Children, young people and families, Legal, criminological and forensic, Memory

False memories of childhood abuse

Are therapists to blame? Chris R. Brewin and Bernice Andrews consider the evidence in a controversial area.

07 June 2017

In 1993 the British Psychological Society convened a working party in response to the concern that some psychologists might be inadvertently implanting false memories of child sexual abuse in their clients. The working party’s conclusions that both genuine recovered memories and false memories were likely to occur were endorsed in a subsequent article by Dan Wright, James Ost and Chris French, published in The Psychologist in June 2006, and a series of guidance documents have since been made available to members. The concerns reflected in these publications remain current, as accusations of historic child sexual abuse continue to increase in the wake of the revelations concerning Jimmy Savile and the setting up of the Independent Inquiry into Child Sexual Abuse by the UK government in 2015.

There is now widespread agreement on the existence of false memories of sexual abuse and on the immense harm they can cause. In our work both for the defence and prosecution, however, we have noticed that the events featuring in these cases by no means always support the original account that primarily identifies therapists as actively setting out to suggest or implant false memories of abuse. In this article we revisit the evidence and ask whether it is time to adopt a broader understanding of the issues involved.

The original view of false memory creation, for example that stemming from research by Elizabeth Loftus, proposed the following typical scenario: Clients with no suspicion of having been abused enter treatment with a therapist who suggests their problems are likely to stem from repressed memories of child sexual abuse. The therapist encourages them to recover the memories using hypnosis, guided imagery or related techniques. The clients are persuaded by the therapist to treat the resulting material as fact, and typically go on to create ever more elaborate ‘memories’ based on suggestion or fantasy.

This account is prominent on websites of groups such as the British False Memory Society, and we have often heard it repeated by experts in court. According to this account, genuine abuse is rarely forgotten and therefore accounts of recovered memories are usually false and the product of inappropriate therapy.

Three claims are typically made to support this argument:
•    Experimental studies show false memories of childhood events are easily created in the laboratory.
•    There is no scientific evidence for a repression mechanism in memory.
•    Surveys show that therapists typically have little understanding of memory and many use inappropriate suggestive techniques with their clients to recover memories.

But to what extent are these claims valid? Here we now briefly review the research conducted by ourselves and others to address each of these issues.

Suggesting false childhood events in the laboratory

Three types of study have been used to assess the ease of experimentally suggesting complete childhood events. (The paradigms used and the results obtained are described in detail in our review article published this year: Brewin & Andrews (2017)) In the imagination inflation paradigm participants are typically given a checklist of distinctive events that might have happened in childhood (such as putting one’s hand through a window) and rate how confident they are that each one occurred. The original 1996 study by Maryanne Garry and colleagues asked participants to imagine events they rated as unlikely to have occurred, to answer questions about the events as if they had happened, and then to re-rate their confidence that they had experienced the events. This mimics the guided imagery thought to be used by some therapists. In the false feedback paradigm, participants rate the confidence with which they believe they experienced certain childhood events (e.g. that they got sick after certain foods). They are then provided with false feedback that the particular experience was likely to have happened to them at that time and re-rate their confidence that they experienced it. This mimics therapists telling clients that their problems are likely to stem from repressed abuse memories.

The most well-known of the three paradigms involves what has come to be known as memory implantation. In these studies the experimenter targets a particular event (such as being lost in a shopping mall, in the original 1995 study by Elizabeth Loftus and Jacqueline Pickrell) that a parent indicates did not happen, and then encourages participants to recall over two or three sessions the details of the false event they are misleadingly told the parent has confirmed as happening. In some cases they may be shown a doctored photograph that supposedly illustrates their presence at the false event. These accounts are then rated for their correspondence to a complete memory by the investigators.

In our review, which is accompanied by expert commentaries, an editorial and our response discussing the nature of this scientific debate, we followed numerous cognitive psychologists in distinguishing beliefs that an event happened (which may be present without any memory) and recollective experiences of the event, noting that such experiences are not necessarily accepted as real. We reasoned that for participants to be judged as having fully accepted a false memory of a childhood event, they must report a recollective experience – usually consisting of a visual image – and be confident as well that this experience corresponds to a real event. The imagination inflation and false feedback studies often succeed in increasing the belief that the suggested event occurred by a statistically significant amount that is typically small in absolute terms, but rarely assess the nature of any recollective experiences. Only eight of the 22 memory-implantation studies we reviewed assessed whether a false recollective experience was accepted as real, and on average about 15 per cent of participants appeared to fully accept a false memory. Studies by researchers such as Henry Otgaar suggest that self-report ratings of accepting a memory as real are more conservative than the observer ratings of false memories that the studies report, so it is possible that the actual figure may be lower than this.

After and during the recall attempts participants in these implantation studies are instructed to provide accounts of the false event, and sometimes of comparison true events that have been confirmed by the parent. In the majority of studies they are also guided to imagine the false events as if they had happened. Rather than childhood memories being easy to implant, therefore, a more reasonable conclusion is that they can be implanted in a minority of people given sufficient effort. Also, does the use of doctored photographs and family members pretending to be eyewitnesses make these experiments unlike the therapeutic situation, in which there is scope for suggestion but not deception on this level? Finally, we note that while researchers have been able to implant single false childhood memories, we are not aware of them implanting memories for repeated events, which often feature in abuse allegations.  

Does repression exist?

Although the original false memory account relied heavily on the lack of any scientific evidence for unconscious repression as an explanation for the forgetting of traumatic events, there was little discussion of alternative mechanisms that might offer a plausible account of how people could forget what appeared to be memorable events. For example, can people choose to forget? Even Freud could not make up his mind whether repression was a deliberate or unconscious process and used the term in both senses.

While evidence for repression as an unconscious mechanism is not yet forthcoming, there is a substantial body of evidence concerning the effectiveness of deliberate strategies of forgetting and their neural underpinnings. Roland Benoit and Michael Anderson have distinguished between direct suppression, the attempt to not think about something (analogous to conscious repression), and thought substitution, its replacement with an alternative thought. Direct suppression makes it harder to retrieve a memory through the mechanism of activity in the prefrontal cortex inhibiting retrieval processes in the hippocampus. Thought substitution, in contrast, involves occupying the limited focus of awareness with a substitute memory. These findings underscore that memory is not a passive process, and that forgetting can be influenced by at least two separate control mechanisms.

Just as experimental research on false memory implantation cannot prove what happens in the clinic, the possible presence of these control mechanisms in the forgetting of childhood abuse has not yet been examined. Nevertheless, we would argue that the principle of the mind inhibiting unwanted memories, as described in psychoanalytic theories of repression and dissociation, is scientifically plausible, and there is evidence that it may occur in response to stress. A related phenomenon is ‘dissociative amnesia’, which involves a more widespread reversible deficit in memory retrieval that is not attributable to brain damage. It typically affects autobiographical memory for events occurring prior to a stressful event and is well recognised in the context of exposure to trauma (Staniloiu & Markowitsch, 2014). A 2010 study led by Hirokazu Kikuchi suggests that the underlying neural mechanisms may be similar to those involved in direct suppression. Finally, it is important to take a developmental perspective, considering for example how memories change qualitatively with age and how early trauma may affect memory by leading to a fragmented sense of self (Brewin, 2012).

Beliefs and practices of psychological therapists

At least 11 surveys since 1994 have questioned therapists about their beliefs concerning the validity of recovered or repressed memories and/or the possibility that such memories could be false. Caution is needed in their interpretation and generalisability as response rates in most surveys are very low – the three surveys since 2000 have not achieved rates above 17 per cent. The vast majority of clinical psychologists and licensed psychotherapists believed that repression existed in the two studies that asked the question. Two surveys that questioned qualified clinical practitioners who were also members of the British Psychological Society found that almost all believed recovered memories were accurate at least sometimes (Andrews et al., 1995; Ost et al., 2013), although few believed they were always so. Forty-three per cent of clinical psychologists in a US survey agreed that ‘repressed memories can be retrieved in therapy accurately’ although none strongly agreed (Patihis et al., 2014). Because the survey questions did not specify whether they referred to the unconscious or deliberate forms of repression, we have questioned whether these results really mean, as the researchers suggested, that there is a science–practice gap, with clinicians being poorly informed (Brewin & Andrews, 2014).

The vast majority of therapists participating in surveys also believed that false memories are possible. The earliest study included family therapists and hypnotherapists and found at least 79 per cent endorsed this possibility, rising to 89 per cent among trained hypnotherapists. More recently, over 95 per cent of clinical psychologists in the US agreed (Patihis et al., 2014). Of the studies reporting lower rates of belief in false memories, two included the already mentioned surveys of BPS members where the question was qualified by asking about the possibility of false memories of repeated childhood sexual abuse. Comparing like with like, this was endorsed by 67 per cent of the Chartered Psychologists and the psychotherapists in Andrews et al.’s (1995) survey, and 68 per cent of the subsample of BPS Chartered Psychologists participating in Ost et al.’s (2013) survey (our calculation, factoring in 27 per cent who didn’t answer the question). This is in contrast to another subsample of BPS Chartered Psychologists included in Poole et al.’s (1995) study of whom 88 per cent endorsed false memory possibility in response to the same question without the ‘repeated’ qualification.

Therapists’ responses to false memory questions seem to depend crucially on specific wording and can change if they are given options beyond the usual yes/no choice or a chance to elaborate. BPS member practitioners from Andrews et al.’s (1995) survey who had seen clients with recovered memories participated in a subsequent in-depth interview study. While 47 per cent of them had originally stated in response to a yes/no question that false memories of repeated abuse were not possible, with extra response options just 15 per cent thought they were not possible, with 73 per cent believing they were possible but unlikely, and 12 per cent that they were possible and likely (Andrews, 2001). This study also provided insight into the reasons behind such beliefs – although half of those who made further comments attributed false memories to therapists and their practices, the other half also implicated the symptoms and difficulties experienced by highly vulnerable and disturbed clients.

Deeper understanding also emerges from the few surveys that have asked therapists whether they actually use particular techniques to help clients remember child sexual abuse. The most highly cited paper-and-pencil survey reported that 71 per cent of US and UK respondents had used at least one therapeutic technique from a specified list for this purpose (Poole et al, 1995). It is difficult to reconcile this substantial rate with the fact that over 90 per cent of all these respondents also believed that false memories were possible. Could the necessarily brief survey items capture the stage at which therapists used such techniques, and whether they were used with clients who had actually forgotten their reported abuse? In our study to address these issues, we found that our sample of BPS member practitioners had used techniques to aid recall in 42 per cent of their recovered memory cases. This rate reduced to 21.5 per cent when they were used before any memory recovery started, with a further reduction to 16 per cent when they were used with clients who did not have any prior memory of abuse (Andrews et al., 1999; Andrews, 2001). These figures suggest that while the prevalence in the 1990s of inappropriate memory techniques was probably less than had been claimed, there was still a significant minority of qualified practitioners who lacked knowledge about good practice.

A broader perspective

Although we agree that the original account of false memory creation remains valid, we think that there are other explanations for many instances of false memories of abuse occurring today. False memories of childhood events can be implanted in the laboratory but this is difficult to do, it relies on procedures such as deception that make it different from therapy, and only a minority of people appear susceptible. The focus on unconscious repression has been superseded by greater understanding of how trauma impacts on the developing self and of how neural mechanisms underpin the deliberate exclusion of unwanted material from consciousness. It appears that many recovered ‘memories’ first occur outside therapy or in the absence of suggestive techniques. The vast majority of practitioner psychologists now have views that are consistent with professional guidelines, although less-qualified therapists are still a major source of concern as they appear to be less well informed about memory than psychologists (Brewin & Andrews, 2014).

Our experience in the courts is consistent with these findings in that we have only rarely come across examples of therapists setting out from the start to recover memories of abuse. We have much more frequently come across complainants who, when they began therapy, had already recovered their ‘memories’ or had started to, or who appeared to recover ‘memories’ spontaneously during a period when they were receiving therapy. As noted by the BPS therapists interviewed in our study, this places greater weight on factors such as reality monitoring, the need to distinguish the products of thoughts, imagination and dreams from what has actually occurred (Johnson, 2006). From this perspective it is important to appreciate how convincing, as well as disturbing, apparent recovered ‘memories’ of traumatic events can be.

Recovered ‘memories’ are often involuntary and can involve repeated reliving of the event, accompanied by marked sensory detail and emotional arousal. High levels of sensory detail are normally associated with true rather than false recollection, but if the apparent recollections are in fact false, the occurrence of this feature increases the likelihood that they are incorrectly labelled as true (Brewin et al., 2012). Clients may also assume that the intensity of their emotional response signifies that the ‘memory’ must correspond to reality.

As has been described in the reality monitoring literature, judgements about whether mental experiences reflect imagined or real events can also be influenced by the person’s cognitive characteristics (e.g. hypnotisability or creative imagination), prior knowledge of similar events, beliefs, cultural factors, repeated imagining and the influence of other people (Johnson, 2006). We think the same constellation of internal and external factors, singly or in combination, contribute to those situations in which people have compelling, yet false, memories of abuse. When disclosed within therapy to a qualified psychologist, these interpretations of experience are likely to meet with a neutral response that preserves the therapeutic alliance and at the same time permits the client to explore the experience in more depth, considering all possible explanations.

We suspect that a minority of therapists who are less qualified and experienced may still uncritically endorse the client’s interpretations without careful consideration of other possibilities; for example because the therapist shares erroneous assumptions about memory or because the therapy is exclusively non-directive and supportive. Other therapists may use techniques involving an element of free association without educating the client about the possibility of false memories. We have commonly found that clients have questioned the veracity of their experiences at some point but often lack the relevant knowledge about how misleading memory may sometimes be. In the context of the disbelief and scepticism that only too often surrounds abuse disclosures, it is perfectly understandable that some clients, convinced their memories are true, may seek out therapists who do not question their beliefs.

Recovered memories lie on a spectrum from being plainly false, being plausible but lacking in corroboration, to being independently corroborated. In seeking to explain those false memories that do arise within therapy, our account in no way excludes the possibility of therapists acting inappropriately, but places more weight on pre-therapy reality monitoring and on the interactions between a therapist and a client struggling to make sense of what are often distressing and confusing experiences. Either uncritically accepting false memories, or disbelieving genuine recovered memories, has the potential to do immense harm. It is therefore essential that we continue to encourage debate and education around these contentious issues.

- Chris R. Brewin is Professor of Clinical Psychology at University College London
[email protected]
- Bernice Andrews is Emeritus Professor at Royal Holloway University of London

Key sources

Andrews, B. (2001) Recovered memories in therapy: Clinicians’ beliefs and practices. In G. M. Davies & T. Dalgleish (Eds.) Recovered memories: Seeking the middle ground (pp.189–204). Chichester: Wiley.
Andrews, B., Brewin, C.R., Ochera, J. et al. (1999). Characteristics, context and consequences of memory recovery among adults in therapy. British Journal of Psychiatry, 175, 141–146.
Andrews, B., Morton, J., Bekerian, D.A. et al. (1995). The recovery of memories in clinical practice. The Psychologist, 8(5), 209–214.
Benoit, R.G. & Anderson, M.C. (2012). Opposing mechanisms support the voluntary forgetting of unwanted memories. Neuron, 76(2), 450–460.
Brewin, C.R. (2012). A theoretical framework for understanding recovered memory experiences. In R.F. Belli (Ed.) True and false recovered memories: Toward a reconciliation of the debate. Nebraska Symposium on Motivation (Vol. 58, pp.149–173). New York: Springer.
Brewin, C.R. & Andrews, B. (2014). Why it is scientifically respectable to believe in repression: A response to Patihis, Ho, Tingen, Lilienfeld, and Loftus (2014). Psychological Science, 25(10), 1964–1966.
Brewin, C.R. & Andrews, B. (2017). Creating memories for false autobiographical events in childhood: A systematic review. Applied Cognitive Psychology, 31, 2–23.
Brewin, C.R., Huntley, Z. & Whalley, M.G. (2012). Source memory errors associated with reports of posttraumatic flashbacks. Cognition, 124(2), 234–238.
Johnson, M.K. (2006). Memory and reality. American Psychologist, 61, 760–771.
Loftus, E.F. (1993). The reality of repressed memories. American Psychologist, 48(5), 518–537.
Ost, J., Wright, D.B., Easton, S. et al. (2013). Recovered memories, satanic abuse, dissociative identity disorder and false memories in the UK: A survey of clinical psychologists and hypnotherapists. Psychology Crime & Law, 19(1), 1–19.
Patihis, L., Ho, L.Y., Tingen, I.W. et al. (2014). Are the ‘memory wars’ over? Psychological Science, 25(2), 519–530.
Poole, D.A., Lindsay, D.S., Memon, A. & Bull, R. (1995). Psychotherapy and the recovery of memories of childhood sexual abuse. Journal of Consulting and Clinical Psychology, 63(3), 426–437.