In a previous article for the Skeptic, I referred to the fact that I sometimes act as an expert witness in cases where allegations of non-recent sexual abuse may potentially be based upon false memories. This is a very small minority of sexual abuse cases and, even then, in about half of the cases that I have dealt with I have concluded that there were no obvious indications in the evidence that I had seen to indicate a heightened probability that this was indeed the case. Every case is different and sometimes a case will raise questions that I cannot answer without a deep dive into the scientific literature. In this article, I want to discuss one such case.
Without going into details, suffice it to say that the complainant, whom I shall refer to as C, alleged at the age of 40 that he had been sexually abused in various ways by the accused, whom I shall refer to as A, when C was between the ages of 8 and 16 years. As an adult, C had recurrent dreams of being sexually abused as a child. The evidence suggested that he had then recovered memories of being so abused at the hands of the accused. The evidence suggested that C believed his nightmares about being sexually abused as a child indicated that he had actually been thus abused, and had repressed those traumatic memories. The idea that repressed traumatic memories might surface in dreams is a common one and one that often features in works of fiction. But is it true?
Much has been written regarding the relationship between memory and dreams, and a comprehensive review of the topic is not possible in this short article. Instead, I will focus upon what I felt were the central questions to be addressed. To start with, I wanted to know: are details of everyday real-life events sometimes replayed during dreams?
There is strong evidence for the idea that details from recent events in waking life are sometimes incorporated into our dreams, indicating that dreams may in some way be involved in consolidation of memories. This effect is strongest two nights after the events in question with a second peak 5-7 days after the events. It should be noted that we are referring here to recent events in the dreamer’s life, not to events from decades earlier. Furthermore, the dreams are not accurate replays of whole episodes. Instead, they simply incorporate elements from the individual’s waking life. For example, Fosse and colleagues (2003) reported that 65% of 299 dream reports from 29 people over a period of a fortnight incorporated some details from recent waking life but only 1-2% preserved the original characters, actions, objects, and setting of the original scene.
Some theorists claim that memory for trauma involves different psychological and neuropsychological mechanisms than memory for non-traumatic events. For example, it is claimed that those suffering from posttraumatic stress disorder (PTSD) really do experience accurate replays of the trauma they experienced in both their recurrent nightmares and their waking “flashbacks”. Van der Kolk and colleagues (1984) reported that eleven of fifteen Vietnam veterans with PTSD reported dreams that were exact replays of the trauma they had experienced during combat. Schrueder, Kleijn, and Rooijmans (2000) reported that 42% of 102 World War II survivors reported accurate replays, albeit that 35% reported experiencing distorted versions of the original trauma. A major limitation of these studies is that they were based upon nothing more than self-report, without asking sufferers to prospectively record their dreams.
However, even studies using an improved methodology that involved getting PTSD sufferers to keep dream diaries for a period and to then assess their entries supported the claim that a minority of PTSD sufferers do indeed report nightmares that they believe replicate actual traumatic events. For example, Mellman and colleagues (2001) had 60 patients, hospitalised as a result of assault or life-threatening accidents, record their dreams during their initial stay in hospital. Eighteen patients reported a total of 21 dreams, of which ten were trauma-related; six were reported as exact replays, four as distorted versions.
It is worth pausing at this point to reflect upon what would be required if PTSD nightmares and flashbacks really were an accurate and exact replaying of the original trauma. This would require that memory worked like a video camera, accurately recording every detail of an experience so that they could be “replayed” later. Although this view of memory is held by a very large proportion of the general public, it is totally incorrect. We typically remember the gist and central details of experiences, and forget the vast majority of the specific details. Thus, it is simply impossible for any nightmares or flashbacks to really be exact replays even though the sufferer may sincerely believe that they are. When a person compares their memory of the original event to their memory of a dream of that event, they are inevitably comparing one unreliable reconstruction with another.
With respect to this specific case, it should also be noted that PTSD sufferers do not, in the absence of physical brain injury, experience any period of amnesia for the events that caused the PTSD. On the contrary, they have great difficulty in keeping the traumatic memories out of consciousness. This observation is itself evidence against the classical psychoanalytic notion of repression; that is, the idea that traumatic memories will tend to be automatically pushed into a non-conscious part of the mind, from whence they could not be retrieved no matter how strong the memory cue. In fact, as discussed in a previous article, there is little, if any, evidence to support the existence of repression, despite it being an idea that is widely accepted by both members of the general public and a range of relevant professional groups.
There is more than one way in which dreams may result in false memories. The most obvious and direct route is that which involves a person mistaking the events that they experienced within a dream for events that took place in objective reality. This would be an example of what psychologists refer to as a reality monitoring error. Reality monitoring refers to the psychological processes that allow us to distinguish between events taking place in objective reality and those taking place solely within our minds, such as dreams, fantasy, and hallucinations. Rassin and colleagues (2001) collected data from two samples of respondents (N = 85 and 255, respectively) and found that,
a nontrivial minority of respondents (11.8% and 25.9%, respectively) reported that they had had the experience of not being able to discriminate between dream and reality.
Page 478
However, in this particular case it was clear that A was always aware when he awoke from his nightmares that he had indeed been dreaming. He subsequently correctly recalled such dreams as being dreams. How then could he have ended up “recovering” detailed false memories of actually being abused? One plausible route is that of imagination inflation. When people repeatedly imagine events which they initially believe they have never actually experienced, this can lead them to subsequently believe that they did indeed experience those events. In some cases, they may also end up with detailed false memories of the events.
This is again an example of a reality monitoring error. In this case, they are confusing events that they have merely imagined with events that have really taken place. Given that C assumed that his dreams of being abused as a child were a possible indication that he really had been, it is natural that he may have imagined various scenarios in which such abuse might have occurred, potentially ultimately resulting in false memories of such abuse.
In this case, as in all such cases, in the absence of independent evidence we cannot know for sure if C’s allegations are true, deliberate lies, or sincere but mistaken. In our adversarial legal system, justice requires that members of a jury make their decision regarding guilt or innocence in the context of a full understanding of relevant factors that can affect human memory.
