adopts the statement of Popper that the criterion for science is falsifiability, refutability, or testability: "Ordinarily, a key question to be answered in determining whether a theory or technique is scientific knowledge that will assist the trier of fact will be whether it can be (and has been) tested ... ('The criterion of the scientific status of a theory is its falsifiability, or refutability, or testability')" (Daubert, 1993, pp.12-13).

The second consideration is whether there has been peer review and publication. The third is the known or potential rate of error and the standards of controlling the technique's operation. Finally, there may be some significance in the level of acceptance in the scientific community. The Supreme Court clearly wants fact finders to have available credible scientific knowledge to snake the most accurate decisions. Therefore, it requires trial judges to think like scientists in order to select admissible evidence (Thomas, 1994).

Although general acceptance in the scientific community (the Frye test; Frye V. United States, 19231) is retained as one consideration, the lack of such by itself does not preclude the proposed testimony. This will make admissible new scientific evidence that was excluded under Frye. At the same time, if properly understood and followed, Daubert should render inadmissible "junk science" in the courtroom (Underwager & Wakefield, 1993). Scientific theories must always be able to make predictions in such a way that they can be falsified, that is, shown to be wrong. Unscientific theories are those, such as Marxism and Freudianism, that explain everything. Nothing can occur that cannot be accommodated by these theories; therefore they are not falsifiable. Science advances by the process of producing hypotheses about how things work, finding a way to test them, and discarding those that do not work out. The number of times a theory has been supported is not the crucial element. Although there may be a large number of supportive studies, a single credible and well-done study that falsifies the theory means the theory must be discarded (Meehl, 1967). The goal now must be to demonstrate that the proposition rests on sound scientific procedure. This involves the factors of falsifiability and level of error as direct evidence while the publication and general acceptance are circumstantial evidence only (Imwinkelried, 1993).

After the Daubert decision on June 28, 1993, there were 40 cases through May 18, 1994, that reached the reporting level in which the new rules establishing what is admissible as scientific evidence were applied (Bernstein, 1994). In most of the 40 cases the courts used Daubert to reject scientifically unsound evidence. Although most of these have involved toxic tort and traditional tort cases, there were four cases in which Daubert was applied to the social sciences, that is, testimony by mental health professionals, psychiatrists, psychologists, and social workers. Testimony based on weakly supported psychological theories has been excluded (e.g., the use of posttraumatic stress disorder and the child abuse accommodation syndrome to buttress, by inference, claims of prior abuse). Because of the recognition of suggestibility, hypnotically refreshed testimony has been excluded.


Beginning in the late 1980s, claims were made that childhood sexual abuse was banished from consciousness through some type of active filtering process termed "repression" or "dissociation." This alleged abuse was often highly traumatic and intrusive, but the victim had no memories until adulthood. The memories generally returned with the aid of a therapist or after reading a sexual abuse survivors' book (e.g., Bass & Davis, 1988; Blume, 1990; Dolan, 1991; Fredrickson, 1992).

These recovered memories claims have created fierce controversy and massive polarization in the professional community. Although some professionals attempt to take a "middle" position, this is difficult to do and those involved are generally on one side or another; A task force established by the American Psychological Association with three researchers and three therapists (Alpert et al., 1996) was unable to reach a consensus on the question of repressed memories and recovered memory therapy.

In 1992, a small group of parents whose adult children accused them of abuse based on recently recovered memories and several sympathetic professionals formed the False Memory Syndrome Foundation (FMSF) in Philadelphia.2 It now has a large professional advisory board, has put on scientific conferences, and disseminates information concerning claims of repressed or dissociated memories and the techniques often used to recover memories. Recovered memory advocates have reacted to the FMSF with anger and defensiveness and have attacked the organization and its director and advisory board in books, newsletters, and conference presentations. The claims of recovered memories as well as the skeptical responses have received significant media attention.

As with all disputes about facticity that cannot be resolved elsewhere in our society, many of these allegations end up in the courtroom. One party claims to have recovered memories of childhood abuse and the other party denies ever committing any abuse. The new rules for scientific evidence are directly applicable to these claims and counterclaims. Therefore, the quest ion becomes which party has the sustainable claim to scientific knowledge, methodology, and techniques under Daubert.

Approximately 200 cases involving claims of recovered memories of childhood abuse have now reached the appellate level. Some have been ruled on while others are still at the appeal level. Four legal postures have emerged from the appellate rulings on recovered memory claims. First, the discovery rules (i.e., the date a person discovers an injury) may not apply to recovered memory claims. Then, recovered memory claims do not extend the statute of limitations as a statutory disability. Next, independent corroboration is required to apply the discovery rule. Finally, and most recently, the reliability of the repressed memory theory must be determined before any extension of the statute of limitations. A pretrial hearing on admissibility is where the Daubert standards of scientific knowledge are most likely to be applied to recovered memory claims. The history of appellate rulings gives some indicati6n of which direction is likely to emerge from such hearings.

In the first repressed memory case to reach the appellate level, the Washington Supreme Court (Tyson v. Tyson, 1986) did not extend the discovery rule to repressed memories because of the lack of "empirical verifiable evidence of the original wrongful act and the resulting injury." The court reasoned that when an alleged recollection of events repressed from consciousness had no way to be independently verified, "the potential for spurious claims would be great and the probability of the court's determination of truth would be unreasonably low" (p. 230). The Michigan Supreme Court (Lemmerman v. Fealk, 1995) also refused to extend the discovery rule because it would be very difficult to resolve the issue "given the state of the art regarding repressed memory and the absence of objective verification." The weakness of scientific support for claims of repressed memories has also been noted by the supreme courts of New Hampshire, Wisconsin, Alabama, Michigan, Texas, and Maryland (Pritzlaff v. Archdiocese of Milwaukee, 1996; Travis v. Ziter, 1996; Lemmerman v. Fealk, 1995; S.V. v. R.V, 1996; Doe v. Maskell, 1996; New Hampshire v. Hungerford/Morahan, 1997; New Hampshire v. Walters, 1997), and appellate courts in Illinois, Tennessee, California, and North Carolina (M.E.H. v. L.H., 1996; Hunter v. Brown, 1996; Engstrom v. Engstrom, 1997; Barrett v. Hyldburg, 1997).

Trial courts in Pennsylvania, North Carolina, and New Hampshire (Pennsylvania v. Crawford, 1996; New Hampshire v. Hungerford, 1995; Barrett v. Hyldburg, 1996), as well as others, have held that claims of recovered memories cannot proceed because of the lack of scientific support for the concepts advanced. The Daubert standards were cited by some trial judges as in North Carolina (Barrett V. Hyldburg, 1996), where the court stated, "This court is of the opinion, considering all of the evidence that has been presented... that the evidence sought to be introduced is not reliable and should not be received into evidence in this trial." The Rhode Island Supreme Court (Kelly V. Marcanonio, 1996) ruled that the scientific reliability and validity of repressed memory theories must first be determined before a trial or extension of discovery. Requiring a judgment about the scientific status of recovered memory claims is the most likely direction for the courts to go. Given the indications of decisions made thus far, when a Daubert analysis is used to determine the scientific status of repressed memory claims, such claims are likely to be found wanting and not allowed in the courtroom. Several trial courts and appellate courts ruled that the statute of limitations applies when a plaintiff has always known of the abuse but asserts lack of knowledge of the damage done (see Marshall V. First Baptist Church of Houston, 1997; John BBB v. Archdiocese of Milwaukee, 1997).

But the final implications of Daubert are not clear because judges will have great latitude in determining whether evidence meets the criterion for admissibility. Some judges may be mistaken, uninformed, or biased. Dershowitz (1982) comments, "Most judges have little interest in justice. ... They want to make sure criminals are convicted and sent away ... many judges will do everything within their lawful power — and some things beyond it — to convict defendants who they believe should be in jail. ... Beneath the robes of many judges, I have seen corruption, incompetence, bias, laziness, meanness of spirit, and plain, ordinary stupidity (pp. xvii-xviii). Huff, Rattner, and Sagarin (1996) report finding a systemic error in the justice system. Higher courts do not reverse lower courts but systematically incorporate their errors into the case law system. "The further a case progresses in the system, the less chance there is that an error will be discovered and corrected, unless it involves a basic issue of constitutional rights and due process" (p.144).

The ruling of the U.S. Court of Appeals for the Second Circuit (Borawick v. Shay, 1995) establishes a legal precedent for interpreting the Daubert decision guidelines by considering the totality of the circumstances in determining admissibility of evidence. This was a recovered memory of childhood abuse claim. While the ruling affirms the district court's summary judgment in the specific case, it leaves open an exercise of judicial discretion that permits evasion of the Daubert ruling's guidelines for admissibility of scientific evidence. Errors by a biased or incompetent judge claiming to consider the totality of the circumstances and admitting evidence with no scientific validity or credibility would be very difficult to reverse. This may effectively prevent the application of the Daubert ruling to prevent junk science from dominating the courtroom. Given the high level of emotional involvement of all participants when there is an allegation of sexual abuse, this may be the type of case in which the potential risk of error is greatest.


The shift represented in the Daubert ruling exposes the science of psychology to what may turn out to be startling influence by the judicial system. With judges passing on the nature of scientific knowledge, scientific methodology, and what is or is not scientific, it is almost assured that what is affirmed by judicial decisions will gain prestige and status whereas what is rejected will lose out. Theories, procedures, and concepts judged to be unscientific may well be vulnerable to malpractice actions, ethical complaints, and loss of insurance coverage.

The period when scientists alone determined what was science and what ideas could be pursued is over. Now if there is any hint of a judicial rejection of scientific recognition, the future of an idea is likely pretty bleak. As science entered the courtroom, it became reasonable for the justice system to hold putative scientists and their testimony to standards of accountability and responsibility. Nobody really wants pseudoscience to be on an equal footing in the courtroom with responsible and empirically supported science.

At the same time, ethical positions taken by the American Psychological Association strongly urge psychologists to maintain the autonomy of the science and the profession. The Committee on Professional Standards (1981) states, "As a member of an autonomous profession, a psychologist rejects limitations upon his [or her) freedom of thought and action other than those imposed by his (or her] moral, legal, and social responsibilities" (p.650 17. 19).

When judges make decisions regarding the admissibility of expert testimony, they cannot avoid becoming participants in a particular construction of scientific facts. Judges can, and most likely, will shape an image of science that is affected by their own preferences and biases as to how the world should work. A recent ruling by the Court of Appeals for the Eighth Circuit (United States v. Rouse et al., 1996) overturned a verdict because the trial judge had inappropriately and mistakenly applied the Daubert standards to the admissibility of expert testimony.3 The process of bringing Daubert to bear on the science of psychology is one that psychologists need to observe carefully and thoughtfully and interact with the judiciary in any way possible. The controversy around claims of recovered memory offers the first opportunity to do so.


Proponents of recovered memories often maintain that from one-third to one-half of all women were sexually abused in childhood. Although the various studies of prevalence yield widely divergent estimates (Miller, Johnson, & Johnson, 1991; Salter, 1992), the higher figures of Russell (1983) and Wyatt (1985) are often cited as support for this claim. (These high prevalence estimates have been sharply disputed by Gilbert, 1991, 1994, and Okami, 1990.) Discussions of prevalence statistics also often include assertions that many instances of child abuse are unreported and many victims never tell anyone. Some professionals believe that up to half of adults who were abused as children have amnesia for their abuse (e.g., Blume, 1990; Demause, 1991; Gleaves, 1994; Maltz, 1990; Summit, 1990).

Advocates for recovered memories claim that "survivors" have a variety of symptoms that indicate abuse even when the person has no memories (e.g., Blume, 1990; Dolan, 1991; Fredrickson, 1992). (See Kihlstrom Chapter 1, this volume, for a discussion of checklists of symptoms.) They believe that sexual abuse survivors must retrieve their memories so they can process the trauma and eliminate their symptoms (Bass & Davis, 1988; illume, ] 990; Courtois, 1992; Dolan, 1991; Everstine & Everstine, 1989; Fredrickson, 1992). Memory recovery work is described as the basic healing force: "The bulk of your repressed memories need to be identified, retrieved, and debriefed for healing to occur" (Fredrickson, 1992, p.223). Without retrieving memories, the person cannot heal and recover (Courtois, 1992).

The recovered memory assumptions are supported by referring to one or more of several concepts. The person is said to have "repressed" the memory because it was too painful, or to have "dissociated" during the abuse as protective mechanism. The victim may have developed "traumatic amnesia" for the abuse and, if the abuse was severe and repeated, may have formed "alter personalities" and will develop "multiple personality disorder;" It is believed that, although not available to conscious memory, childhood abuse can be observed through a variety of later problems and "body memories," "flashbacks, or "nightmares" (see Wakefield & Underwager, 1994b, for a detailed discussion of these concepts).

Vague feelings about abuse are interpreted to mean that the person has been abused but cannot remember it. Demands for details or corroboration are seen as unreasonable and, once the person begins "recovering" memories, the veracity of these memories is seldom questioned. In The Courage to Heal, sometimes referred to as the "Bible" of the recovered memory movement, Bass and Davis (1988) state:

"If you think you were abused and your life shows the symptoms, then you were" (p.2).

"If you don't remember your abuse you are not alone. Many women don't have memories, and some never get memories, This doesn't mean they weren't abused" (p.81).

"You are not responsible for proving that you were abused" (p.137).

Information about the basic assumptions and therapy that elicits recovered memories comes from several sources. Descriptions of the type of treatment offered are found in the reports of people who have undergone such treatment (Goldstein & Fanner, 1993; Nelson & Simpson, 1994; Wakefield & Underwager, 1994b). Several therapists describe their techniques in books, articles, and workshop presentations (e.g., Bass & Davis, 1988; Blume, 1990; Courtois, 1992; Dolan, 1991; Fredrickson, 1992; Grand, Alpert, Safer, & Milden, 199 1; Lundberg-Love, 1989, n.d.). There have been instances in which a journalist joins a survivors" group or goes to a therapist and records what happens (e.g. Nathan, 1992) or a parent sends a private investigator to their adult child's therapist to act as a pseudopatient (Wakefield & Underwager, 1994b). In a questionnaire project conducted by the FMSF, lengthy questionnaires were sent to people whose adult children accused them of recovered memories of repressed childhood sexual abuse (Freyd, Roth, Wakefield, & Underwager, 1993; Wakefield & Underwager, 1992, 1994b). Many of the respondents were able to provide information about the type of therapy their adult child received.

Many different techniques may be used to help patients recover memories of sexual abuse. These include direct questioning, hypnosis, age regression, reading survivors' books, attending survivors' groups, free association, massage therapy, dream interpretation, ideomotor signaling with the unconscious, and expanding on imagistic memories. In the FMSF questionnaire, respondents also were aware of a variety of other unconventional techniques, including prayer, meditation, neurolinguistic programming, reflexology, channeling, psychodrama, casting out demons, yoga, trance writing, and primal scream therapy.

After memories are retrieved, the "survivor" may be encouraged to express her rage at the perpetrator in a variety of ways, such as throwing darts at his photograph, writing him angry letters, or confronting him during a family gathering. The patient is often referred to a survivors' group, where further abuse memories may develop in response to group norms and influence. The goal of the therapist is to be accepting, reassuring, encouraging, and validating of the disclosures.

The number of therapists accepting these assumptions and techniques appears to constitute a significant minority. Poole, Lindsay, Memon, and Bull (1995) found that 71% of three random samples of doctoral-level therapists taken from the National Register of Health Services Providers in Psychology (NRHSPP) in the United States and the Register of Chartered Clinical Psychologists (RCP) in Britain used techniques to help clients recover suspected repressed memories of sexual abuse. Out of their total sample of 202 therapists, 25% reported a constellation of beliefs and practices suggestive of a focus on memory recovery, and this latter group reported relatively high rates of memory recovery in their clients. If the Poole et al. (1995) sample is typical of doctoral-level clinical psychologists, there are thousands of clinical psychologists performing recovered memory therapy. Poole et al. (1995) estimate that the number of clients working with recovered memory-focused therapists from the NRHSPP and the RCP during the time covered by the survey would exceed 100,000. This number does not include other psychologists, social workers, psychiatrists, and assorted "counselors." Poole et al. (1995) observe that their findings argue against the claim that these therapists are a small group of uncredentialed and untrained therapists. Dawes (1995) notes that using the Poole et al. estimates, a conservative estimate is that in a 2-year period, 1,475,833 women have seen therapists who use two or more recovered memory techniques.

More recently, some clinicians have shifted toward a less dogmatic position and are more open to discussion and modification of previous positions. For example, in 1992, Courtois recommended hypnosis, guided imagery, drawing, guided movement, body work, and psychodrama for retrieving memories. But in 1996, she acknowledges that some of the techniques previously used in recovering memories may not be well supported and states that such techniques are used mostly by untrained and unlicensed therapists. She now describes a consensus on treatment that is phase oriented and aimed at healing rather than full remembering or reexperiencing the trauma and abreaction, and she declares that therapists must pay careful attention to social influence and risk factors that may generate erroneous beliefs. There is no mention of the techniques she advocated in 1992.

In 1991, Alpert was a presenter in an APA symposium (Grand et al., 1991) in which the role of the therapist was declared to be to help the patient become convinced of the historical reality of the recovered memory of abuse even when there was no external corroboration and even when the patient doubted the memory was real. Techniques of body work, dream analysis, imaging, and the like were touted. But in 1996, Alpert states that the "search for buried memories is not promoted by professional programs in psychology or by the mainstream professional literature on treatment of adult survivors.... There is no training program or mainstream literature that presents memory retrieval to the exclusion of other therapeutic tasks as the treatment goal, or that promulgates the utilization of techniques that are suggestive" (Alpert, 1996, p.328).

Alpert and Courtois were members of the American Psychological Association's working group on investigations of memories of childhood sexual abuse (Alpert et al., 1996). In the three clinicians' report, Alpert, Courtois, and Brown4 conclude that the majority of sexually traumatized individuals remembered their abuse, that there are no symptoms pathognomic of childhood sexual abuse, and that clinicians must use care to avoid errors of commission or omission.

There now appears to be general agreement among both researchers and clinicians that most sexual abuse victims remember their abuse, that some times abuse many be forgotten and later remembered, but that it is possible to construct convincing pseudomemories of abuse that never happened.


The nature of memory itself is central to evaluating a recovered memory claim. The popular view of memory is that it operates like a videotape in which everything that happens to us is recorded and stored in our brain, waiting for the correct playback button to be found so that [he memory can be retrieved. This view of memory is basic to the belief in recovered memories and is widely accepted by both laypersons and professionals (Loftus & Ketcham, 1991).

But this view of memory is completely mistaken. In reality what we remember is a combination of the original encoding of the event, intervening events that happen to us since the original event occurred, and our current beliefs and feelings Dawes, 1988; Goodman & Hahn 1987; Loftus & Ketcham, 1991; Loftus, Korf, & Schooler, 1989). Reconstructed memories can some-times include detailed and subjectively real pseudomemories of events that never happened (Loftus & Ketcham, 1991; Wakefield & Underwager, 1994b). Many things that happen are forgotten. People regularly forget significant life events even a year after they occurred. Studies indicate that even events such as job loss, injuries, robbery, assaults, hospitalizations, or accidents are sometimes forgotten (e.g. Loftus, 1993; Raphael, Cloitre, & Dohrenwend, 1991; Rettig, 1993). However, although forgotten, such events are generally readily recalled when there are cues.

But even with cues, people are seldom able to remember incidents from before the age of 3 or 4 because of the phenomenon of infantile amnesia. Infantile amnesia is part of the normal process of growth and development and has nothing to do with dissociation, repression, or traumatic amnesia. Although some researchers report slightly younger estimates of how far back adults can remember, these bits of memory may well represent educated guesses about what was likely to have happened (Loftus, Garry, Brown, & Rader, 1994). No researchers report reliable memories in adults (hr events that occurred before the age of 2 (see Malinoski, Lynn, & Sivec, Chapter 5, this volume).

Sexual abuse may be forgotten — not "repressed," not "dissociated," not remembered only by an "alter personality," but simply forgotten. No one disputes this. Not all sexual abuse is traumatic. Almost every study on the effects of sexual abuse reports a substantial group of subjects with little or no symptomology (Finkelhor, 1990). Some adults do not view their childhood abuse as traumatic and may not even define themselves as sexual abuse victims. They may have been too young at the time of the abuse to fully understand what was happening, especially if the abuse consisted only of fondling. For many children, the abuse may have been an unpleasant but relatively unimportant event in the same category as countless other unpleasant childhood events (Spence, 1993).

In addition, as Schacter (1996) observes, people may attempt to avoid or suppress painful experiences. Because sexual abuse is unlikely to be discussed if not disclosed, victims may be deprived of opportunities to talk about their abuse, which could in turn weaken their memories for those experiences. Schacter notes that this kind of explanation could apply to cases in which the initial experience was not highly traumatic but was disturbing or confusing enough to make the victim avoid thinking about it. The forgotten memories then may return years later through prompted recall. We have described an example of corroborated abuse that was forgotten until the victim's recall was triggered by reading a letter she had written years before at the time of the abuse (Wakefield & Underwager, 1994b).

It is therefore not surprising that some sexual abuse victims forget about it until they are reminded in some way. It is not necessary to hypothesize a mechanism of repression, dissociation, or traumatic amnesia for such cases.

It is also unlikely that persons for whom the abuse was highly traumatic will forget it. There is information on the reactions of people to documented trauma, such as fires, airplane crashes, terrorist attacks, automobile accidents, hurricanes, and being held hostage. Such trauma victims may report feelings of unreality, detachment, numbing, disorientation, depersonalization, and flashbacks, but a response of total amnesia for the entire event is not reported in the literature (e.g., Eth & Pynoos, 1985; Schacter, 1996; Spiegel, 1991; Wilson & Raphael, 1993). Although memories may be fragmented and impaired, complete amnesia for traumatic episodes is extremely rare.

Terr (1985,1988, 1990, 1991, 1994) studied many children who underwent verified trauma. Terr differentiates repression from simple forgetting and believes that sexual abuse can be repressed or dissociated , and later recovered. But, despite her beliefs about recovered memories, there are 110 reported cases in her research in which children over the age of infantile amnesia have no memories for the trauma.

Terr's findings that children subjected to documented trauma remember the trauma are consistent with other studies of traumatized children. In discussing the effects on children who witnessed acts of personal violence, including homicide, rape, or suicide, Pynoos and Eth (1985) state that such children "do not display traumatic amnesia" (p.24). Children who witnessed a parent being murdered have not been found to develop amnesia; instead, they are likely to be troubled by intrusive thoughts (Black, Kaplan, & Hendriks, 1993; Malmquist, 1986). Gordon and Wraith (1993) studied more than 100 families affected by disasters and traumas, including physical and sexual abuse, and there are no accounts of anyone developing amnesia. The types of trauma occurring in wartime will include some cases of repeated trauma, such as when the child is subjected to repeated bombardment and shelling. Macksoud, Dyregrov, and Raundalen (1993) discuss the effects of such wartime experiences on children and amnesia for the trauma is never mentioned.



Repression is the original concept used to explain how people can have no memories of sexual abuse but can later retrieve these memories in accurate detail (e.g. Blume, 1990; Briere & Conte, 1989, 1993; Fredrickson, 1992; Maltz, 1990; Terr, 1991, 1994, 1996; Williams, 1992). Although repression is usually differentiated from dissociation and traumatic amnesia, these concepts ire often used interchangeably.

Repression is seen as a psychological defense that results in the person losing all memory for events that were too traumatic to be borne by the conscious mind. This protective mechanism is thought to be powerful enough to completely block out repeated instances of sexual abuse, rape, torture, and murder. It is assumed that tile repressed abuse still affects us in powerful ways. The long list of "symptoms" of childhood sexual abuse described earlier are said to result from such repressed abuse.

Repression comes from Freudian psychodynamic theory and is seen as an active, filtering process that is different from ordinary forgetting. The concept of repression that involves the banishment from consciousness of a series of traumatic and intrusive events that take p lace in different circumstances over a number of years has been termed "robust repression" (Ofshe & Watters, 1993).

Repression is used in different ways by different researchers and theorists (Singer & Sincoff, 1990). If it is defined narrowly as intentional suppression of thoughts about an event, there is no dispute that it exists. But, if repression is defined as an unconscious defense mechanism used to block out all memories of overwhelming events, it is highly controversial and most experimental psychologists do not believe it exists. There are no empirical quantified data to support the theory of repression; the only evidence comes from impressionistic clinical case studies and anecdotal reports. Although hundreds of studies have tested the concept of repression (Hoch, 1982; Holmes, 1974, 1990, 1994; Hornstein, 1992; Mackinnon & Dukes, 1963), none has produced unequivocal support for the concept. In 1974, Holmes published a review in which he concluded that there was no reliable evidence for repression. In 1990, he stated that he has not seen anything new in the literature to change his conclusion. In 1994, he observed that "the many elegant theoretical explanations or repression may seem convincing, but they are meaningless because its existence has never been demonstrated" (p. 5).

In addition, traditional psychodynamic theorists object to the way repression has been used by recovered memory proponents. Traditional analytical therapists are concerned with the patient's perceptions of reality rather than the historical accuracy of the material uncovered in therapy. They do not assume that childhood memories retrieved in therapy are historically truthful (Hedges, 1994; Nash, 1992; Wakefield, 1992).

There is nothing in the literature on repression supporting the belief that repeated episodes of sexual abuse can be "repressed" and inaccessible to memory and to be only remembered years later in bits and pieces. There is no support for a concept of "robust repression" that underlies the use of recovered memory techniques.

Dissociation, Psychogenic Amnesia, and
Posttraumatic Stress Disorder

As the concept of repression was attacked, many people began abandoning it as tile operative mechanism and dissociation is now often used to describe the process by which traumatic memories are banished from consciousness. Dissociation refers to the failure of the person to integrate all relevant aspects of an experience so that it becomes difficult or impossible to recall the experience.

There is no disagreement about the concept of dissociation comparable to that surrounding repression. All of us are familiar with minor forms of dissociation, such as daydreaming, becoming lost in a book, or "spacing out" while driving. Dissociation can be seen as lying on a continuum from such ordinary forms to pathological forms such as amnesia, depersonalization, and fugue states (Bernstein & Putnam, 1986). Trauma victims often report having had dissociative experiences while in the midst of a rape, accident, terrorist attack, and so on. They may feel unreal or detached during the trauma and the experience may feel like a dream or like it happened to somebody else. Later they may show some memory impairment or perceptual distortions (Spiegel, 1991; Terr, 1991).

Persons undergoing repeated trauma are believed to learn to use dissociation as a defense against the trauma (e.g., Courtois 1992, 1996; Dolan, 1991; Fredrickson, 1992; Putnam, 1991a; Putnam & Trickett, 1993; Terr, 1991, 1994, 1996). Because psychiatric patients report disproportionately high abuse histories (Bernstein & Putnam, 1996), this is seen as supporting the trauma-dissociation hypothesis. The repeatedly abused child is thought to learn to dissociate as a way of defending against the trauma so that the trauma is lost from conscious awareness. Sometimes, after under-going repeated severe trauma, the child is said to develop alter personalities and multiple personality disorder; It is assumed that adults cannot recall their childhood sexual abuse because they dissociated the abuse and were in an altered state of consciousness. Therefore, hypnosis and age regression are used to retrieve these memories by attempting to replicate the state in which the abuse occurred.

There are difficulties with this use of dissociation as a way of supporting recovered memories assumptions. If up to one-quarter of the women in the United States have been abused but do not remember it because they dissociated it, we would expect the symptoms of childhood dissociation to be frequently seen and for dissociation to be a common diagnosis in the literature on psychopathology in children. But review articles (Lahey & Kazdin, 1988, 1989, 1990) on childhood disorders do not even mention dissociative disorders. In addition, the connection between childhood trauma and dissociation has been questioned (Tillman, Nash, & Lerner, 1994). People who have dissociative symptoms have more psychopathology in general and the a-c-search claiming a link between trauma and dissociation is plagued by conceptual and design problems.

There are changes between the revised third edition and the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-III-R and DSM-IV; American Psychiatric Association, 1987, 1994) in the dissociative amnesia disorder that reflect the attention given to recovered memory claims. A new type, systematized amnesia, is added. Whereas DSM-III-R states that psychogenic amnesia (now called dissociative amnesia in DSM-IV) "is rarely diagnosed under normal circumstances; it is more common in wartime and during natural disasters" (p. 274), according to DSM-IV the acute form that occurs during wartime and in natural disasters is less common. The main manifestation is now said to be a gap or series of gaps for aspects of the person's life history. The phrase "series of gaps" is added in DSM-IV. DSM-IV notes that there has been an increase in the reports of dissociative amnesia that involve previously forgotten early childhood traumas, but it acknowledges that this increase is subject to very different interpretations, with some professionals maintaining that the increase is due to improved identification but others due to overdiagnosis in suggestible individuals.

Although Loewenstein (1991) broadens the concept of psychogenic amnesia to include a group of events, which allows psychogenic amnesia to account for repeated instances of sexual abuse, no research supports this assumption. No empirical data support a concept of psychogenic amnesia for a category of events stretching across several years. Literature reviews on the consequences of sexual abuse (Beitchman, Zucker, Hood, daCosta, & Akman, 1991; Beitchman, Zucker, Hood, daCosta, Akman, & Cassavia, 1992; Cole & Putman, 1992) do not include psychogenic amnesia as a sequelae of sexual abuse.

In addition, neither Loewenstein nor anyone else has explained just how traumatic amnesia (or "repression" or "dissociation") is supposed to work in eradicating sexual abuse memories. Does the person completely dissociate the abuse and therefore develop traumatic amnesia immediately following each event? If this is the case, each new instance of abuse would be like the very first time as the child would have no memories of any of the previous incidents. Or, at some point after the abuse stops, does the person suddenly develop total amnesia for all memories of all the abuse incidents that were previously remembered? Proponents of memory-recovery techniques have not addressed these questions.

In clinical case studies of psychogenic amnesia, the person suffers a trauma that can be verified. But corroboration is seldom found in recovered memory claims. In contrast, people who have experienced documented trauma rarely develop amnesia unless they are physically injured. Although dissociation may occur and memories may be fragmented and incomplete, this is quite different from having no memory whatsoever of the incident. With the exception of dissociative identity disorder (discussed later), nothing in the literature on dissociation describes selective amnesia for a series of traumatic events that occur at different ages and at different times and places.

The diagnosis of posttraumatic stress disorder (PTSD) is often found in recovered memory claims where it may be used to explain the lack of memories. Survivors of severe sexual abuse are often reported to have PTSD symptoms, including disturbances of memory (Briere, 1996; Harvey & Herman, 1996). According to DSM-IV (American Psychiatric Association, 1994), the PTSD diagnosis is given when a person develops characteristic symptoms following exposure to an extremely traumatic event that involves actual or threatened death or serious injury, or a threat to the physical integrity of self or others and the person's response involves intense fear, helplessness, or horror; The symptoms involve reexperiencing the traumatic event, avoidance of stimuli associated with the event or numbing of general responsiveness, and increased arousal. But although the criteria for PTSD include numbing and efforts to avoid thoughts or feelings along with psychogenic amnesia for an important aspect of the event, there is no mention of total amnesia for the entire event.

Also, to diagnose PTSD, there must be a known stressful event (Fisher & Whiting, 1996). The diagnosis cannot be given on the basis of the symptoms alone without verification of the event. A task force of the American Psychiatric Association (Halleck, Hoge, Miller, Sadoff; & Halleck, 1992) makes it clear that it is inappropriate to use the presence of symptoms related to PTSD as evidence that prior abusive events such as rape and child sexual abuse took place. Nevertheless, Fisher and Whiting (1996) note that. PTSD symptoms are frequently misused as diagnostic criteria for the validation of sexual abuse.

Dissociative Identity Disorder

Dissociative identity disorder (DID; formerly multiple personality disorder, or MPD) often appears in recovered memory cases when the alleged abuse is violent and sadistic. DSM-IV (American Psychiatric Association, 1994) defines DID as the presence of two or more distinct identities or personality states. The disorder is believed to begin early in life and mainly to affect women.

It is assumed that most people with DID were abused as children (e.g., Dunn, 1992; Kluft, 1987, 1991; Putnam, Guroff, Silberman, Barban, & Post, 1986; Ross et al. 1990). A "protector" personality is believed to develop and take over for the child, who therefore escapes psychologically from the abuse (Spiegel, 1991). The theory is that the traumatized child learns to dissociate from the repeated abuse so that DID is found in people with a history of severe physical or sexual abuse.

But the main support for this belief comes from clinical case reports. In a review of the empirical literature on the long-term effects of child sexual abuse, Beitchman et al. (1991) conclude that there is insufficient evidence to confirm a relationship between childhood sexual abuse and multiple personality disorder;

In addition, even though it is found in DSM-III-R and DSM-IV, DID itself is controversial (Dunn, 1992; North, Ryall, Ricci, & Wetzel, 1993). This controversy has increased as this diagnosis was given to people claiming recovered memories of child sexual abuse. Many researchers and clinicians believe there is little empirical evidence supporting DID as a distinct mental disorder and that it is heavily dependent on cultural influences for both its emergence and its diagnosis (Aldridge-Morris, 1989; Fahy, 1988; Frankel, 1993; McHugh, 1995; Merskey, 1992, 1995; Underwager & Wakefield, 1996; Weissberg, 1993). Spanos (1994, 1996) notes that few patients show clear signs of MPD at the beginning of therapy but learn to act out the role of the multiple personality patient during therapy. He and his colleagues have performed a series of role-playing and hypnosis experiments that support the interpretation or DID as a learned role.

The interpretation of DID as an artifact of psychotherapy is criticized by researchers and clinicians who see DID as a distinct mental disorder and who maintain that although some of the symptoms of DID can be created by therapists, there is no evidence that the disorder itself can be created. Spiegel and Cardeña (1991) state, "Although it is possible that the inappropriate handling by a therapist of a highly suggestible person may give rise to inaccurate reports of early abuse and MPD-like symptomatology, this mechanism does not seem sufficient to explain all or even most of the cases of MPD" (pp.371-372). Gleaves (1996) maintains that Spanos (1994) makes "numerous false assumptions about the psychopathology, assessment, and treatment of DID" (Cleaves, 1996, p.42). Lie believes that treatment recommendations arising from such assumptions may be harmful to patients, since posttraumatic symptoms will not be addressed.

But, although the existence of DID as a distinct mental disorder is debated, there is agreement that suggestible patients, with unwitting encouragement from their therapists, can learn to show symptoms of DID (e.g., North et al., 1993).

Body Memories

The "body memory" concept assumes that although there are no conscious memories, the body remembers and the client has physical symptoms that correspond to the childhood abuse. For example, Fredrickson (1992) describes a man whose anus hurt when he talked about his father because his body remembered the father's anal rapes. Grand et al. (1991) state that the analyst can observe the body memories and therefore be certain that abuse happened, even though the patient is confused.

Body memories are believed to occur when memories are repressed because the body has no intellectual defenses and therefore cannot screen out memory imprints. Later, a "survivor," who may have no abuse memories, will retrieve colors, hear sounds, or experience smells, odors, and taste sensations and her body may react in pain reflecting the purported abuse as the memory is retrieved. The body memory can even be an actual physical representation of an event, such as hand prints appearing around the neck of a person who was allegedly choked during the abuse. There is no empirical support for the body memory concept. (See Smith, 1993, for a critical analysis of the body memory concept.)

Trauma Memory

An assumption underlying recovered memory theories is that trauma memory is different from other memories. Herman (1992) says that whereas normal memory is "the action of telling a story," traumatic memory is "wordless and static" (p. 175), and Harvey (1993) maintains that traumatic memories are acquired in an altered state and are not stored in the same way as ordinary memories. Freyd (1993) postulates a "betrayal-trauma theory" in which childhood incest produces conflict between external reality and a necessary system of social dependence. When the abuser (betrayer) is someone we depend on, such as a parent, we need to ignore the betrayal so we block out information about sexual abuse in order to preserve our attachments. There is no convincing empirical support for these trauma-memory theories.

van der Kolk (1988a, 1988b; van der Kolk & van der Hart, 1989) developed a model of the biological changes associated with trauma along with the effect of this on later memory. He postulates a complicated biological theory to explain how traumatic events are handled differently from nontraumatic events. But van der Kolk's biological trauma memory concepts do not explain how repeated, intrusive abuse, occurring in different circumstances and places and over a number of years, can completely disappear from memory and then be recovered many years later. His explanations of how biological changes can explain PTSD symptomatology and contribute to later psychopathology do not need the assumption of' some type of repression process during the trauma.


Lawsuits against Parents

Survivor's groups and books may suggest suing the alleged perpetrator (Bass & Davis, 1988; Crnich & Crnich, 1992; Nohlgren, 1991). Some therapists and attorneys believe such lawsuits are an essential part of the healing process. One attorney, quoted in The Courage to Heal, says, "In my experience, nearly every client who has undertaken this kind of suit has experienced growth, therapeutic strengthening, and an increased sense of personal power and self-esteem as a result of the litigation" (Bass & Davis, 1988, p.310). Mallia (1993) believes civil litigation can bring closure to the victim and claims that winning the lawsuit "can cleanse the victim's tortured psyche" (p.129). Lamm (1991) lists several benefits, including exoneration of the victim through publicly blaming and punishing her abuser, deterring future abuse, financial rewards and healing for the victim, and increasing public awareness of the problem of abuse. Clute (1993) describes the effect on a survivor she represented who won $3 million: "It was clearly a victory of the human soul, a liberation from a bondage that had changed him for a long time.... I knew I had participated in a significant healing experience" (p.122). Walker (1992) describes such lawsuits as a "new and exciting tool in the war against child sex offenders" (p. 125).

But others (e.g., Ewing, 1992; Thompson, 1993) are pessimistic about the psychological benefits of civil litigation and observe that, especially if the abuse never haippened, such a lawsuit can be harmful not only to the falsely accused but to the person initiating the lawsuit. Nevertheless, there is evidence that such lawsuits are not uncommon. The FMSF reports that 1 out of 16 of the accused parents who have contacted them have had lawsuits filed against them.

Statutes of Limitations

The statute of limitations restricts the period during which lawsuits may be filed after the date of injury. This protects defendants from having to defend themselves in court years later when witnesses have died, evidence is lost, documents have disappeared, and memories have faded. When the plaintiff is a minor, the statute of limitations does not begin to run until the age of majority.

The law, however, recognizes circumstances in which the person could hot reasonably have discovered the injury within the statutory time period. Common examples are in medical malpractice (i.e., a former surgery patient exlperiences pain and discovers that a sponge was left in his abdomen years before) and asbestos damage. To account for such circumstances, courts have developed a "delayed discovery rule." This rule delays tolling the statute of limitations until the plaintiff discovers or reasonably should have discovered the injury. Many courts have applied the delayed discovery rule to lawsuits brought by plaintiffs claiming childhood sexual abuse. Some courts have limited the delayed discovery rule to cases in which the plaintiff claims to have completely repressed her memory of the abuse, only to recall it as an adult. In others, the plaintiff only has to claim that she did not discover within the statutory period that her current psychological problems were caused by the abuse (Lamm, 1991).

Several states have passed legislation extending the statutory period in civil cases so that the statute of limitations does not begin until 2 or 3 years after the alleged abuse is remembered and/or after the claimant understands that the abuse caused injury (Bulkley & Horwitz, 1994; Colaneri & Johnson, 1992; Loftus, 1993; Loftus & Rosenwald, 1993; Slovenko, 1993). Courts and legislatures in many states have created legal mechanisms for both criminal and civil actions based on recovered memories (Loftus & Rosenwald, 1993; Sargeant, 1994).

Recovered Memory Claims and Daubert

The concepts and theories used by proponents of recovered memories are vulnerable under Daubert. For example, the concept of repression is not falsifiable because all the behaviors of the client can be accommodated by this concept. Innumerable symptoms are said to be caused by repressed sexual abuse, even when there are no memories. Parents who are outwardly normal as well as those who are clearly dysfunctional are seen as capable of abusing their children in a way that produces repression. When there is no corroboration from the accused, the accused is said to be either denying or repressing the act.5 If the "survivor" later retracts her recovered memories, she is said to be "rerepressing" the abuse or retracting it under pressure from her family. If the alleged abuse is implausible (i.e., bizarre satanic ritual abuse), it is claimed that these things really happen and that the satanists are clever conspirators with positions in high places who therefore can cover it tip. If the alleged abuse is completely impossible (i.e., a woman who is found to have never been pregnant alleges multiple forced births as a breeder in a cult), the memories are said to be distorted but based on actual abuse or to be symbolic of emotional abuse. If everything can be explained by the theory and it is not capable of being tested or falsified, the theory should not be allowed under Daubert.

As mentioned previously, the concept of repression is controversial atid not supported by empirical research. There are a few studies examining amnesia for childhood abuse (Briere & Conte, 1993; Feldman-Summers & Pope, 1994; Herman & Schatzow, 1987; Loftus, Polonsky, & Fullilove, 1994; Williams, 1994), but none assess repression or provide credible scientific evidence to support the assumptions underlying recovered memory therapy. They all have methodological problems concerning the manner in which trauma and amnesia are documented. None support the view that large numbers of people are so amnesiac for actual abuse that intrusive therapy techniques must be used to help them remember and heal. (for a critique of these studies see Kihlstrom, 1996; Lindsay & Read, 1994; Pope & Hudson, 1995; Wakefield & Underwager, 1994b.)

Faced with both the lack of support and the lack of testability for repression, the court should rule that testimony based on the concept is not scientific and cannot be relevant or helpful to the finder of fact. Therefore, it is not admissible. This could make it impossible for civil litigation based on a claim of recovered repressed memories to be pursued. The other concepts should be treated similarly.

Lawsuits against Therapists

An increasing number of people who recovered "memories" of abuse in therapy are now retracting these memories (Freyd, 1994). Some of them are suing their former therapists for malpractice. A suit by a Texas woman, Laura Pasley, against the therapists who persuaded her that she had been sexually abused by her mother, brother, grandmother, and a neighbor ended in a significant out-of-court settlement and a Minnesota psychiatrist, Diane Humenansky, is currently being sued by five women who claim that she implanted false memories of satanic ritual abuse (Gross, 1994). Two of the cases went to trial and the juries awarded multi-million- dollar settlements. Dr. Humenansky's license has also been suspended (Gustafson, 1994).

Many therapists are becoming worried about ethical complaints and malpractice suits. At a conference entitled "Memories of Abuse" in Minneapolis in June 1993, several therapists expressed anxiety over this. But the worried therapists do not appear to be seriously considering the possibility that their techniques may be creating false memories and harming their clients. Peterson (1994) describes her "inescapable" pain and "shattered soul" after her patient "sadistically turned on [her]" by filing a malpractice suit, but she still believes her client was abused. "Kill the messenger. Lie. This client relived the trauma by victimizing me" (pp. 26-27).

Accused parents often want to sue the therapist they hold responsible for their adult children's false memories of abuse. However, they lack legal standing to sue for malpractice because, as they were not patients, the therapist owes them no duty. For a cause of action for negligence against a therapist to succeed, the plaintiff must establish (1) the duty of the professional to use such skill, prudence and diligence as other members of his profession commonly possess and exercise; (2) a breach of that duty; (3) a proximate causal connection between the negligent conduct and the resulting injury; and (4) actual loss or damage resulting from the professional's negligence. The difficulty in third-party professional negligence cases is in establishing the duty of the therapist to the parent. If the law fails to recognize a duty between the therapist and the parent, the parent lacks standing to sue, no matter how egregious the behavior of the therapist (Simons, 1994).

But in Ramona v. Isabella (1994), the judge ruled that as a parent and as someone who had been substantially affected by the therapist's alleged malpractice, the accused father, Gary Ramona, did have standing to sue his daughter's therapists. In May 1994 the jury awarded Ramona $500,000 in damages.

In Ramona, the daughter, who sought therapy for bulimia nervosa, originally had no memories of sexual abuse. However, the therapist told the girl's mother that 70% to 80% of bulimics had been sexually abused and the daughter was placed in group therapy where abuse was discussed. Eventually she began having vague flashbacks of abuse by her father; The therapist then referred her to a psychiatrist for a sodium amytal interview. During this interview, she accused her father of repeated rape from the time she was age 5 to age 7. Afterwards, the therapist and the psychiatrist assured her that it was impossible to lie under the influence of sodium amytal and that her memories were real. They arranged a confrontation session with the father (Ewing, 1994; Loftus & Rosenwald, 1993).

As a result of the accusations, Ramona was fired from his $500,000-a-year job and his wife divorced him. In granting Gary Ramona standing to sue his daughter's therapists, the trial judge referred to a California Supreme Court ruling (Molien v. Kaiser Foundation Hospitals, 1980). In this case, the court ruled that the husband of a woman misdiagnosed as having syphilis had standing to sue his wife's physicians because, as he had to be treated for syphilis also, they had a duty to care for him as well as the wife (Gross, 1994). In similar cases of lawsuits by parents or relatives, settlements have been reached in several, including one against an attorney for wrongful use of civil proceedings (Bean v. Peterson, Peterson, 1995; Downing v. McDonough, 1997).

Ramona is seen as a landmark case, both because a third party was granted standing to sue and because it was the first courtroom challenge to the recovered memory assumptions and techniques. It has been interpreted by skeptics as a judgment against recovered memory therapy and a warning to psychotherapists who engage in it. It is expected to result in more malpractice suits. It is also expected to change the behavior of many therapists.


The only way to test the accuracy of a memory is through independent corroboration (Schacter, 1996; Spiegel & Scheflin, 1994). Mental health experts have no special abilities, to tell whether a memory is true or is the product of imagination, social influence, or fabrication. Without external corroboration, there is no truly satisfactory way to reliably determine the truth of a given "memory." At the same time, we believe psychologists can provide helpful information to the finder of fact. This information includes facts about the nature of memory, the scientific credibility and status of the assumptions underlying memory-recovery techniques, the reliability and validity of these techniques, the nature of social influence and suggestibility, and the probabilities of the behaviors alleged. This information is not generally known by either laypeople or many professionals.

When a mental health professional becomes involved in a civil lawsuit involving recovered memories, it is necessary to get as much information as possible about the circumstances surrounding the disclosure and accusations. Here is a paraphrasing of the information Daly and Pacifico (1991) suggest should be gathered:

1. All medical, psychiatric, and school records of the person claiming abuse from childhood to the present.

2. Any information concerning relationships with peers, siblings and parents, or any childhood behavior problems of the person claiming abuse.

3. Any information concerning the sexual history of the person claiming abuse, including rapes, other childhood sexual abuse, abortions, and so on.

4. The nature and origin of the disclosure, in as much detail and specificity as possible.

5. Information about any current problems or stresses in the life of the person claiming abuse.

6. The nature of any current therapy, e.g. whether techniques such as hypnosis and survivors' groups were used, the training and background of the therapist, and whether lie or she specializes in treating MPD or "recovered" abuse.

7. Any books, television shows, or workshops about sexual abuse or rape to which the person claiming abuse may have been exposed.

8. Any exposure to recovered memory cases though a highly publicized case in the media or through frjends who may have reported that this happened to them.

9. The work history of the person claiming abuse, including any problems with supervisors or coworkers, especially any allegations of sexual harassment.

10. The psychological characteristics and social and family history of the accused adult(s), including any drug or alcohol use, sexual history, family relationships, and job history.

11. Any criminal record or prior behaviors in the accused adult which would support or undermine the credibility of the allegations.

12. A detailed description of the behaviors alleged to have occurred.

13.Possiblle ways by which the person making the accusation might benefit from or receive reinforcement from making the accusation (e.g., a civil lawsuit, an explanation for why life has not gone well, the expression of anger for perceived childhood injustices, power over a dominant parent, attention, acceptance, new friends [in survivor group], etc.).

After obtaining the available documents and information, it is helpful to create a chronology of events. In the chronology, it is particularly important to note information about how the memory returned and how it was disclosed; the impact of therapy, books, or survivors' groups, and so on; any changes in the nature of the allegations over time; and any possible secondary reinforcement.

Some professionals have proposed ways of evaluating claims of alleged sexual abuse based on recently recovered memories (Gardner, 1992a, 1992b; Rogers, 1992, 1994; Wakefield & Underwager, 1992). There is little empirical research on this, so all such suggestions are based primarily on existing knowledge about memory, social influence, suggestibility, conformity, the psychotherapy process, hypnosis, and the characteristics and behavior of actual sexual abusers.

The Nature of the Alleged Behaviors

When the allegations are of extremely deviant, low-probability behaviors rather than of behaviors more typical of actual abusers, the memory is less likely to be for a real event. This appears to be self-evident, but cases of alleged sexual abuse often involve highly implausible allegations. For example, preliminary data from a survey by the the American Bar Association indicated that approximately one-fourth of local prosecutors have handled cases involving "ritualistic or satanic abuse" (Victor, 1993).

Recovered memory claims may include allegations of highly deviant abuse, such as anal or vaginal rape of toddlers, violence, sadism, sex with animals, feces and urine, abuse by several adults, and satanic ritual abuse. The accused are not just fathers; mothers and other relatives are also accused, often of acting together; These allegations can be assessed in terms of what is known about the behavior of actual child sexual abusers (Wakefield & Underwager, 1994a, 1994b).

Information is available about the behavior of actual sexual abusers, although the studies vary somewhat as to the behaviors reported. This is not surprising because the studies differ in terms of the sample studied (community, college, clinical, prison, hospital, etc.), whether victims or offenders are sampled, the method of obtaining the data (interviews, questionnaires, hospital records, etc.), the sex of the victim, the definition of terms, the specificity of the description of the behavior, whether the abuse is intrafamilial or extrafamilial, and the adequacy of the verification of the abusive acts. Despite these considerations, the research literature provides helpful information about what actual child molesters do when they sexually abuse a child (Erickson, Walbek, & Seely, 1988; Gebhard, Gagnon, Pomeroy, & Christenson, 1965; Kendall-Tackett & Simon, 1987; Kinsey, Pomeroy, Martin, & Gebhard, 1953; Tollison & Adams, 1979).

Most sexual abuse victims are girls (Erickson et al., 1988; Tollison & Adams, 1979). Although some preschoolers are abused, most are older. Tollison and Adams (1979) report that average age of female victims is ages 6 to 12 and male victims appear to be somewhat older; Erickson et al. found that one-fourth of the victims of both sexes were under age 6, one-fourth were between 6 and 10, and half were between the ages of 11 and 13.

Most sexual behavior consists of fondling, exhibitionism, masturbation, and oral or genital contact. Anal and vaginal penetration of very young children is rare but penetration becomes more likely with an older child. Males are more likely to be victims of attempted or actual anal penetration than alt females (Erickson et al., 1988; Gebhard et al., 1965; Kinsey et al., 1953; Tollison & Adams, 1979). Vaginal penetration is more common in clinical samples compared to community samples (Kendall-Tackett & Simon, 1987). Most victims know the offender; abuse by strangers is much less common. It is rare for the offender to have a partner who participates in the abuse or to molest children in groups.

Many professionals believe that a grooming process is generally involved in incest (e.g., Christiansen & Blake, 1990; Erickson et al., 1988; Gebhai4 CL al., 1965; Kendall-Tackett & Simon, 1987, 1992). Aggression and violence are not usually part of [he behavior, although Lang and Langevin (1991) indicate that at least one in five child victims appear to be subjected to force or "gratuitous physical violence" as part of the abusive act.

Although sadistic, bizarre, or homicidal forms of child sexual abuse do occur (e.g., Dietz, Hazelwood, & Warren, 1990), they are extremely rare. Dietz et al. (1990) observe that such cases occur so infrequently in any one jurisdiction that it is difficult for researchers to gather information about them. Kinsey et al. (1953) state that in only one case was appreciable physical injury done to the child. Gebhard et al. (1965) report that only 6.6% of their sample used force. Erickson et al. (1988) note that there were no reports in their sample of sadistic or bizarre abuse. There is no evidence supporting Allegations of organized satanic ritual abuse conspiracies (Hicks, 1991; Lanning, 1992; Putnam, 199 lb; Richardson, Best, & Bromley, 1991; Victor, 1993).

Contrasted with this are the results from 398 surveys from the questionnaire project from the FMSF (Freyd et al., 1993; Wakefield & Underwager, 1994a, 1994b). The respondents were people who contacted the FMSF after their adult child recovered a memory of sexual abuse that the caller denied. No effort was made to make an independent determination of the veracity of the denial or of the information obtained. Many of the purported behaviors reported by the respondents are impossible and others have such low base rates that the probability of their actually occurring is extremely small. Examples of such allegations include the following:

· Abused by parents and grandparents in satanic rituals while wearing hooded black robes. Forced to drink urine and blood. Raped by grandfather while grandmother and mother watched. Hung by her heels. Abused with a hot poker, freezer, and washer wringer;

· Abused by mother, father, strangers, blind uncle, and nursery school teacher; Raped with scissors, killed babies, worshiped Satan, and ate ears and other organs. Was sold into child prostitution.

· Forced to have sex with a neighbor's dog and subsequently had a baby that was half dog.

In the FMSF survey, although fathers were usually the ones accused, mothers were often accused along with the fathers and in one-third of the cases, a variety of other persons were accused, most often along with the parents. Over half of the respondents appeared to have little idea concerning just what it was they were supposed to have done, but when this information was known, the allegations included a very high proportion of extremely deviant and intrusive behaviors. In the 203 cases in which the respondents knew the nature of the allegations, violence was alleged in 41%, rape in 44%, witnesses to the abuse in 42%, and satanic ritual abuse in 34%.

Therefore, in evaluating a case of alleged childhood abuse, when the allegations are of extremely deviant, implausible, and low-probability behaviors, the memory is unlikely to be for a real event. Allegations of ritual abuse by intergenerational satanic cults are especially unlikely to be true.

Because there is no scientific evidence to support a mechanism of "robust repression," if there are allegations of a series of violent and abusive incidents across time in different places and situations, the recovered memory is less likely to be true than if it is for a single traumatic incident. When the disclosures progress across time to ever more intrusive, abusive, and highly improbable behaviors, the growth and embellishment of the story may represent the suggestions and reinforcement in therapy. Anecdotal evidence (Wakefield & Underwager, 1994b) suggests this can happen in survivors' groups where group members spend the sessions talking about their newly recovered memories and flashbacks.

The Age the Abuse Is Said to Have Occurred

If the memory is for abuse that occurred at a very young age, such as abuse during infancy or under age 3 or 4, the memory is less likely to reflect a real event. This is younger than documented sexual abuse victims whose average age is between 6 and 12. In addition, the phenomenon of childhood or infantile amnesia makes it less likely that the memory is of a real event because adults and older children seldom remember incidents from their lives that happen prior to age 3 or 4.

But in the FMSF survey, the abuse typically was said to have begun at a very young age. For 29% of the cases, the alleged events began at tinder age 2. For 55%, the alleged events began from ages 2 to 6. Only in 16% did the alleged events first happen at age 6 or older; The median age for the age the accusing child claims the abuse began is between ages 3 and 4 (Freyd et al., 1993; Wakefield & Underwager, 1994a, 19941)).

Gardner (1992a, 1992b) believes that an important factor in evaluating the truthfulness of recently remembered abuse is the length of time the alleged abuse took place. The longer the period of abuse, the less the likelihood of its being forgotten. He sees it as more credible that a person would forget abuse that occurred over a 2- or 3-year period at age 6 or 7 compared to abuse that took place from ages 2 to 18. The age a which the abuse is said to have stopped is another factor. Although events taking place at age 5 or 6 may be forgotten, it is much less likely that events taking place during the teen years would not be remembered.

Characteristics of the Accused

Especially when the behaviors alleged are highly deviant, the allegations 'lit less likely to be true when they include the mother and when a psychological evaluation of the defendant indicates no discernible pathology. There is no single child sex offender personality type. But despite the fact that sex offenders are heterogeneous in personality characteristics, they tend to have psychological problems (Ballard et al., 1990; Kalichman, Shealy, & Craig, 1990; Langevin, 1983; Overholser & Beck, 1986; Wakefield & Underwager, 1988; Weinrott, & Saylor, 1991).

Psychological evaluations are less helpful when the behaviors alleged are of nonviolent fondling and other less deviant behaviors. It is not unusual in such cases for the perpetrator to show a "normal" personality based on psychological testing. Therefore, the Minnesota Multiphasic Personality Inventory (MMPI) or other assessment techniques do not rule out the possibility that the person is a sexual abuser because some sex offenders produce normal MMPIs.

Although no psychological test or evaluation procedure can determine whether or not a given individual has, in fact, sexually abused a child or committed any other specific behavior (Erickson, Luxenberg, Walbek, & Seely, 1987; Myers, 1992; Nagayama Hall & Crowther, 1991), a psychological evaluation of the person accused can provide information concerning the likelihood that an individual would engage in the behaviors alleged. Normal, functional persons do not ordinarily act in bizarre, unusual, and totally idiosyncratic ways. Therefore, it may be helpful to evaluate the accused, especially when the allegations are of more intrusive, deviant, or sadistic behaviors. The person must be evaluated in light of the specific behaviors he is accused of committing. When it cannot be demonstrated that an accused person has the level of pathology expected given the behaviors alleged, the likelihood of a false accusation increases.

Accusations that a woman has sexually abused a child must be treated very cautiously. Although awareness about women perpetrators of sexual abuse has increased in recent years, sexual contact between children and women represent a minority of child-adult sexual contacts. Many studies depict women who sexually abuse children as being loners, socially isolated, alienated, likely to have had abusive childhoods, and apt to have emotional problems, although most are not psychotic (see Wakefield & Underwager, 1991, for a review). It is unlikely that a psychologically healthy and well-adjusted woman would sexually abuse a child.

Research claiming large numbers of outwardly normal female perpetrators must be examined very cautiously because false allegations may have contaminated the samples. For example, a large study using. a sample of day-care facilities where there were abuse allegations (Finkelhor, Williams, & Burns, 1988; Finkelhor, Williams, Burns, & Kalinowski, 1988) reported that 40% of the perpetrators were intelligent, educated, highly regarded women who had no histories of known deviant behavior; But their sample includes many cases that ended in dismissals, acquittals, or convictions or that were overturned on appeal (e.g., Kelly Michaels and McMartin preschool). In fact, the authors admit that charges were filed and arrests were only made in half of their cases and less than a tenth resulted in guilty pleas or convictions. In another example, Waterman, Kelly, Oliveri, and McCord (1993) describe research on the effects of ritualistic sexual abuse, but their sample was mostly the McMartin preschool children.

The Type of Therapy and Therapist Characteristics

When memories first emerge following reading The Courage to Heal or therapy with a practitioner who uses memory-recovery techniques, the possibility of influence must be considered. Techniques such as hypnosis, survivors' group participation, dream analysis, ideomotor questioning, and guided imagery may increase the risk of creating memories for events that never happened, especially when the therapist believes abuse occurred. Therefore, the medical notes and records should be carefully examined. Rogers (1992, 1994) notes that valid claims may arise in therapy, but in these cases the therapist did not use intrusive techniques and the individual was not placed in group treatment until the abuse had already been frilly detailed and documented.

Although some professionals believe that a lack of credentials and inadequate scientific training are major factors in false recovered memories (Dawes, 1992; Gardner, 1992a, 1992b; Gravitz, 1994; Rogers, 1992, 1994), therapists who have erroneous ideas of concepts such as the nature of memory, suggestibility and social influence, and the veracity of memories uncovered through techniques such hypnosis include those with advanced degrees. In the FMSF survey (Freyd et al., 1993), the therapists reported by the parents included all disciplines, including psychiatrists and social workers as well as less credentialed therapists. Bottoms, Shaver, and Goodman (1991) surveyed doctoral-level clinical psychologists and report that of those who encountered cases of ritual or religious abuse, the great majority, (,)3%, accepted the clients' claims as true. Many of these cases involved adult survivors who reported amnesiac periods. Yapko (1994a, 1 994b) found that 22% of therapists who believed in age regression and past lives held PhDs and 28.1% held MDs. All of the Poole et al. (1995) sample mentioned earlier were doctoral-level therapists with good credentials; one quarter of these reported beliefs and practices consistent with recovered memory therapy.

Therefore, the fact that the therapist has a PhD or MD from a major university does not mean that the therapist has avoided using misguided and suggestive therapy techniques. However, the lack of training and under-standing in relevant scientific issues may make some therapists more vulnerable to engaging in poor clinical practices.

Psychiatric History and Characteristics of the Plaintiff

Although psychopathology in some individuals may well make them more susceptible to developing pseudomemories, the data from the FMSF survey (Freyd et al., 1993) suggest that many people who recover memories of childhood abuse are not psychologically disturbed. In fact, the questionnaire responses suggest that most of the accusing adult children are well educated and occupationally successful and come from functional, intact, and successful families. Most of the accusers had no history of significant psychiatric problems prior to the recovered memories. These data are consistent with the work of Spanos, Cross, Dickson, and Dubreuil (1993), who found that subjects reporting UFO experiences were not psychologically disturbed.

Claims that the individual must have been abused because of problems in her life must be viewed cautiously. The existence of eating disorders, sexual dysfunction, anxiety, depression, or low self-esteem cannot be used to support abuse claims because these can be caused by a variety of factors. Beitchman et al. (1992) concluded there is insufficient evidence to confirm a relationship between childhood sexual abuse and borderline or multiple personality disorder; Pope and Hudson (1992) reviewed studies on bulimia and sexual abuse and i'eport that these studies did not find that bulimic patients show a higher prevalence of childhood sexual abuse than do control groups.

Rogers (1992, 1994) notes that in evaluating a case for civil litigation, an important consideration is whether the claimant is a bona fide patient or is in treatment for reasons other than pain or dysfunction.

Corroborating Evidence

Corroborating evidence, such as a childhood diary with unambiguous entries, pornographic photographs, or an uncoerced admission by the perpetrator, clearly makes the allegations much more likely to be true. Some cases may have this type of' corroboration.

Ambiguous evidence, however, such as a childhood story or drawings now reinterpreted in light of the believed-in abuse, cannot be used as proof that the abuse actually occurred. The nature and quality of the corroboration must be looked at carefully. For example, Herman and Schatzow (1987) claim that 3 out of 4 of 53 women in their group were able to validate their memories by obtaining corroborating evidence from other sources (p. 1), but most of the women in their group hid either full or partial recall of the abuse prior to therapy; only a minority of their sample addresses the issue of repressed memories. Ii) discussing the claimed corroboration, no distinction is made between women who always remembered the abuse and those who (lid not recall it until entering therapy. In addition, the details of the corroboration are vague and depended on the reports of the women in group therapy. Herman and Schatzow (1987) presented two case examples to describe the validation process in women who had no memories prior to therapy. In one, there was no corroboration. For the other, the corroboration consisted of the women's report in group therapy of discovering her brother's pornography collection and diary after he was killed in Vietnam. But there is no indication that anyone else saw the diary or verified what the woman said she found. Herman (Joseph DeRivera, personal communication, 1993) acknowledges that the diary was not seen by the group leaders.

Daubert Analysis

An examination of the scientific status of the competing claims of recovered memories and the assertion that therapists can teach vulnerable patients to have false memories suggest that the claim that abuse can be repressed and later recovered is not supported. In addition, the concept of repression as it is generally used is so broad that nothing can occur that cannot be accommodated by the concept and therefore it is not falsifiable.

The level of error in recovered memory claims is dealt with by Lindsay and Read (1994), who offer a Bayesian analysis of the error involved in such claims. Although they view the implantation of false memories by therapists as the best explanation for the claims, to assess the level of error, they accept the hypothetical possibility that the phenomenon may occur as the proponents describe it. Using the most liberal figures offered by the proponents of recovered memories for base rates of occurrence and assuming an unrealistic accuracy for the test of 90%, the level of error is five false positives (or every false negative. One-third of the diagnoses of repressed memory will be wrong. If the accuracy of test is dropped to 80%, still unrealistically high, half of the diagnoses of repressed memories will be wrong.

There is no general acceptance of the claims of recovered repressed memory in the science of psychology. There may be publications in peer-reviewed journals, but they are subject to criticism as primarily anecdotal and case study material and demonstrating severe weaknesses and flaws in design and procedures. Using the criteria of the Daubert decision, expert testimony advancing the concept of recovered memories as accurate accounts of prior historic events should not be allowed in the courtroom.

The position that people can develop subjectively real memories for events that did not happen is supported by laboratory research (Ceci & Bruck, 1995; Hyman, Husband, & Billings, 1995; Loftus, 1993; Loftus & Ketcham, 1994) as well as anecdotal and case study evidence (Gavigan, 1992; Goldstein & Farmer, 1993; GondoIf; 1992; Nelson & Simpson, 1994; Wright, I 993). It must be noted, however, that the extent to which the laboratory research can be generalized to psychotherapy is debated (e.g., Alpert et al., 1996; Oho, 1994; Pezdek & Banks, 1996).

When there is no corroboration, denial by the alleged perpetrator with support of the denial by others, and claims of improbable, bizarre, or even impossible acts, the level of error is likely to be very low. This points to a much lower level of error than the proponents of uncorroborated recovered memory can suggest. The Daubert circumstantial evidence of general acceptance and publication are also met. Therefore, scientific testimony refuting claims that the human mind can produce accurate memories with the help of memory-recovery techniques should be permitted in the courtroom.


Some survivors of childhood sexual abuse forget about their childhood sexual abuse, particularly if the abuse was not highly traumatic. This lack of memory is most likely due to some combination of the normal forgetting process, conscious suppression, and lack of rehearsal if the abuse was never talked about. But there is no credible evidence that people subjected to repeated incidents of violent and traumatic abuse after early childhood will lose all memories for the abuse. The research on the nature of memory and forgetting does not support the assumption that some abuse is so traumatic that all memories for it will be removed from consciousness by an active filtering process of repression or dissociation.

Sexual abuse may not be talked about until the victim is an adult and real abuse may be forgotten until the person is reminded in some way. In such cases, there is no need for concepts of repression, dissociation, or traumatic amnesia. Intrusive and traumatic abuse that occurs after early childhood is likely to be remembered. The research on victims of documented trauma supports t his assumption.

Schacter (1996) notes that because of ethical considerations, there can be no conclusive scientific evidence from controlled laboratory studies demonstrating that false sexual abuse memories can be created. However, there is good scientific evidence that pseudomemories for other events can be created in some people. The difficulty, as Spiegel and Scheflin (1994) note, is that it is just as possible to dissociate and retrieve a real memory as it is to convince oneself of a false belief. But there is no easy way to know the difference. Without independent corroboration, memory cannot be trusted. Although there is dispute over the prevalence, there is also no doubt that some therapists have helped create false memories of sexual abuse. The techniques used for memory recovery risk creating memories for events that never happened.

Many recovered memory cases are appearing in court. In such cases, although there is no substitute for external corroboration, mental health professionals can provide information to help the finder of fact sort out the truth and falsity of. an allegation. Although each case must be evaluated on its own merits, we are skeptical of the truthfulness of allegations involving satanic ritual abuse, abuse said to be remembered from infancy, claims that repeated, traumatic abuse has been completely repressed only to be remembered years liter, and abuse that is not recalled until after intrusive therapy techniques.


I. Under the Frye test a scientific technique is not admissible unless the technique is "generally accepted" in the scientific community. Giannelli (1980) notes that the Frye rule envisions a process by which a novel technique must pass through an "experimental" stage where it is scrutinized by the scientific community. Only after it has passed successfully through this process and has entered into the "demonstrable" stage can it be admissible. Under the Frye rule it is not enough that a qualified expert or experts believe the technique is valid and reliable; it must have been generally accepted by the relevant scientific community.  [Back]

2. The False Memory Syndrome Foundation, a tax-exempt research and education institute, is located at 3401 Market Street, Suite 130, Philadelphia PA 19104.  [Back]

3. This ruling was later reconsidered and overturned by the full Eighth Circuit. Nevertheless, it remains a good example of a Daubert analysis.  [Back]

4. The three clinicians were Judith L. Apert, Laura S. Brown, and Christine A. Courtois. The three researchers who prepared a separate statement were Peter A. Ornstein, Stephen J. Ceci, and Elizabeth F. Loftus.  [Back]

5. In some cases, the accused may be hypnotized or helped in other ways to "remember" his abusive behaviors. Ofshe (1992) provides an account of the Paul Ingram case where such efforts resulted in Ingram falsely confessing to abusing his daughters in a satanic cult.  [Back]


IPT Home Page Up One Level

Copyright © 1989-2014 by the Institute for Psychological Therapies.
This website last revised on April 15, 2014.
Found a non-working link?  Please notify the Webmaster.