Remembrance of things past : the cognitive psychology of remembering and forgetting trauma

Several authors have argued that traumatic experiences are processed and remembered in a qualitatively different way from neutral events. To investigate this issue, we interviewed 121 Croatian war veterans diagnosed with posttraumatic stress disorder (PTSD) about amnesia, intrusions (i.e., flashbacks and nightmares), and the sensory qualities of their most horrific war memories. Additionally, they completed a self-report scale measuring dissociative experiences. In contrast to what one would expect on the basis of theories emphasizing the special status of traumatic memories, amnesia and high frequency intrusions were not particularly typical for our sample of traumatized individuals. Moreover, traumatic memories were not qualitatively different from neutral memories with respect to their stability and sensory qualities. The severity of PTSD symptoms was not significantly correlated with dissociative experiences. Our findings do not support the existence of special memory mechanisms that are unique to experiencing traumatic events. I N T R O D U C T I O N In clinical literature, there is little consensus about how victims remember traumatic experiences (e.g., Kihlstrom, 2004; Loftus, 1997b; McNally, 2003a). According to one popular view, memories of traumatic events have special properties that distinguish them from ordinary memories. Thus, many authors have argued that traumatic memories are qualitatively different from other types of memories, involving unique mechanisms not related to general memory functioning (e.g., van der Kolk, 1996, 1997; for a review, see Zoellner & Bittenger, 2004). For example, these authors assume that traumatic events are processed in a predominantly perceptual fashion, leaving survivors with traumatic memories that are not readily accessible (i.e., dissociative or psychogenic amnesia) and/or flashbacks that possess strong sensory qualities. This dissociative style of processing would also create a substantial overlap between dissociative and posttraumatic stress disorder (PTSD) symptoms. There are several versions of this theoretical stance (Brewin, Dalgleish & Joseph, 1996; Ehlers & Clark, 2000), but the core assumption they have in common is that trauma has a special impact on the way in which memories of the traumatic event are organized. For that reason, this position has been dubbed the traumaTRAUMATIC WAR MEMORIES 37 memory argument (for a balanced discussion, see Kihlstrom, 1996, in press; Shobe & Kihlstrom, 1997; for a reply, see Nadel & Jacobs, 1998). Although the trauma-memory argument has gained popularity among many clinicians, some findings argue against this view. As Pendergrast (1998) has pointed out, classic monographs like those of Spiegel and colleagues (Grinker & Spiegel, 1943/1945; Kardiner & Spiegel, 1941/1947) include only isolated case reports of soldiers suffering from amnesia (see also Lilienfeld & Loftus, 1998). Moreover, more systematic studies suggest that only a small minority of war victims report dissociative amnesia. For example, Kuch and Cox (1992) studied 124 holocaust survivors and found that dissociative amnesia, with an estimated lifetime prevalence rate of 3.2%, was quite rare in this group. Likewise, Merckelbach and colleagues (Merckelbach, Dekkers, Wessel, & Roefs, 2003a, 2003b) found in a group of 29 Dutch concentration camp survivors only one survivor reporting mnemonic experiences that might be taken as evidence for dissociative amnesia1. Goodman and co-workers found that 15.5% of victims of documented childhood sexual abuse (CSA) failed to report the target incident during a telephone interview conducted approximately 13 years after the pertinent events (Goodman, Ghetti, Quas, Edelstein, Alexander et al., 2003; Goodman & Paz-Alonso, in press). However, nondisclosure dropped to 8.3% after follow-up by a mailed questionnaire and a telephone interview2. Both rates of reporting failure are clearly lower than those found in earlier, less meticulous studies of CSA victims (e.g., Williams, 1994). Even when one accepts these lower rates of non-reporting, the question arises to what extent they do reflect genuine dissociative amnesia. People might not report the traumatic events for other reasons than amnesia (e.g., shame, ordinary forgetting; see McNally, 2003a). Recent literature is also critical about the idea that traumatic memories invariantly produce flashbacks. For example, using patient archives, Jones and colleagues showed that flashbacks were an uncommon symptom in generations of British war veterans before the 1980’s (Jones, Vermaas, McCartney, Beech, Palmer et al., 2003). So far, empirical support for the trauma-memory argument is not very compelling. On the other hand, the studies cited above suffer from various shortcomings and thus, one could argue that they provide only suboptimal tests of the trauma-memory argument. One shortcoming is that it is not always clear whether the events (e.g., war experiences) were really traumatic to the participants (Jones et al., 2003). Another limitation is that the traumatic events