Symptoms of Trauma and Traumatic Memory Retrieval in Adult Survivors of Childhood Sexual Abuse

Abstract

We examined posttraumatic stress symptoms and the memory retrieval process in 2 groups of adult survivors of childhood sexual abuse: 29 participants who reported having memories of their abuse prior to entering therapy (PM) and 13 who reported no memories of abuse prior to therapy (NPM). Participants were asked to indicate on checklists symptoms of constriction, hyperarousal, and intrusion experienced (a) prior to entering therapy and (b) during the surfacing of a memory while in therapy. Overall, the findings indicate that for both groups the same cluster of posttraumatic stress symptoms occurred prior to therapy and during therapy and that there was a continuity of symptoms over time. Participants were also asked to fully describe details of their traumatic memories as these details emerged prior to and/or during therapy. We determined that (a) there were striking similarities in the detailed recall of trauma memories for both groups; (b) memories of abuse emerged in substantial perceptual, somatic, and emotional detail over time before developing into a narrative; (c) the amount of detail remembered increased in the PM group during therapy; (d) members of the NPM group were more kinesthetic than visual in their orientation to the world and may not have had access to the visual information that would associate their symptoms to their abuse; and (e) triggers of traumatic memories were largely the result of internal rather than external stimuli, and these triggers happened primarily outside of therapy sessions.

KEYWORDS:

• posttraumatic stress disorder
• trauma symptoms
• adult survivors of childhood sexual abuse
• traumatic memory retrieval

INTRODUCTION

To study the consequence of childhood sexual abuse (CSA) and the process of retrieving traumatic memories, we developed a survey, “Symptoms of Trauma and the Memory Retrieval Process,” designed for adult survivors in therapy. Given that retrieved traumatic memory has generated controversy in the scientific community, we chose to compare results between two groups: individuals who reported having memories of sexual abuse prior to entering therapy (PM) and those who reported having no memories of abuse prior to therapy (NPM). This article describes our findings.

The effects of CSA in adult survivors are reflected in symptoms of posttraumatic stress disorder (PTSD). The Diagnostic and Statistical Manual of Mental Disorders (4th ed. [DSM–IV]; American Psychiatric Association, 1994) outlines the criteria for PTSD as follows: (a) a traumatizing event, usually involving bodily injury or threat to life; (b) intrusive reexperiencing of symptoms; (c) generalized numbing of responsiveness; and (d) physiological reactivity. Herman (1992) identified three types of complex PTSD symptoms: hyperarousal (the persistent expectation of danger), constriction (the numbing response of surrender), and intrusion (the indelible imprint of the traumatic moment). She detailed numerous ways in which each type of symptom can manifest in cases resulting from continuous and repeated abuse. Explaining that after a traumatic incident, the physiological arousal system of self-preservation goes into permanent alert, Herman cited Kardiner, who used the term physioneurosis to describe the psychosomatic complaints that, like other hyperarousal behaviors, result from chronic arousal of the autonomic nervous system. When people find themselves completely helpless, the self-preservation system shuts down entirely—constricts. Escape is experienced by a change in consciousness, the most severe aspect of which is dissociation, considered to be a reliable predictor of chronic PTSD (D. Brown, Scheflin, & Hammond, 1998). Intrusion is experienced when constriction fails and aspects of the traumatic memory leak into consciousness.

Unlike the DSM–IV, which focuses on single-incident trauma, Herman (1992) differentiated two types of PTSD: simple (resulting from single-incident trauma) and complex (resulting from continuous and repeated abuse, such as often occurs in childhood). With complex PTSD, symptoms can be chronic and persistent and can involve severe memory disturbance, both hypermnesic and amnesic (Horowitz & Reidbord, 1992). D. Brown et al. (1998) outlined research documenting the existence of functional amnesia in survivors of CSA, including studies by Cameron (1996), Draijer (1990), Ensink (1992), Herman and Schatzow (1987), and Roe and Schwartz (1996). Terr (1991) divided childhood trauma into two basic types: Type I includes full, detailed memories, “omens,” and misperceptions; and Type II includes denial and numbing, self-hypnosis, dissociation, and rage. Freyd (1996) proposed a two-dimensional model of trauma in which terror results in the hyperarousal aspects of PTSD symptoms, and betrayal by a parent or trusted caregiver results in the amnesia aspect of constriction symptoms.

Critics of the concept of traumatic amnesia have held therapists responsible for implanting memories of abuse in unsuspecting clients. However, Williams's (1994) prospective study on memory in survivors of CSA determined that even in cases when abuse had been documented in medical records, women had amnesia for their abuse. She concluded, “Having no memory for childhood sexual abuse is a common occurrence” (p. 1173). Freyd's investigation and reanalysis of Williams's and others' papers determined that amnesia rates were higher for survivors who had been abused by a family member (Sivers, Schooler, & Freyd, 2002). Fergusson, Horwood, and Woodward (2000) determined that the unreliability of the reporting of child abuse is due not to false reports by people who were not abused but to people who were abused often providing false-negative reports.

Although there has been considerable debate regarding amnesia for CSA, trauma experts agree that memory disturbance resulting from extreme trauma is best described as dissociation, a concept first used by Janet (1904). Putnam (1989) defined dissociation as “a normal process that is initially used defensively by an individual to handle traumatic experiences and evolves over time into a maladaptive or pathological process” (p. 9). The DSM–IV (American Psychiatric Association, 1994) defines dissociative amnesia as “episodes of inability to recall important personal information, usually of a traumatic or stressful nature, that is too extensive to be explained by ordinary forgetfulness” (p. 481). Phillips and Frederick (1995) conjectured that at the extreme end of the dissociation continuum traumatic memory is not simply forgotten, it does not even register in the conscious mind because it is stored in a different part of the brain and is assigned to different aspects of the personality—ego-states.

Current research has explored the different functions of dissociation (R. J. Brown, 2006; Sar & Ozturk, 2007) and provides new perspectives with implications for trauma treatment. We have discussed elsewhere how the natural ability to dissociate, in addition to being a coping strategy, can be used for healing in psychotherapy, for example when one imagines a safe place in the mind (Malmo & Laidlaw, 1996). Van der Hart and Nijenhuis (as cited in D. Brown et al., 1998) identified several types of secondary dissociation to account for partial amnesia for trauma and for the way in which traumatic memory can be divided into BASK components (behavior, affect, somatosensory, kinesthetic). Van der Hart (as cited in Van der Kolk, 1996) also described tertiary dissociation, which allows people to maintain a sense of self while separate states of mind process the traumatic event. Nijenhuis, Van der Hart, and Steele (2004) theorized that structural dissociation is the phenomenon at the basis of alter personalities in dissociative identity disorder. It is generally agreed that traumatic dissociation is an unconscious and automatic process that disconnects the conscious mind from a traumatic experience when a person's cognitions and feelings are overwhelmed, thereby protecting the individual by causing amnesia. Only when amnestic persons are safe enough or when a circumstance reminiscent of the trauma triggers the memory does dissociation begin to break down and memory begin to return (Grassian & Holtzen, 1996, as cited in D. Brown et al., 1998).

Sivers et al. (2002) defined recovered memory as “the recollection of a memory that is perceived to have been unavailable for some period of time” (p. 169). Rossi (1986) proposed that traumatic memory, when dissociated from consciousness, is state dependent (the origin of symptoms is embedded in emotional and somatic states), is centralized in the limbic–hypothalmic system (the link between mind and body), and therefore is often only retrievable when the person is in the same emotional state as he or she was when traumatized. Consistent with this theory, Van der Kolk (1996) argued that traumatic memory is organized differently from ordinary memory as implicit and perceptual rather than explicit and narrative. D. Brown et al. (1998) outlined numerous studies (Cameron, 1996; Davies & Frawley, 1994; Kristiansen, Felton, Hovdestad, & Allard, 1995; Roe & Schwartz, 1996; Van der Kolk & Fisler, 1995) that have reported the return of traumatic memory as sensory (flashbacks, somatic experiences, images, dreams, sudden and intense feelings, fragments, reenactments, avoidant behaviors) rather than narrative. According to Van der Kolk and Fisler, narrative memory emerges over time for most people only after the emergence and gradual integration of sensory memory.

Brewin, Dalgleish, and Joseph (1996) and Brewin (2001) proposed a dual representation theory of PTSD that involves two separate processing systems for trauma: (a) verbally accessible memory, which allows for deliberate and conscious recall of limited aspects of trauma; and (b) situationally accessible memory, which involves visual, sensory, physiological, and motor reenactment of trauma memory. Situationally accessible memory is accessed spontaneously or unconsciously via external or internal stimuli and involves far more detailed and extensive memories, with emotions being reexperienced at their original intensity. Not only do these two systems result in qualitatively different memories, but situationally accessible memories are more voracious and remain intact because of hormonal effects experienced with acute trauma that diminish neural activity in conscious processing while enhancing nonconscious perceptual and memory processing.

METHOD

Participants

In our previous research (Laidlaw & Malmo, 1995), 143 participants (ages 20–64) from across Canada and the Northwest Territories were identified by 82 therapists as being survivors of CSA. These participants gave multiple reasons for entering therapy: psychological issues—depression, anxiety, overwhelming sadness or fear, feeling something was wrong or missing, severe dissociation, suicidal thoughts or urges, rage, low self-esteem; awareness of having been sexually abused; relationship issues—marriage problems, abusive relationship, parenting difficulties; crisis issues—feeling unable to cope, feeling out of control, problems with school or career, previously failed treatment, suicide attempt, physical illness, sexual abuse of own child; seeking support for healing; concerns about possible sexual abuse due to flashbacks, obsessive thinking, awareness of physical abuse; recommendations to get help; addiction issues. We contacted the 78 participants who had agreed to follow-up contact and invited them to take part in this study. Of the 47 who responded, 5 were eliminated because of incomplete surveys, leaving a total of 42 respondents (41 women, 1 man). Names and identities of participants were coded to ensure anonymity. All participants gave informed consent. The project was vetted and approved by the Human Ethics Committee, Faculty of Graduate Studies, Dalhousie University.

Because traumatic memories retrieved in therapy have been controversial, we compared responses of participants who reported having had conscious memories of CSA prior to entering therapy (PM; n = 29) with those of participants who reported having no memories of CSA prior to therapy (NPM; n = 13).

Survey

The survey consisted of four parts: (a) updated demographic background and therapeutic experiences of participants; (b) awareness of experiences of CSA and perpetrators prior to entering therapy; (c) symptoms of trauma experienced prior to therapy and during the course of memory retrieval in therapy, details of how traumatic memories emerged during the course of therapy, and triggers of memories; and (d) the role of the therapist in memory retrieval. This article focuses on the results obtained from part (c).

Based on Herman's (1992) descriptions, we developed a checklist of posttraumatic stress symptoms within three categories: (a) constriction (including dissociation), (b) hyperarousal (including medical/physical symptoms of somatization), and (c) intrusion. Participants indicated which symptoms they experienced (a) prior to entering therapy and (b) during the surfacing of a memory while in therapy. In response to a series of open-ended questions, participants described in detail the emergence of one traumatic memory recalled prior to entering therapy (if applicable) and one recalled during the course of therapy. They also described the trigger(s) of their memory and indicated where the trigger(s) occurred—in a therapy session, outside of therapy, or both.

Analysis

The quantitative component of the study focused on the symptoms experienced by participants prior to and during therapy. Because numbers were small, no significant differences were found using standard methods of statistical analysis. Therefore, we used only percentages to compare symptoms of constriction, hyperarousal, and intrusion.

The qualitative component provided for an in-depth examination of the emergence of participants' traumatic memories. We categorized responses according to the kind of sensory material that was reported:

1. Context—aspects of the environment in which the abuse occurred.

2. Images—visual pictures in the mind.

3. Sensations—all sensory information other than images or sounds.

4. Feelings—emotional state of the survivor and perpetrator.

5. Behaviors—actions, words, or sounds of the survivor and perpetrator.

6. Cognitions—recollections of thoughts, meanings, intentions, intuitions.

7. Narratives—descriptions of what happened over time, including events preceding the abusive incident, the abusive experience itself, and events following the abusive incident.

We also analyzed triggers that precipitated the emergence of a memory, categorizing them as internal (originating inside the participant) or external (originating outside in the environment) and as simple (single) or complex (multiple). We compared the results between groups for details of traumatic memory and triggers of them, and we compared the memories of the PM group prior to and during therapy.

RESULTS

Demographic Information

Certain differences were noted between the NPM group (n = 13) and PM group (n = 29). Specifically, 54% of the NPM group had received a graduate degree compared to 21% of the PM group, 62% of the NPM group had full-time employment outside the home compared to 28% of the PM group, and 8% of the NPM group made less than $20,000 per year compared to 41% of the PM group.

Complex Posttraumatic Stress Symptoms

Constriction

Large numbers of participants in both groups indicated that they experienced constriction symptoms prior to as well as after entering therapy (see Figure 1). The most obvious difference between the groups was that all members of the NPM group reported total amnesia before therapy, whereas almost two thirds of the PM group reported partial amnesia for details related to sexual abuse and for associated feelings. In the NPM group, total amnesia completely disappeared during therapy; in the PM group, amnesia for memory details dropped dramatically (from 62% to 2%), but amnesia for feelings remained relatively high (41%). Approximately half of participants in both groups reported generalized amnesia for childhood prior to therapy that decreased during therapy. Experiences related to dissociation were consistently reported in greater numbers by the NPM group before therapy. An altered sense of time was reported in greater numbers by both groups during therapy compared to prior to therapy, and this increase was most dramatic for the NPM group. Although there was a decrease in the reporting of emotional constriction for both groups, the decrease was greater for the PM group.

FIGURE 1 Complex posttraumatic stress symptoms: constriction. PM, n = 29; NPM, n = 13. PM = participants who reported memories of abuse prior to entering therapy; NPM = participants who reported no memories of abuse prior to entering therapy.

Hyperarousal

Large numbers of participants in both groups reported many hyperarousal symptoms—including anxiety, hyperalertness, startle responses, extreme reactivity, medical/physical symptoms, sleep problems, panic attacks, phobias, and explosive anger—both before entering therapy and during the surfacing of traumatic memories in therapy (see Figure 2). Anxiety was consistently reported by a greater percentage of participants in the PM group than in the NPM group at both time periods—83% compared to 62% prior to therapy and 86% compared to 69% during therapy, respectively. Medical/physical symptoms included a variety of gastrointestinal problems (nausea, gagging, choking, stomach pains, rectal bleeding) as well as psychogenic, muscular, neurological, gynecological, respiratory, and cardiovascular problems. Changes in medical/physical symptoms during therapy were noteworthy: Symptoms decreased in the PM group (from 72% to 59%) but doubled in the NPM group (from 46% to 92%). Another noteworthy change was the reduced reporting of phobias by the PM group during therapy compared to prior to therapy.

FIGURE 2 Complex posttraumatic stress symptoms: hyperarousal. PM, n = 29; NPM, n = 13. PM = participants who reported memories of abuse prior to entering therapy; NPM = participants who reported no memories of abuse prior to entering therapy.

Intrusion

Large numbers of participants in both groups for both time periods indicated numerous symptoms of intrusion, including overwhelming feelings of sadness or pain, fear or terror, guilt or shame, and anger or rage; visual intrusions—flashbacks, dreams, nightmares; intrusive thoughts and words; kinesthetic intrusions—body sensations and intrusive smells, tastes, and sounds; and compulsive urges, compulsive reenacting, and risk-taking behaviors (see Figure 3). A notable difference was the decreased reporting of overwhelming sadness or pain in the NPM group after entering therapy (from 62% to 38%) and the increased reporting of overwhelming fear or terror (from 46% to 85%). The number of participants in the PM group who reported these symptoms remained constant at about 66%. A notable increase for both groups was the increased reporting during the surfacing of memories in therapy of intrusive images (NPM: from 23% to 69%; PM: from 45% to 76%) and intrusive body sensations (NPM: from 15% to 69%; PM: from 38% to 66%). A somewhat decreased reporting of compulsive urges and behaviors was seen for both groups during therapy.

FIGURE 3 Complex posttraumatic stress symptoms: intrusion. PM, n = 29; NPM, n = 13. PM = participants who reported memories of abuse prior to entering therapy; NPM = participants who reported no memories of abuse prior to entering therapy.

Memory Content

Comparisons of memory detail (context, images, sensations, feelings, behaviors, cognitions, narratives) revealed that a great amount of detail was recalled by the PM group before entering therapy and by both groups during therapy. A comparison of the PM group's memories before and during therapy revealed an overall increase in details remembered during therapy. A comparison between the PM and NPM groups' memories during therapy revealed a similar pattern of reporting numerous details for memory content in all categories, with little variation.

Context

The place or environment in which the abuse occurred was the aspect of context most often recalled (by more than four fifths of both groups during therapy), followed by the time of day or year, clothing of self and of the perpetrator, awareness of others, voices of others, and sounds from the environment (see Figure 4).

FIGURE 4 Details of memory content: context. PM, n = 29; NPM, n = 13. PM = participants who reported memories of abuse prior to entering therapy; NPM = participants who reported no memories of abuse prior to entering therapy.

Images

The majority of participants reported seeing images of themselves, followed by images of the perpetrator, body parts, the perpetrator's facial expression, themselves naked, the perpetrator naked, and symbolic images (such as a snake representing a penis; see Figure 5).

FIGURE 5 Details of memory content: images. PM, n = 29; NPM, n = 13. PM = participants who reported memories of abuse prior to entering therapy; NPM = participants who reported no memories of abuse prior to entering therapy.

Sensations

Relational position (such as whether the perpetrator was approaching from the left or right side or was experienced as being above or beside) was consistently reported by about three fourths of participants across groups. However, there was a striking difference in the reporting of kinesthetic memory. For the PM group, reporting of kinesthetic memory doubled from 38% prior to therapy to 76% during therapy, whereas 100% of the NPM group reported kinesthetic memory. The reporting of sense of size of self and smell followed a similar pattern of increased reporting. Only a small number of the PM group reported remembering taste and only during therapy (see Figure 6).

FIGURE 6 Details of memory content: sensations. PM, n = 29; NPM, n = 13. PM = participants who reported memories of abuse prior to entering therapy; NPM = participants who reported no memories of abuse prior to entering therapy.

Feelings

There was an almost identical reporting of feelings by both groups at both time periods: More than four fifths of participants reported their own feelings (“fear,” “terror,” “pain,” “confusion,” “guilt,” “shame,” “sadness,” etc.), and one fifth reported their perpetrator's feelings (“anger,” “hate,” “friendly,” “loving”).

Behaviors

The behavior of the perpetrator was consistently reported by more than three fourths of participants in all cases, and the voice of the perpetrator was consistently reported by more than half in all cases. The most notable differences were in the PM group's increased reporting of their own behavior, dissociation, resistance, and vocalizations during therapy compared to prior to therapy and the NPM group's even greater reporting of their own resistance and verbalizations/vocalizations (see Figure 7).

FIGURE 7 Details of memory content: behaviors. PM, n = 29; NPM, n = 13. PM = participants who reported memories of abuse prior to entering therapy; NPM = participants who reported no memories of abuse prior to entering therapy.

Cognitions

More than four fifths of the NPM group and the PM group before therapy reported the identity of the perpetrator. There was a slight drop in awareness of the abuser's identity for the PM group after entering therapy, reflecting the awareness of additional perpetrators. Their own age at the time of the abusive incident was reported by two thirds or more of the participants in all cases. Other aspects of cognitions reported in similar numbers were beliefs, intentions, intuitions, an awareness of power difference, and threats and intimidation by perpetrators (see Figure 8).

FIGURE 8 Details of memory content: cognitions. PM, n = 29; NPM, n = 13. PM = participants who reported memories of abuse prior to entering therapy; NPM = participants who reported no memories of abuse prior to entering therapy.

Narratives

More than half of participants recalled the beginning of the narrative in all cases. The middle of the story—the abusive incident itself—was recalled by 79% of the PM group before therapy, which increased to 88% during therapy, compared to 92% of the NPM group. Although only 28% of the PM group described the end of the incident in the memory recalled prior to therapy, their reporting increased to 64% during therapy compared to a similar reporting of 62% by the NPM group.

Triggers

Triggers of memories of abuse were categorized as internal (kinesthetic, images, repetitive thoughts, dreams, nightmares, spontaneous) or external (context, talking about abuse, hearing about abuse in media, seeing the perpetrator, being touched, being raped) and as single or multiple. Although some categories are self-explanatory, others require explanation. A kinesthetic trigger refers to body sensations or experiences such as bleeding, bruising, legs twitching, throat filled with phlegm, chest constricted, feeling cold, and other similar experiences. A context trigger refers to environmental factors, such as time of day or year, quality of the light, details of a particular room, or an object or an animal related to the abuse.

For the PM group prior to therapy, triggers were largely external (talking about abuse, context, perpetrator, media), and during therapy they were largely internal (kinesthetic, dreams, nightmares, and relaxation with image or kinesthetic sensations). Single and multiple triggers were reported about equally by the PM group, and multiple triggers most often involved context. In the NPM group, the majority of participants reported single internal triggers (kinesthetic), and when multiple triggers were reported they included kinesthetic sensations.

Finally, for the time period during therapy, 85% of the PM group indicated that their memory was triggered outside of therapy, 7% indicated that triggers occurred during therapy, and another 7% indicated that triggers occurred both outside of and during therapy. A similar finding for the NPM group was even more dramatic: 92% indicated that their memory trigger was located outside of therapy.

DISCUSSION

Posttraumatic Stress Symptoms

This article has presented the results of a study that examined symptoms of complex posttraumatic stress. Participants who reported memories of abuse prior to entering therapy (PM group) and those who reported no conscious memory of abuse prior to entering therapy (NPM group) experienced many similar complex posttraumatic stress symptoms of three types—constriction, hyperarousal, and intrusion—before entering therapy. These numerous similarities indicate that regardless of whether survivors of CSA had conscious memories of having been abused prior to entering therapy (i.e., whether they were amnestic for the abuse incidents), they have experienced many of the same kinds of symptoms.

After entering therapy, both the PM and NPM groups continued to experience the same symptoms during the surfacing of traumatic memory, with anxiety remaining high for both groups. This suggests that particular symptoms that may have brought people into therapy may continue to occur during the therapeutic process while trauma memories are entering consciousness and being worked through. The decreased reporting of amnesia and other constriction symptoms is consistent with the increased reporting by both groups of intrusive symptoms. This change is to be expected, given that a goal of therapy is to reconnect people to their dissociated memories. These results are consistent with the findings of Cameron (1996) that many symptoms of trauma increase with memory recovery.

The reporting of total amnesia for CSA prior to therapy by the NPM group and partial amnesia by the PM group supports findings (Draijer, 1990; Ensink, 1992; Harvey & Herman, 1994; Herman & Schatzow, 1987) that amnesia for abuse is not an either/or phenomenon and can involve partial amnesia. That amnesia for details of the abuse dropped dramatically but amnesia for feelings remained high in the PM group suggests that connecting to feelings associated to CSA is more difficult than knowing cognitively that abuse has occurred.

Being aware of their sexual abuse and being generally more symptomatic may have interfered with achievement in the PM group, whose members reported a lower level of education and income than those of the NPM group. This result supports the findings of Grassian and Holtzen (1996; as cited in D. Brown et al., 1998), who determined that dissociation could provide for a higher level of functioning in those survivors of abuse who coped by internalizing their trauma compared to those who externalized, remembered, and consistently experienced problems, including conduct and addictive behaviors as well as lower grades in school. About those people who successfully dissociate their memories from consciousness for a period of time, Grassian and Holtzen concluded, “We are able to view both the adaptive function of the failure to remember—the avoidance of overwhelming, disorganizing affect—and also the heavy price the individual pays for such a defensive adaptation—emotional constriction, numbing, and compulsivity” (as cited in D. Brown et al., 1998, p. 174).

Once in therapy, the NPM group reported more changes in symptoms than the PM group—a decrease in emotional numbing, a sense of something missing, and generalized amnesia for childhood—which suggests that in therapy they were reconnecting with their past experiences. There was also a decreased reporting of dangerous, risk-taking behavior, suggesting that during therapy they were learning to take care of themselves and to express feelings and needs appropriately. At the same time, certain negative symptoms increased dramatically for this group, including medical/physical symptoms, overwhelming fear and terror, intrusive images and flashbacks, and intrusive body sensations. These findings suggest that for those trauma survivors who are most dissociative, particular symptoms may increase as the trauma is reassociated. This result supports the findings of Grassian and Holtzen (1996; as cited in D. Brown et al., 1998) that for severely amnestic people, remembering trauma can precipitate an emotional crisis that results in those people becoming more symptomatic.

Almost three quarters of the PM group reported medical/physical symptoms prior to entering therapy, and this number decreased a little after groups entered therapy, whereas in the NPM group the numbers doubled from 46% to 92%. The high reporting of medical/physical symptoms highlights the prevalence of somatization, which has long been understood to be an indicator of posttraumatic stress. Van der Kolk, McFarlane, and Weisaeth (1996) identified numerous studies that have linked somatization to childhood sexual or physical abuse, dissociation, and PTSD. They concluded,

Traumatized patients come to experience emotional reactions merely as somatic states, without being able to interpret the meaning of what they are feeling. Unable to “know” what they feel, they become prone to undifferentiated affect storms and psychosomatic reactions which are devoid of personal meaning and cannot lead to adaptive responses. (p. 60)

It would appear that the group with no conscious memory of having been abused prior to therapy became overwhelmed by the realization that they had been abused (and, according to Freyd, 1996, that they had been betrayed) and that their distress was converted into physical pain and medical problems.

Traumatic Memory Retrieval

The most striking aspect of the participants' descriptions of their abuse memories was the amount and quality of detail that was remembered. Participants in both groups were able to recall a great deal of detail in every category, with the amount of detail increasing for the PM group once participants were in therapy. The vast majority of memory detail described by participants in all cases was perceptual. These results support the position of Van der Kolk and Fisler (1995) that trauma is initially remembered as somatosensory information and emotional flashbacks—that traumatic experiences are remembered intrinsically in the body. The results are also consistent with the view that traumatic memory is organized differently from ordinary memory as implicit and perceptual rather than explicit and narrative (Van der Kolk, 1996).

Somatosensory memory is primal and is typical of infants and young children who do not yet have the cognitive structures with which to contextualize their experiences. That the majority of participants in all cases recalled their relational position with respect to the perpetrator is not surprising, then, because differentiating the self in relation to others is one of the earliest developmental tasks of an infant. Our results support Herman's (1992) supposition that “in their predominance of imagery and bodily sensation, and in their absence of verbal narrative, traumatic memories resemble the memories of young children” (p. 38).

That traumatic memory is different from ordinary memory is substantiated by research that used mental imaging to demonstrate that different parts of the brain fire during the recalling of autobiographical narratives than fire during the experiencing of traumatic flashbacks (Ogden, Minton, & Pain, 2006; Van der Kolk & Fisler, 1995; Van der Kolk et al., 1996). ¹ This is significant because research that claims that false memories can be suggested has demonstrated only that narrative memory, not traumatic memory, can be manipulated (D. Brown et al., 1998).

It has already been established that adults who have experienced more severe trauma and who possess a greater ability to dissociate have less conscious awareness of having been abused in childhood (Brown et al., 1998). Chu, Frey, Ganzel, and Mathews (1999) found that being abused at an early age and being abused frequently was associated with higher levels of dissociation. Freyd determined that the element of betrayal is another factor affecting level of amnesia (Sivers et al., 2002). Our research may have uncovered another factor that can assist in understanding why some people are aware of having been sexually abused in their childhoods whereas others remain largely dissociated from their traumas. The 100% reporting of kinesthetic memory detail by the NPM group suggests that people with no conscious memories of having been abused prior to entering therapy (but with numerous other trauma symptoms similar to those experienced by people with conscious memories) may be more kinesthetic in their orientation to the world. This group's dramatic increased reporting (92%) of medical/physical symptoms during therapy as dissociation was reversed seems to be congruent with this interpretation—traumatic memories were experienced physically, and unresolved trauma manifested somatically. Conversely, it may be that participants with conscious awareness of their abuse prior to therapy are more visual. Perhaps the PM group was aware of having been abused because they had images of their trauma and thus “knew” what had happened to them. During memory retrieval in therapy, most participants in both groups reported the perpetrator's facial expression, body parts, voice, behaviors, and identity. It is this crucial information that the NPM group members did not have prior to entering therapy in order to conclude that they had been abused.

Participants generally explained that their memory detail emerged as symptoms and fragments over long periods of time, from days to months (and for initial memories even years), and only became a narrative at the end of the remembering process when fragments merged. There was a dramatic increase in reporting of the end of the narrative in memories described during therapy compared to prior to therapy for the PM group. This suggests that safety and support in the therapeutic relationship enables people to stay with or tolerate the recollection of a traumatic memory to its completion, that is, through the process of gathering together fragments of emotional and sensory material until a complete narrative emerges into consciousness.

The differing nature of memory triggers for the groups is interesting. For the PM group, the movement from primarily multiple to single triggers suggests that once in therapy these participants were working more quickly with their initial memory indicators. The movement from primarily external to internal triggers in this group suggests that participants were in the process of learning to work with the emerging material from their unconscious minds and bodies rather than simply reacting to events from the environment, as they had done prior to therapy.

That the members of the NPM group primarily indicated single triggers suggests that they may be less likely to be flooded with intrusive symptoms, or that they were continuing to dissociate, or that their greater kinesthetic awareness allowed for keener focus. Their reporting of primarily internal triggers (arising from their own somatic and affective experience) is consistent with their 100% reporting of kinesthetic memory during traumatic memory retrieval in therapy and their 92% reporting of somatic symptoms during therapy.

Finally, the vast majority of participants in both groups reported that triggers of their memory occurred outside of therapy, challenging those who claim that retrieval of traumatic memory is the result of suggestions by therapists. This finding supports Chu et al. (1999), whose participants generally reported recovering their memories at home, alone, or with family or friends.

CONCLUSION

Our findings indicate that the same cluster of posttraumatic stress symptoms occurs in adult survivors of CSA prior to therapy and during the course of therapy whether they report having awareness of their abuse prior to therapy or no awareness of abuse. In other words, psychological consequences of abuse are evidenced prior to as well as during therapy whether or not the survivors are aware of the origin of their symptoms. Although the numbers are too small to draw any definitive conclusions, the results also suggest that people with no conscious awareness of having been abused prior to therapy, despite exhibiting numerous severe PTSD symptoms, may be more kinesthetic than visual in their perceptual orientation and may be missing the visual memory fragments that would provide them with knowledge of having been abused.

For both those people who had memories of abuse prior to entering therapy and those who reported none, memories of abuse emerged in substantial perceptual, somatic, and emotional detail over time before developing into a narrative. Greater detail and a more complete narrative were reported by the PM group after entering therapy compared to before. Triggers for memories were more internal than external for both groups, demonstrating that retrieved memories of abuse emerge primarily from the mind–body of survivors of abuse and not from the environment. This is especially the case for individuals with no conscious memories of their abuse before therapy, who appear to be largely kinesthetic in their perceptual orientation. This result challenges the position of critics who argue that retrieved memories are necessarily suspect and arise from suggestions by the therapist.

Notes

1. Traumatic flashbacks involve emotional and sensory aspects of the traumatic experience, including visual images and auditory, olfactory, and kinesthetic sensations.