How is memory affected by traumatic events? An evaluation of the basic tenet of repression theory


The Freudian concept of repression and its modern heir «dissociation» have been much debated in contemporary psychology. The debate has centered on the effect of trauma on memory. In this essay, Mona Hide Klausen reviews briefly elements of the debate and contemporary research findings.

By Mona Hide Klausen (published 17.01.2009)

» Last ned PDF-versjon av denne artikkelen

1. Introduction

The nature of traumatic memories was at the core of a debate that raged during the 80s and 90s (e.g. McNally, 2003, pp. 1-2), and which is lingering at the present date (e.g. Rofé, 2008). The debate, referred to as “The Memory Wars”, was marked by deep and profound disagreement between professionals. On one side were those that supported the traumatic memory argument; that traumatic memories could be repressed, in order to protect the mind of the trauma victim. This notion was based on the psychoanalytic theories of Sigmund Freud (1856-1939,, 2008). During these two decades a series of women and men came forth claiming to have been sexually abused in childhood, and to have repressed these events up until entering therapy, where the memories had been dug out, in pristine form (e.g. Loftus & Ketcham, 1994; McNally, 2003; Schacter, 2001). On the opposing side of “the war” were parents claiming to have been falsely accused of abuse, alongside concerned researchers and psychologists, who suspected that some of the abuse memories were false, having been created in therapy as a side effect of suggestive therapy techniques (Schacter, 2001, pp. 123-130).


The theoretical basis for the debate is the assumption that traumatic memories affect the brain in a different manner than “normal” events do. The memories are supposedly “split up”, and stored in the brain in a fragmented manner. The victim is therefore unable to integrate these pieces into a coherent whole, in order for a memory to arise in consciousness. The memories are thus inaccessible (e.g. Conway, 1997; Herman, 1992, in Porter & Peace, 2007). The opposing view states that traumas do not affect the brain in a special manner, and they are subject to the same kinds of distortions and deteriorations as memory for ordinary events. In addition they claim that traumas are rather difficult to forget (Porter & Peace, 2007).


Since both cannot be right, which viewpoint is best supported by evidence?

The aim of this paper is to look into empirical studies pertaining to answer this question. For simplicity; those supporting repression theory will be referred to as ‘traumatic memory theorists’, and the opposing view ‘ordinary memory theorists’.1


A third view that has gained support in more recent years has been dubbed ‘the trauma superiority theory’, claiming that traumatic memories are special, but in the sense that trauma enhances and preserves memory, rather than impair it (e.g. Peace & Porter, 2004; Shobe & Kihlstrom, 1997). This theory will also be evaluated here.


Trauma has been defined in a number of ways, but generally they involve an extraordinarily stressful event, where the life or well being of the victim, or someone close, is at risk, and involves overwhelming stress or fear. Some also highlights the interpretation of the event (see Cordon et al., 2004).


The paper will be split in 4 parts starting with defining and explaining central concepts, followed by a description of the underlying assumption of repression (that traumas are made inaccessible to consciousness; i.e. amnesia), before evaluating empirical evidence based on this assumption. Finally, a brief mention of alternative explanations will be given.


2. Repression theory

The theory of repression was first introduced by Sigmund Freud, and was thought to be an unconscious defence mechanism that banned unwanted thoughts, desires and trauma from consciousness. The memories were supposedly hidden away in a corner of the brain, “out of sight and out of mind”, and completely inaccessible. These memories caused various problems for the person repressing the aversive memories, and therefore they had to be uncovered so they could be processed properly, and the person could move on (e.g. Geraerts & McNally, 2008; Loftus & Ketcham, 1994 pp. 49-50; McNally, 2003, pp. 5-6; 159-168). (The various techniques used to recover memories will not be dealt with in this paper; for a brief review see Loftus & Ketcham, 1994, pp. 150-175).



Today, most researchers talk about dissociation, not repression (c.f. McNally, 2003, p. 172). Dissociation can be defined as “a disruption in the usually integrated functions of consciousness, memory, identity, or perception of the environment” (DSM IV, APA 1994, p. 766, in McNally, 2003, p. 172). The definition of dissociation makes it rather difficult to interpret, since the nature of the disruptions seems unclear. Conway (1997) defines a dissociated state as “[narrowed attention] to very specific aspects of the environment or internally generated cognitions” (Conway, 1997, p. 172), while de Prince and Freyd (2004) seem to claim that high dissociators have troubles in selectively focussing attention.

However; since the result of dissociation is thought to be somewhat the same as with repression (unconscious blocking of painful stimuli, for decades, upon retrieval in therapy, Shobe & Kihlstrom, 1997) I will focus on the “classic” repression theory, since that is more clearly defined. The main focus of this paper is the possibility of “removing”, or blocking, memories from conscious awareness, not the name of the concept one infers to be behind it. Falsifying the assumption that traumas are more easily and completely forgotten, will undermine both concepts.


It is also important to stress the difference between repression, the unconscious blocking of painful memories, and suppression, consciously choosing not to think about an event. Suppressed memories are accessible and may surface with appropriate clues, while repressed memories may never return (Golding & Long, 1998, in Epstein & Bottoms, 2002).


3. Amnesia

The reason for the apparent unavailability of the memories is thought to stem from what is called traumatic amnesia, and any lack of memory has been taken as evidence of repression (McNally, 2003; 2004). The concept of amnesia lies at the core of the debate about repression and recovery of traumatic memories (e.g. McNally, 2003; 186). McNally (2003; 2004) points out that amnesia indicate an inability to remember even if one tried to. I will adhere to that definition here. Some theorists talk of “partial or complete forgetting” (e.g. Freyd, 1996), but how can forgetting of parts of an incidence be juxtaposed to forgetting ever having experienced a traumatic event at all (e.g. McNally, 2003)? Various types of amnesia have been implicated in the recovered memory debate, some will be described briefly, in addition to other phenomena that have been confused with amnesia.


Amnesia by betrayal

Jennifer Freyd (1996) postulated that abuse committed by a trusted caretaker would be more likely to lead to amnesia, because it involved “betrayal by a trusted, needed other” (Sivers, Shooler & Freyd, 2002; 169). In order for the child to keep a relationship with the caretaker the child would become unaware of the betrayal for the sole purpose of survival (Freyd, 1996).


Yes, maternal support is significantly related to disclosing abuse (e.g. Epstein & Bottoms, 2002; Goodman et al., 2003), but a more plausible explanation would be lack of rehearsal, not necessarily betrayal. Not talking about something diminishes its possibility of being properly consolidated in long term memory (e.g. Epstein & Bottoms, 2002). Goodman & Melinder (2007) also found that children with secure attachment told about more negative events than children with unsecure attachment. Although it could be argued that the children had secure attachment to the caretaker because they had amnesia for abusive incidents, they probably would not be talking about other negative events if this was the case.


Psychogenic amnesia

Psychogenic amnesia (aka dissociative amnesia) is thought to result from “events whose psychological or emotional meaning produces memory loss without damaging the brain” (McNally, 2003; 186), and therefore thought to be the cause of amnesia for traumatic events. However, in psychogenic amnesia, the amnesia begins immediately after the traumatic event, lasts for a relatively short period of time (rarely more than a few weeks), recovery is sudden, and psychotherapy is rarely required to restore memory. One therefore cannot compare it to the traumatic amnesia were memory is supposedly gone for decades, and returns gradually in therapy (McNally, 2003; 189-190). In addition, the person suffering psychogenic amnesia typically has severe retrograde amnesia, and may even forget who (s)he is. In” traumatic amnesia”, the memory deficit is selective (involving only memories of abusive events), and the sufferer does not loose personal identity. Psychogenic amnesia can also be triggered by a non-traumatic event. They are therefore not the same (McNally, 2004).


Organic amnesia

In order to confirm a memory loss as resulting from trauma, other factors must be ruled out. One such factor is organic deficits that results in amnesia. Direct physical insult to the brain that results in memory loss, cannot be taken as evidence of traumatic amnesia that is purely psychological in nature (McNally, 2004).


Incomplete encoding

Memory is based on the three processes of encoding, storage and retrieval (Passer & Smith, 2008). If something is not properly encoded, it cannot be properly stored, therefore retrieval is unlikely. In a traumatic event our attention may also be focused on a central aspect of the situation, thus not noticing other, less important, details, i.e. being robbed at gunpoint leads to weapon focus (“the victim is so focused on the weapon that (s)he does not remember the robber”, e.g. Magnussen, 2004, p. 84) This does not constitute amnesia, however (e.g. McNally, 2004), since that would indicate that the memory is stored, but unavailable (e.g. Geraerts & McNally, 2008; McNally, 2003, p. 186).


Not thinking about it

Some therapists have used retrospective assessment of possible amnesia, asking participants if there “ever was a time where they could not remember the abuse” (Briere & Conte, 1993, p. 24, in McNally 2004, p. 99). This ambiguous question is not a very good assessment of repressive amnesia, since not thinking about an event, or even trying not to think about an event because it was too upsetting, could be plausible interpretations (e.g. Epstein & Bottoms, 2002; McNally, 2004). It therefore cannot be taken as proof of traumatic amnesia. In addition, some also forget having remembered (McNally, 2004), undermining the validity of the assessment even further. Also; “not being able to remember” could also indicate an attempt to remember, but if something has been repressed, i.e. lost from conscious awareness, how can you know what you are supposed to try to remember (e.g. Laney & Loftus, 2005; McNally, 2004)?



Another problem with studies of memory for sexual abuse is that the victim may not be interested in talking about it, for various reasons. This hardly equates with an amnesia for the event, however (e.g. Goodman et al, 2003; McNally, 2004), and might lead to inflated estimates of the prevalence of forgetting trauma.


Childhood amnesia

Childhood amnesia refers to the lack of episodic memory before a certain age, usually 3-4 years (e.g. McNally, 2003, p. 43).


Some trauma theorists have claimed that it is possible to remember events that took place during this period, and this has made its way into popular self-help books (e.g. Bass & Davis, 1988; 29). However, memories of events taking place during the childhood amnesia period are rarely, if ever, transferred into adulthood, and numerous studies have found that adults rarely remember any events prior to their 3rd or 4th birthday, some even later (e.g. McNally, 2003, pp. 43-48). Including children younger than 3 or 4, as in Williams’ (1994) study would therefore yield an inflated amnesia rate, because the children are too young to remember. The memories are not unconsciously repressed.


4. Traumatic memories vs. “normal” memories

In order to judge traumatic memories as different from other memories they should differ in significant ways. Various studies have attempted to compare negative, emotional memories with memories for both positive and neutral events. Neutral events differ from positive and negative events on emotional activation, and positive and negative events differ on valence. In order to judge negative emotional memories as “special” they should be unique on both emotional activation and valence. In other words they have to be qualitatively different from both neutral (lacking activation) and positive (a different valence) memories. According to repression theory, there should be less memory for strongly negative stimuli (cf. 2.0).


Let’s first look at emotionality. Do traumatic (highly emotional) stimuli differ from neutral (lacking emotional activation) stimuli? A study by Payne & Corrigan (2007) used a directed forgetting paradigm to see if emotional stimuli are more easily blocked from consciousness than neutral stimuli.


The directed forgetting paradigm usually consists of two lists of words that the participant is told to memorize. After the first list has been presented they are instructed to forget the first list, because it is supposedly only a practice list. They are then told to remember the second list. Control groups are instructed to remember both lists. On a later remembering task they are suddenly tested for words from both lists, including the one they had been instructed to forget. If they report fewer words from the “forget” list than the “remember” list, directed forgetting is said to have occurred. In addition they tend to remember more words from the “remember” list than control groups, indicating that the first list have not interfered with learning the second list. Together they make up the directed, or intentional, forgetting effect (e.g. Payne & Corrigan, 2007).


Studying word lists have been common in these types of studies, but reading a list of words may not actually invoke that strong emotions, thus lack ecological validity (e.g. Payne & Corrigan, 2007). Payne & Corrigan (2007) therefore used images in their study, in order to elicit stronger emotions. A pilot-test indicated that the images they had selected succeeded in producing an emotional state. They were mostly interested in emotional significance, and therefore the emotional condition also included positive images. One of the lists contained emotional pictures, the other neutral, and the presentation were counterbalanced. They found a main effect of emotion; more images were remembered from the emotional list compared to the neutral list. In addition they found a main effect of list in that more words were remembered from list 2 than list 1. This is in line with previous research on directed forgetting.


The interesting part of this study is that an intentional/directed forgetting effect was observed when the first list was neutral (more easily forgotten), but not when it was emotional. In that condition the directed forgetting seemed to break down. In fact: pictures from the first list (which they had been instructed to forget) interfered with the learning of words in the second list (which contained neutral words) (Payne & Corrigan, 2007). They therefore concluded that “the findings support the claim that intentional forgetting is more difficult for emotional events than for emotionally neutral ones” (Payne & Corrigan, 2007; 784). The study got results that seem in line with studies on people experiencing traumatic events – emotional stimuli are rather difficult to forget.


Porter & Birt’s (2001) study provides insight into the nature of traumatic vs. positive events, i.e. valence. They tested three underlying principles of the traumatic memory argument:


1)    Because of the fragmentation of traumatic memories, they should be more emotional and sensory. They would lack in visual vividness, narrative details, and having poorer overall quality.

2)    They should be less often talked, or thought, about than non-traumatic memories.

3)    More forgetting would be reported for traumatic events than the positive events (Porter & Birt, 2002).


They asked respondents to describe their most traumatic memory, and their most positive memory. Afterwards the participants were asked several questions pertaining to the quality of their memories.


They reported that traumatic memories “had fewer sensory components … had been thought about more often … contained significantly more details … and more references to emotional state at the time of event” (Porter & Birt, 2002; 107). This partially contradicts the three underlying assumptions. In line with the first assumption, traumatic memories were rated as more emotional. Although a significant difference of sensory components were reported, it turned out that this only applied to taste, with fewer taste components in the traumatic memories, compared to positive memories. Apart from this, positive and traumatic memories were quite similar (Porter & Birt, 2001).


However, those that reported more stress also reported thinking about the trauma more, discussing it more, and the memories were more vivid/clear, than for those reporting lesser stress in reaction to the trauma (Porter & Birt, 2001). This seems to contradict the notion that the more severe the trauma, the less of it is remembered (e.g. Peace & Porter, 2004).


A longitudinal study by Peace & Porter (2004) also examined the difference between positive and traumatic events. They interviewed participants on two occasions separated by approximately 3 months, about a recent traumatic event (less than a year ago) and a positive event, to see if either memory changed over time, and whether traumatic memories changed more than positive. The memories did not differ significantly on ratings of emotionality (p>.05).


The study found that traumatic memories were significantly more consistent over time and more detailed than positive memories, which seemed to decline over time, supporting the trauma superiority theory, but not the trauma equivalency theory. In addition, severity of trauma (low, medium, high) did not affect memory consistency; more severe trauma did not impair the memory (Peace & Porter, 2004), contradictory to the trauma memory argument.


However, although Peace & Porter (2004) argued that three months should be enough time for the memories to fade, it might be that this is to short an interval to accomplish this. A follow up study addressed this issue. In Porter & Peace (2007) the same sample in Peace & Porter (2004) were retested approximately 3.5 and 5 years later, to see if traumatic events and positive events had declined in consistency. They found that “traumatic memories were recalled more vividly, were of higher quality, and contained more sensory components than positive memories” (Porter & Peace, 2007; 439) further supporting the trauma superiority argument.


Payne & Corrigan (2007) ran an additional analysis where they compared pleasant and unpleasant pictures in the emotional list, and found a significant difference, in that more negative than positive images were remembered, although both emotional pictured resisted the forgetting instruction (Payne & Corrigan, 2007). This study seems to answer our initial presumption; traumatic events should differ both on emotionality (emotional pictures remembered better than neutral) and valence (negative remembered better than positive). However, the results run opposite of the assumptions of repression theory.


5. Post Traumatic Stress Disorder

A number of people who have experienced a traumatic event develop post-traumatic stress disorder (PTSD). In order to be diagnosed with PTSD certain symptoms have to be present, in addition to the (obvious) requirement of having experienced a traumatic event (Bodkin, Pope, Detke & Hudson., 2007). The remaining criteria are:

2. Re-experiencing trauma through recurrent and intrusive images, thoughts, perceptions, or dreams of the trauma. Included is also flashbacks (reliving the trauma), and enhanced reaction to trauma related stimuli.

3. Avoidance of internal or external stimuli associated with trauma, such as places, thoughts, people, etc. In addition: being unable to recall important aspects of the traumatic event, blunted affect, and loss of interest to formerly enjoyable activities. Feels separated from other people, and is also unable to imagine a future.

4. Increased arousal, such as irritability and anger, hyper vigilance, and is easily startled. In addition, difficulty falling or staying asleep, and difficulty concentrating. (, 2008).

5) Symptom duration more than 1 month

6) Significant impairment in functioning (Bodkin et al., 2007)

People suffering from PTSD are haunted with memories from the traumatic event, and they seem unable to forget (e.g. McNally, 2003; p. 9). Even so, the avoidance symptom has been taken as indicator of dissociation or repression of trauma by some therapists (see van der Kolk, 1997, in Zoellner, Sacks & Foa, 2003). Especially being unable to remember important aspects of trauma has been theorized to be the result of some form of traumatic amnesia (McNally, 2003, p.10). This aspect has been debated, however. Various researchers have found that the worst moments of a traumatic event, so called “hotspots”, correspond to the intrusive memories (see Braisby & Gellatly, 2005, p. 537). One would think that the most traumatic moment of an event would be “an important aspect”, thus should have been forgotten rather than remembered too well. In addition; if traumas lead to amnesia in order to protect the psyche, as postulated by Freud, then, logically, the most traumatic aspects should be the ones to get repressed.


What has also been found is that central details are better remembered, but more peripheral details are forgotten (e.g. Laney & Loftus, 2005; McNally, 2003), i.e. weapon focus (Magnussen, 2004, p. 84). Forgetting peripheral details can hardly be taken as evidence of traumatic amnesia for “important aspects of a traumatic event”. It could be argued that the more traumatic parts of the event are taking up so much “mental energy” that little is left to think about the less emotional moments.


Zoellner, Sacks & Foa (2003) investigated the claim that PTSD patients actually alternate between an avoidant and an arousal state (e.g. van der Kolk 1987, in Zoellner et al, 2003), and that previous studies thus may have failed to find the link between avoidance and PTSD. They used mood induction to make sure the PTSD group was in a dissociated state while participating in the directed forgetting task, to see if an avoidant state impacted their performance. In addition to the PTSD group they had a control group that did not suffer from PTSD, and they also included a serenity mood induction.


The hypothesis was that those with PTSD in a dissociated state would show an avoidant encoding style, thus ‘threat’-words they had been instructed to forget would be remembered to a lesser extent than positive or neutral words, which would show the directed forgetting effect. Contrary to initial hypothesis, the forget negative words were not forgotten more than other word types, but they did exhibit diminished memory for words they had been instructed to remember, yielding a similar forgetting rate for both forget and remember words. For the PTSD group dissociation seemed to worsen memory altogether, rather than affecting it in unique ways. The control group showed the directed forgetting effect, and it was also observed in the serenity condition. They concluded that dissociation may impair encoding processes (Zoellner, Sacks & Foa, 2003). If this is so, then dissociating during trauma should not cause memories to be tucked away safely in the brain for later recovery; the memories should not be able to enter the brain in the first place, hence there would be nothing to recover.


To complicate matters further, the rationale for the very existence of the PTSD diagnosis is also debated (e.g. Bodkin et al., 2007). Recall that experiencing a traumatic event is an essential prerequisite for acquiring the diagnosis. If that is the only thing separating PTSD-patients from people with other mental disorders could we say that PTSD is so unique to trauma sequelae that it can be said to be a diagnosis on its own terms?


A study by Bodkin et al. (2007) investigated the difference between patients with major depression disorder with and without additional PTSD diagnosis, to see if they would differ on the remaining PTSD criteria (apart from experiencing trauma). The patients were rated as having experienced trauma, not having experienced trauma, and equivocal (uncertain). For those not reporting a traumatic incident, a proxy for trauma was created, such as worrying thoughts or fears. They were then scored on the remaining PTSD criterions. Surprisingly, PTSD clinical syndrome had equal prevalence in all groups (80 %) for the other criteria, indicating that even non-traumatized people could reach diagnostic level for PTSD (Bodkin et al., 2007).


If PTSD were solely connected to experiencing traumatic events, then it should affect traumatized people in a unique way, and non-traumatized people should not be able to fulfil the other symptom cluster criteria. These findings seem to raise doubt about the validity of the diagnosis. It should be noted that a PTSD diagnosis demands “significant impairment in functioning (6)”. It is unclear how a non-traumatic, albeit troubling event, could cause this significant impairment. However, the sample used were diagnosed with major depressive disorder, thus may be more troubled by such worrying thoughts, than people in the non-clinical population.


6. Alternative explanations for “recovery” of trauma memory

Yes, memories of sexual abuse can be “not present” for several years, and then suddenly enter awareness (e.g. Brenneis, 2000; McNally, 2003, p. 219-226). Is this recovery of repressed memories?


Memory recovery can be divided into therapy recovered, where the patient has spent long periods attempting to dig out memories, and naturalistically recovered, were an external cue has lead to memory retrieval. The latter is more plausible, and they have also been found to be easier to corroborate (e.g. Geraerts, Schooler, Merchelbarch, Jelicic, Hauer & Ambadar, 2007).


Forgetting and subsequent remembering does not have to mean that the event was repressed, though. Some people are actually able to suppress (deliberately not thinking about) negative memories (Geraerts & McNally, 2007), but not thinking about something does not mean that it is unavailable (e.g. Epstein & Bottoms, 2002; McNally, 2003, p. 2). Clancy & McNally (2005-06) also found that people who had recovered memories outside of therapy was actually better at “suppressing anxious autobiographical thoughts” (in Geraerts & McNally, 2008, p. 620). If the memories had been truly repressed they should have had to be recovered in therapy, not “recovered” on their own (e.g. McNally, 2003, in Geraerts & McNally, 2008).


Clancy & McNally (2005-06) also found that if the abuse had not been considered traumatic at the time it happened, the event was more likely to have been forgotten. This is also consistent with other research indicating that severity of abusive event to be significantly correlated with better memory for that event (Alexander et al., 2005; Goodman et al., 2003), contradicting repression theory. If something is not considered traumatic when it happened, then there should not be any motivation to forget it (Clancy & McNally, 2005-06).


Researchers have tried to find what mechanisms might explain the forgetting of traumatic events. Epstein & Bottoms (2002) asked trauma victims to indicate the reasons why they had temporarily forgotten previous abusive and traumatic events. They found that common cognitive mechanisms (such as active cognitive avoidance, relabeling and retrieval failure) were used more often as explanation for forgetting than repression. Relabeling was most often used as explanation for sexual abuse victims (Epstein & Bottoms, 2002).


This is in line with Clancy & McNally (2005-06). They asked victims of sexual abuse to indicate the severity of their abusive incident, and also to indicate how they might have forgotten the abuse. Only two (of 27) labelled the abuse traumatic. The majority reported actively trying not to think about the abuse (14/27). Only three implicated an unconscious defence mechanism, while the remaining of the sample indicated normal cognitive processes. Of the three implying some form of repression, all had been in therapy (Clancy & McNally, 2005-06), and may therefore have been influenced by therapists’ terminology or theoretical explanations.


7. Discussion

The memory war may not be as visible now as it used to, but the theories and practices are still lingering (e.g. Rofé, 2008). The assumption that traumatic events can lead to amnesia has been extensively investigated during these past 20 years. So what have we found out? How is our memory affected by traumatic events?


Various reasons for lack of memory, or lack of disclosing memory, have at some point been taken as evidence for amnesia (e.g. Loftus, Pollonsky & Fulliove, 1994; McNally, 2003; 2004). It is imperative that we manage to separate “true traumatic amnesia” from “everything else”. If therapy is based on the assumption that memories have to be dug out, when this assumption proves to be faulty, then therapists might inadvertently cause those memories to arise (e.g. McNally, 2003). Such memories are called “false memories” (“memory for an event that did not occur”, Sivers, Schooler & Freyd, 2001, p. 169) and affects patients as well as innocent people accused of something they did not do (e.g. Loftus, 1994; Schacter, 2001). False memories have caused devastation on numbers of people throughout the decades. Families have been shattered, and people have been imprisoned for gruesome crimes they never committed (e.g. Loftus & Ketcham, 1994).


As we have seen, the various types of amnesia, and other reasons why people do not report a specific event, does not equate with a psychologically induced amnesia, due to excess negative, emotional activation (e.g. McNally, 2004). If traumatic events truly did affect memory negatively, then one should observe differences between negative stimuli, and positive and neutral stimuli. Neither of the studies investigated here found any evidence to support the claim, they actually found the exact opposite (Payne & Corrigan, 2007; Peace & Porter, 2004; Porter & Birt, 2001; Porter & Peace, 2007). The important aspect of these studies is that they did not rely on wordlists, which might not be very ecologically valid (e.g. Payne & Corrigan, 2007). Porter et al. (Peace & Porter, 2004; Porter & Birt, 2001; Porter & Peace, 2007) even used autobiographical memories to assess the differences between positive and negative memories, making the studies more generalizable. Contrary to predictions based on repression theory, negative stimuli were actually remembered better (e.g. Payne & Corrigan, 2007; Peace & Porter, 2004; Porter & Birt, 2001; Porter & Peace, 2007), and it seems like the trauma superiority theory is gaining support.


Studies on post traumatic stress disorder also seem to support the trauma superiority theory. PTSD is mostly marked by intrusive memories of the event, and even in a dissociated state they fail to support the notion that the avoidance symptoms might support repression/dissociation theories (e.g. Zoellner, Sacks & Foa, 2003). “Not remembering important aspects of trauma” seems like a dubious proof of repression theory. One would think that the most important aspect of a traumatic event was, precisely, the most traumatic aspect, thus should have been forgotten, not haunting the victim (e.g. Braidsby & Gellately, 2005; McNally, 2003). When even the PTSD diagnosis is drawn into question (e.g. Bodkin et al., 2007) supporting evidence for repression theory seems virtually non-existent.  


When asked to indicate the reason for their lack of memory for an event, the majority chose “normal” memory mechanisms, over repression (Epstein & Bottoms, 2002). However, one cannot preclude the possibility that people interpret a seemingly lack of memory as repression (e.g. Epstein & Bottoms, 2002). Repression is a common term, especially in Norwegian language (e.g. Magnussen, Endestad, Koriat & Helstrup, 2007), and this may cause people to grasp this explanation because it is readily available, even though their temporal lack of memory could be best explained by other factors (e.g. Epstein & Bottoms, 2002).


Not only does cognitive theory explain the phenomenon of suddenly remembering an event that has been forgotten for years, it is also able to explain the phenomena that repression theory is unable to explain. Cognitive theory would predict that more emotional events would be remembered better, and negative events most of all. This is in line with years of research on emotional impact on memory (e.g. McNally, 2003). It follows from this that the less traumatic an event, the better the chance of the event being forgotten (e.g. Clancy & McNally, 2005-06). Even when the traumatic events have been proven to have been forgotten, and then suddenly resurfaced (e.g. Brenneis, 2000) it does not logically follow that an unconscious mechanism has been operating (e.g. McNally, 2003, p. 2). Brenneis (2000) also found the allegedly validated stories to have been released by external cues, not through therapeutic “digging”.


The law of parsimony dictates that the theory that gives the best, and simplest, explanation of a certain phenomenon should be the favoured one (, 2008). Cognitive explanations seem to hold their ground. As for repression theory; considering all the studies contradicting the basic assumptions (e.g. Peace & Porter, 2004; Porter & Birt, 2001; Porter & Peace, 2007), and the potential dangers of false memories arising in certain therapeutic settings (e.g. Loftus, 1994; Loftus & Ketcham, 1994; McNally, 2003, pp. 229-259), it may be due time to retire the concept, and the presumptions attached to it, which does not seem to be properly validated:


“[I]t may now be time to stop looking for unremembered traumatic events and begin asking what we can learn from where we’ve been. If we grant […] that a significant number of memories for trauma recovered in therapy are artifacts of the treatment process, and not memories in any sense, this phenomenon in itself should be understood” (Brenneis, 2000, p. 75).

 (Mona Hide Klausen is a Master student at the Cognitive Neuropsychology program at UiO, where she specializes in false and recovered memories.)


Alexander, K.W., Quas, J.A., Goodman, G.S., Ghetti, S., Edelstein, R.S., Redlich, A.D., Cordon, I.M. & Jones, D.P.H. (2005). Traumatic impact predicts long-term memory for documented child sexual abuse. Psychological science, 16, 33-40.

Bass, E. & Davis, L. (1988). The courage to heal. London: Vermillion.

Brenneis, C.B. (2000). Evaluating the evidence. Can we find authenticated recovered memory? Psychoanalytic psychology, 17, 61-77.

Bodkin J.A., Pope, H.G., Detke, M. & Hudson, J.I. (2007). Is PTSD caused by traumatic stress? Journal of anxiety disorders, 21, 176-182.

Braisby, N. & Gellatly, A. (2005). Cognitive psychology. New York: Oxford University Press.

Clancy, S. & McNally, R.J. (2005-06). Who needs repression? Normal memory processes can explain “forgetting” of childhood sexual abuse. The scientific review of mental health practice, 4, 2, 1-8.

Conway, M.A. (1997). Past and present: Recovered memories and false memories. In Conway, M.A., Recovered memories and false memories, 150-191. Oxford: UK

Cordon, I.M., Pipe, M-E., Sayfan, L., Melinder, A. & Goodman, G.S. (2004). Memory for traumatic experiences in early childhood. Developmental review, 24, 101-132. 

De Prince, A. & Freyd, J.J. (2004). Forgetting trauma stimuli. Psychological science, 15, 488-492.

Epstein, M.A. & Bottoms, B.L. (2002). Explaining the forgetting and recovery of abuse and trauma memories: Possible mechanisms. Child maltreatment, 7, 210-225.

Geraerts, E. & McNally, R.J. (2008). Forgetting unwanted memories: Directed forgetting and thought suppression methods. Acta psychologica, 127, 614-622. (author’s personal copy)

Geraerts, E., Schooler, J.W., Merckelbach, H., Jelicic, M.,Hauer, B.J.A., & Ambadar, Z. (2007). The Reality of Recovered Memories: Corroborating Continuous and Discontinuous Memories of Childhood Sexual Abuse. Psychological Science – in press.

Goodman, G.S., Ghetti, S., Quas, J.A., Edelstein, R.S., Alexander, K.W., Redlich, A.D., Cordon, I. M. & Jones, D.P.H. (2003). A prospective study of memory for child sexual abuse: New findings relevant to the repressed-memory controversy. Psychological             science, 14, 113-118.

Goodman, G.S. & Melinder, A. (2007). The development of autobiographical memory: a new model. In Magnussen, S. & Helstrup, T. (Eds.). Everyday memory. 111-134. East Sussex: Psychology Press.

Laney, C. & Loftus, E. (2005). Traumatic memories are not necessarily accurate memories. Canadian Journal of Psychiatry, 50, 823-828.

Loftus, E. (1994). Repressed memory accusations: Devastated families and devastated patients. Applied cognitive psychology, 11, 25-30.

Loftus, E. & Ketcham, K. (1994). The myth of repressed memory. New York: St. Martin’s Griffin.

Loftus, Pollonski & Fullilove (1994). Memories of childhood sexual abuse. Remembering and repressing. Psychology of women quarterly, 18, 67-84.

Magnussen, S. (2004). Vitnepsykologi. Pålitelighet og troverdighet i dagligliv og rettssal. Oslo: Abstrakt forlag.

Magnussen, S., Endestad, T., Koriat, A. & Helstrup, T. (2007). What do people believe about memory and how do they talk about memory? In Magnussen, S. & Helstrup, T. (Eds.). Everyday memory. 5-25. East Sussex: Psychology Press.

McNally, R. (2003). Remembering trauma. U.S.A.: Belknap Harvard.

McNally, R.J. (2004). Is traumatic amnesia nothing more but psychiatric folklore? Cognitive behaviour therapy, 33, 97-101.

Payne, B.K. & Corrigan, E. (2007). Emotional constraint on intentional forgetting. Journal of experimental psychology, 43, 780-786.

Passer, M.W. & Smith, R.E. (2007). Psychology. The science of mind and behaviour. 3rd ed. New York: McGraw-Hill.

Peace, K.A. & Porter, S. (2004). A longitudinal investigation of the reliability of memories for trauma and other emotional experiences. Applied cognitive psychology, 18, 1143-1159.

Porter, S. & Birt, A.R. (2001). Is traumatic memory special? A comparison of traumatic memory characteristics with memory for other emotional life experiences. Applied cognitive psychology, 15, 101-117.

Porter, S. & Peace, K. A. (2007). The scars of memory. A prospective, longitudinal investigation of the consistency of traumatic and positive emotional memories in adulthood. Psychological science, 18, 435-441. (Uncorrected proof version)

Rofé, Y. (2008). Does repression exist? Memory, pathogenic, unconscious and clinicalevidence. Review of general psychology, 12, 63-85.

Rosen, G.M. & Taylor, S. (2007). Pseudo-PTSD. Journal of anxiety disorders, 21, 201-210.

Schacter, D. L. (2001). The seven sins of memory: How the Mind Forgets and Remembers. New York: Houghton Mifflin.

Shobe, K. K. & Kihlstrom, J.F. (1997). Is traumatic memory special? Current directions in psychological science, 6, 70-74.

Sivers, H., Schooler, J. & Freyd, J.J. (2002). Recovered memories. Encylopedia of the human brain, 4, 169-184.

Williams, L.M. (1994). Recall of childhood trauma: A prospective study of women’s memories of child sexual abuse. Journal of consulting and clinical psychology, 62, 1167-1176.

Zoellner, L.A., Sacks, M.B & Foa, E.B. (2003). Directed forgetting following mood induction in chronic posttraumatic stress disorder patients. Journal of abnormal psychology, 112, 508-514.

 Web: 16.11.2008  20.11.2008 21.11.2008



1 It is unfortunately, due to the highly polarized public debate, necessary to stress that ordinary memory theorists obviously do not claim that all accounts of sexual abuse are false.