Psychology & Law 2000words essay

Research Report

The Reality of Recovered Memories Corroborating Continuous and Discontinuous Memories of Childhood Sexual Abuse Elke Geraerts,1,2 Jonathan W. Schooler,3 Harald Merckelbach,1 Marko Jelicic,1 Beatrijs J.A. Hauer,1 and

Zara Ambadar4

1 Maastricht University, Maastricht, The Netherlands;

2 Harvard University;

3 University of British Colombia, Vancouver,

British Columbia, Canada; and 4 University of Pittsburgh

ABSTRACT—Although controversy surrounds the relative

authenticity of discontinuous versus continuous memories

of childhood sexual abuse (CSA), little is known about

whether such memories differ in their likelihood of cor-

roborative evidence. Individuals reporting CSA memories

were interviewed, and two independent raters attempted

to find corroborative information for the allegations.

Continuous CSA memories and discontinuous memories

that were unexpectedly recalled outside therapy were

more likely to be corroborated than anticipated discon-

tinuous memories recovered in therapy. Evidence that

suggestion during therapy possibly mediates these differ-

ences comes from the additional finding that individuals

who recalled the memories outside therapy were markedly

more surprised at the existence of their memories than

were individuals who initially recalled the memories in

therapy. These results indicate that discontinuous CSA

memories spontaneously retrieved outside of therapy may

be accurate, while implicating expectations arising from

suggestions during therapy in producing false CSA mem-

ories.

There are individuals who say that they had completely forgotten

their traumatic life experiences, only to recover them years later.

How accurate are such discontinuous memories? Over the past

15 years, this issue has given rise to one of the most intense

controversies to afflict the fields of psychiatry and psychology

(e.g., Loftus, 1997; McNally, 2003; Schacter, 1995). This debate

has focused mainly on the validity of discontinuous memories of

childhood sexual abuse (CSA). Whereas some researchers have

forcefully argued that such reports should be taken at face value

(Brown, Scheflin, & Whitfield, 1999), others have countered

with equal vehemence that they are likely to be the product of

therapist suggestion (e.g., Kihlstrom, 2004; Loftus, 2003).

A relatively straightforward approach for assessing the au-

thenticity of discontinuous CSA memories is to examine whether

such memories can be corroborated independently, by other

individuals who were abused by the same perpetrator, learned of

the abuse soon after it occurred, or confessed to having partic-

ipated in the abuse themselves. Although such evidence does

not incontrovertibly demonstrate the validity of any single

memory report, the existence of independent corroborative ev-

idence certainly increases the likelihood that a memory report

corresponds to an actual event. Using this approach, Schooler

and his coworkers (e.g., Schooler, Ambadar, & Bendiksen, 1997;

Schooler, Bendiksen, & Ambadar, 1997; Shobe & Schooler,

2001) described corroborative evidence for several case studies

of individuals who experienced the ‘‘discovery’’ of apparently

long-forgotten memories of abuse. These case studies demon-

strated that it is possible to find corroborative information for the

memories of individuals who report suddenly remembering that

they had been victims of abuse. However, these findings do not

preclude the possibility that other discontinuous memories are

illusory. It is important to bear in mind that none of the dis-

continuous memories in these cases were recalled in the context

of aggressive recovered-memory therapy employing visualiza-

tion, suggestion, hypnosis, or other techniques in order to fa-

cilitate the recall of unknown ‘‘memories’’ of abuse. It has

frequently been suggested that such techniques are a major

P S C I 1 9 4 0 B Dispatch: 25.5.07 Journal: PSCI CE: BlackwellJournal Name Manuscript No. Author Received: No. of pages: 5 PE: TMS Op: gsravi

Address correspondence to Elke Geraerts, Department of Psychol- ogy, Harvard University, 1232 William James Hall, 33 Kirkland St., Cambridge, MA 02138, e-mail: [email protected].

P S Y C H O L O G I C A L S C I E N C E

564 Volume 18—Number 7Copyright r 2007 Association for Psychological Science

PSCI 1940(B W

U S

P S

C I

19 40

.P D

F 2

5- M

ay -0

7 19

:5 9

11 24

64 B

yt es

5 P

A G

E S

n o

pe ra

to r=

R av

is ha

nk ar

)

source of false CSA memories (Lindsay & Read, 1994, 1995;

Loftus, 2004; McNally, 2003).

Despite the huge controversy surrounding this issue, no

systematic large-scale research has examined the relative

likelihood of corroborative evidence for continuous versus dis-

continuous CSA memories. Accordingly, in the current study,

we attempted to determine whether continuous CSA memories

(i.e., abuse memories characterized as having been continuously

accessible to the individual) differ from discontinuous memories

that are alleged to have been forgotten and subsequently re-

called (for different approaches, see Andrews et al., 2000;

Dalenberg, 1996; Williams, 1995). If discontinuous memories

are inherently unreliable, then continuous memories should be

more likely to be corroborated than discontinuous memories,

regardless of whether the latter are recalled in or out of therapy.

If discontinuous memories are inherently reliable, then they

should not differ from continuous memories in their likelihood of

corroboration. Finally, if discontinuous memories can be the

product of therapist suggestions, then discontinuous memories

recalled in therapy should be less likely to be corroborated than

those recalled out of therapy.

We also examined whether other characteristics of memory

descriptions predict the relative likelihood that CSA memories

will be corroborated. A systematic comparison of the factors

associated with corroborated versus noncorroborated memories

can speak to mechanisms that could produce false memories.

For example, if deliberate efforts to find suspected CSA mem-

ories increase the likelihood of generating false memories, then

reported surprise at the discovery of a CSA memory may be

predictive of whether or not the memory will be corroborated

(Schooler, 2001).

METHOD

For this study, which is part of an ongoing research project on

discontinuous memories (Geraerts, Arnold, et al., 2006, Ge-

raerts, Smeets, Jelicic, Merckelbach, & van Heerden, 2006;

Geraerts, Smeets, Jelicic, van Heerden, & Merckelbach, 2005),

we used advertisements in local newspapers to recruit subjects

with discontinuous and continuous CSA memories. Subjects

were told that the research pertained to CSA and memory. CSA

was defined as physical sexual contact ranging from fondling to

penetrative acts until the age of 12. We screened subjects during

an initial telephone conversation so as to acquire similar sample

sizes for the two groups. After subjects provided written in-

formed consent, a semistructured memory interview was con-

ducted to verify their classification into one of the two groups.

The discontinuous-memory group consisted of 57 subjects

(mean age 5 41.5 years, SD 5 10.3; 45 women) who responded

affirmatively to the question, ‘‘Do you believe there was a time

when you were completely unaware that you had ever been a

victim of abuse, and that you later came to remember that you

were abused?’’ We note that just because these individuals said

they forgot the memory does not mean they necessarily did

(Geraerts, Arnold, et al., 2006, Schooler, Ambadar, & Bendik-

sen, 1997). The continuous-memory group comprised 71 sub-

jects (mean age 5 41.2 years, SD 5 11.5; 55 women) who

responded negatively to this question.

Subjects were asked to complete an extensive questionnaire

that assessed their memory for the abuse. Additionally, for

subjects reporting discontinuous memories, the questionnaire

covered the degree of prior memory loss, the context of the

memory recovery, and the use of specific techniques during

recovered-memory therapy.

After subjects had completed the questionnaire, they were

systematically queried regarding their knowledge of potential

corroborative evidence for the alleged abuse events. Subse-

quently, two raters, blind to the type of abuse memories of each

subject, interviewed other individuals who could potentially

provide corroborative evidence for the abuse. Memories were

characterized as corroborated if one or more of the following

three criteria were met: (a) another individual reported learning

about the abuse soon (i.e., within the next week) after it oc-

curred, (b) another individual reported having also been abused

by the alleged perpetrator, or (c) another individual reported

having committed the abuse him- or herself. For each subject,

these blind raters assessed whether or not the testimonies met

these criteria. Interrater agreement was .96; disagreements were

resolved by discussion.

RESULTS

To examine whether continuous CSA memories and discontin-

uous memories recalled outside therapy are more likely to be

corroborated than discontinuous memories reported to have

been recalled in therapy, we divided the discontinuous-memory

group into individuals reporting having recalled CSA memories

out of therapy (n 5 41) and individuals reporting having recalled

the abuse events in therapy (n 5 16). There was relatively little

difference in the frequency of corroborative evidence between

the continuous-memory group (45%) and the recalled-out-of-

therapy group (37%), w2(1, N 5 112) < 1. Moreover, as Table 1 shows, the type of corroboration did not differ between these two

groups, w2(1, N 5 112) 5 .195. However, there was a marked difference in the frequency of corroborative evidence for dis-

continuous memories reported to be recalled out of therapy

(37%) and discontinuous memories reported to be recalled in

therapy (0%), w2(1, N 5 57) 5 7.94, prep 5 .97. Indeed, no corroborative evidence could be found for any of the abuse

events reported by subjects in the recalled-in-therapy group.

We had hypothesized that discontinuous memories recalled in

therapy would be less likely to be corroborated than those re-

called out of therapy because the therapy context often involves

an explicit effort to unearth forgotten memories (Polusny &

Follette, 1996) and thereby raises the opportunity for sugges-

tion. To test this hypothesis, we examined how surprised indi-

Volume 18—Number 7 565

E. Geraerts et al.

PSCI 1940(B W

U S

P S

C I

19 40

.P D

F 2

5- M

ay -0

7 19

:5 9

11 24

64 B

yt es

5 P

A G

E S

n o

pe ra

to r=

R av

is ha

nk ar

)

viduals indicated they were when they recalled the memory of

abuse. A logistic regression analysis restricted to the discon-

tinuous-memory group revealed that subjects’ reported surprise

at the existence of the memory significantly predicted the

probability of corroborative evidence, b 5 .84, SE 5 .39, prep 5 .91. Furthermore subjects whose memories were recalled out-

side therapy were markedly more surprised at the existence of

their memories than were individuals whose memories were

recalled in therapy, t(55) 5 7.41, prep >.99. This finding is what

would be expected if expectations of recovering long-lost

memories are shaped during therapy and lead clients to recall

unfounded memories. Further support for the claim that thera-

pists’ suggestions contributed to the disparity between the cor-

roboration rates of memories recalled in therapy and memories

recalled out of therapy comes from the finding that the recalled-

in-therapy and recalled-out-of-therapy groups did not differ

with respect to a host of other variables that could in principle

contribute to differences in corroboration rate (including re-

ported age of abuse, b 5 .13, SE 5 .08, prep 5 .81; relationship to perpetrator, b 5 .02, SE 5 .22, prep 5 .15; and severity of abuse, b 5 .32, SE 5 .52, prep 5 .47).

DISCUSSION

This study made use of a corroborative approach to examine

discontinuous memories of abuse. Although the sources of cor-

roboration were initially based on our subjects’ self-reports, the

corroborative information itself reflects independent corrobo-

ration. Admittedly, this corroborative evidence did not reach the

‘‘beyond reasonable doubt’’ threshold necessary for criminal

convictions, as false memories could have been independently

corroborated with the criteria used in this study. For example, in

cases in which other individuals reported being abused by the

same perpetrator (i.e., our second criterion), the corroborating

source could have generated a false memory after learning of the

subject’s reported abuse. Notably, however, the proportion of

cases corroborated according to this criterion did not differ

between discontinuous memories recalled outside therapy and

continuous memories. This result is evidence against the pos-

sibility that the corroboration rate for discontinuous memories

was unduly biased by this source of evidence. Moreover, our

corroboration criteria are consistent with those used in other

rigorous investigations of CSA memories (McNally, Ristuccia, &

Perlman, 2005), led to very high agreement between indepen-

dent raters, and powerfully discriminated between discontinu-

ous memories that on a priori grounds were expected to be more

versus less likely to be factual. Altogether, it seems appropriate

to conclude that the present corroborative approach offers

genuine clues regarding the factors that tend to discriminate

authentic CSA memory reports from pseudomemories about

CSA.

We found that continuous CSA memories and memories re-

called unexpectedly out of therapy were comparable in their

likelihood of corroborative evidence. This finding indicates that

discontinuous memories are not, as has sometimes been sug-

gested, inherently unreliable. This idea is supported by recent

research showing that people reporting spontaneously recovered

memories show a striking tendency to forget prior incidences of

remembering when those prior retrievals have taken place in a

different retrieval context. This finding suggests that this group,

as a whole, may simply be failing to remember their prior

thoughts about a genuine incidence of CSA (Geraerts et al.,

2007).

At the same time, discontinuous memories that were recalled

in the context of therapy were significantly less likely to be

corroborated than were either continuous memories or discon-

tinuous memories recalled outside therapy. Indeed, of the 16

therapy-based discontinuous memories, not a single one could

be corroborated.

The difference in the corroboration rate for discontinuous

memories recalled in therapy versus out of therapy does not

logically require that memories recalled outside therapy were

necessarily more likely to be false. It is possible that some other

factor associated with individuals recalling memories in therapy

led to the failure to corroborate these memories. For example,

factors such as age at the time of the event, relationship to the

perpetrator, or severity of the abuse could in principle have

contributed to the likelihood of corroboration, independently of

whether or not the memories were factual. However, none of

these variables differed for discontinuous memories recalled in

versus out of therapy. In fact, the only factor besides corrobo-

ration that discriminated discontinuous memories recovered in

therapy from those remembered out of therapy was surprise at

the existence of the memory, with memories recalled in therapy

being associated with markedly less surprise. Indeed, degree of

surprise proved to be a powerful predictor of corroboration.

The general lack of surprise at the existence of uncorrobo-

rated memories recovered in therapy is consistent with the view

that expectations (that were generated by the therapist or that

led the subject to therapy in the first place) contributed to the

TABLE 1

Percentage of Continuous Memories and Discontinuous

Memories Recovered Out of Therapy That Were Corroborated by

Each of the Three Criteria

Type of corroborative information

Memory type

Individuals abused by the

same perpetrator

Individuals who learned of the abuse soon after it occurred

Perpetrator him- or herself

Continuous 53% (17) 31% (10) 16% (5)

Recovered out of

therapy 60% (9) 27% (4) 13% (2)

Note. Frequencies are given in parentheses.

566 Volume 18—Number 7

The Reality of Recovered Memories

PSCI 1940(B W

U S

P S

C I

19 40

.P D

F 2

5- M

ay -0

7 19

:5 9

11 24

64 B

yt es

5 P

A G

E S

n o

pe ra

to r=

R av

is ha

nk ar

)

generation of false memories. It is possible, however, that some

other factor was responsible for the tendency for individuals to

be unsurprised by the existence of memories recovered in

therapy. For example, the existence of abuse-related symptoms

might have increased the likelihood of seeking therapy and also

reduced the experience of surprise when a possible source of

those symptoms was discovered. Similarly, it is possible that

some undetected factor could have both increased the likelihood

that subjects sought therapy and decreased the likelihood that

their abuse experiences were corroborated. For example, indi-

viduals who failed to disclose their abuse might have been more

likely to seek therapy (because of their symptoms) and less

likely to have experienced abuse that could be corroborated

(because there were no other individuals to whom the abuse was

disclosed). Such an account could in principle be one reason for

the differential corroboration rate for memories recovered in and

out of therapy. However, the present data provide no direct

evidence for this hypothesis, as there was no relation between

whether or not a memory was recovered in therapy and the

frequency with which subjects reported talking about the abuse

with other people.

A limitation of our study deserves attention. Our findings

pertain to adults recruited through advertisements in local

newspapers. Corroborative evidence for abuse events might be

different in clinical samples. It is important for future research

to examine whether the patterns observed in this study gener-

alize to the many abuse survivors who would be unlikely to

actively seek out participation in research associated with their

experience.

Together, the current findings suggest that expectations re-

garding the possible existence of a forgotten memory of abuse

contributed to the difference in the corroboration rates for

memories recalled in versus out of therapy. Given the correla-

tional nature of the data, it is difficult to assess whether such

expectations were caused by the therapists, or whether indi-

viduals who suspected they possessed forgotten memories of

abuse were more likely than others to seek therapy. However,

given that therapists are known to use aggressive recovered-

memory techniques (for a review, see Andrews, 2001) and that

such forms of suggestion can produce false memories (e.g.,

Mazzoni, Loftus, Seitz, & Lynn, 1999), it seems likely that one

reason memories recalled in therapy were not corroborated is

that some of these memories were false memories generated from

an interaction between expectations induced in therapy and

intrinsic source-monitoring difficulties in the clients (Johnson,

Hashtroudi, & Lindsay, 1993). Consistent with this idea, a re-

cent study in our lab has shown that as a group, people who come

to believe they have recovered a memory of CSA through sug-

gestive therapy generally show a pronounced tendency to in-

correctly claim that they have experienced events when they

have demonstrably not experienced those events, as measured

by simple, well-documented cognitive tests of false-memory

formation (Geraerts et al., 2007).

In sum, this study provides support for both sides of the so-

called recovered-memory controversy. Evidence that discon-

tinuous memories can be genuine comes from the observation

that discontinuous memories recalled outside the context of

therapy were not significantly less likely to be corroborated than

were continuous memories. Evidence that discontinuous mem-

ories can be false comes from the finding that memories recalled

in the context of therapy were markedly less likely to be cor-

roborated than were continuous memories or discontinuous

memories recalled outside of therapy. The present results also

offer an important clue for discerning which discontinuous CSA

memories are likely to be factual. Seemingly forgotten memories

whose recall was associated with a sense of surprise were much

more frequently corroborated than discontinuous memories

whose existence was anticipated. This latter finding suggests

that whereas deliberately recovered memories are apt to be

suspect, spontaneously discovered memories (Schooler, 2001)

are more likely to be true.

Acknowledgments—This study was supported in part by a

grant from the Netherlands Organization for Scientific Research

(NWO Grant 446-06-002) awarded to Elke Geraerts. We would

like to thank Danny Kostons and Kim Hamers for their help

testing the subjects.

REFERENCES

Andrews, B. (2001). Recovered memories in therapy: Clinicians’

beliefs and practices. In G.M. Davies & T. Dalgleish (Eds.),

Recovered memories: Seeking the middle ground (pp. 95–151). Chichester, England: John Wiley & Sons.

Andrews, B., Brewin, C.R., Ochera, J., Morton, J., Bekerian, D.A.,

Davies, G.M., et al. (2000). The timing, triggers and qualities of

recovered memories in therapy. British Journal of Clinical Psy- chology, 39, 11–26.

Brown, D., Scheflin, A.W., & Whitfield, C.L. (1999). Recovered

memories: The current weight of the evidence in science and in

the courts. Journal of Psychiatry & Law, 27, 5–156. Dalenberg, C.J. (1996). Accuracy, timing and circumstances of dis-

closure in therapy of recovered and continuous memories of

abuse. Journal of Psychiatry & Law, 24, 229–275. Geraerts, E., Arnold, M.M., Lindsay, D.S., Merckelbach, H., Jelicic,

M., & Hauer, B. (2006). Forgetting of prior remembering in

people reporting recovered memories of childhood sexual abuse.

Psychological Science, 17, 1002–1008. Geraerts, E., Lindsay, D.S., Jelicic, M., Merckelbach, H., Raymaekers,

L., Arnold, M.M., & Schooler, J.S. (2007). Cognitive mechanisms underlying recovered memory experiences of childhood sexual abuse. Manuscript submitted for publication.

Geraerts, E., Smeets, E., Jelicic, M., Merckelbach, H., & van Heerden,

J. (2006). Retrieval inhibition of trauma-related words in women

reporting repressed or recovered memories of childhood sexual

abuse. Behaviour Research and Therapy, 44, 1129–1136. Geraerts, E., Smeets, E., Jelicic, M., van Heerden, J., & Merckelbach,

H. (2005). Fantasy proneness, but not self-reported trauma is

related to DRM performance of women reporting recovered

Volume 18—Number 7 567

E. Geraerts et al.

PSCI 1940(B W

U S

P S

C I

19 40

.P D

F 2

5- M

ay -0

7 19

:5 9

11 24

64 B

yt es

5 P

A G

E S

n o

pe ra

to r=

R av

is ha

nk ar

)

memories of childhood sexual abuse. Consciousness & Cognition, 14, 602–612.

Johnson, M.K., Hashtroudi, S., & Lindsay, D.S. (1993). Source moni-

toring. Psychological Bulletin, 114, 3–28. Kihlstrom, J.F. (2004). An unbalanced balancing act: Blocked, re-

covered, and false memories in the laboratory and clinic. Clinical Psychology: Science and Practice, 11, 34–41.

Lindsay, D.S., & Read, J.D. (1994). Psychotherapy and memories of

childhood sexual abuse: A cognitive perspective. Applied Cog- nitive Psychology, 8, 281–338.

Lindsay, D.S., & Read, J.D. (1995). ‘‘Memory work’’ and recovered

memories of childhood sexual abuse: Scientific evidence and

public, professional, and personal issues. Psychology, Public Policy, and Law, 1, 846–908.

Loftus, E.F. (1997). Dispatch from the (un)civil memory wars.

In J.D. Read & D.S. Lindsay (Eds.), Recollections of trauma: Scientific evidence and clinical practice (pp. 171–198). New York: Plenum.

Loftus, E.F. (2003). Make-believe memories. American Psychologist, 58, 867–873.

Loftus, E.F. (2004). Dispatch from the (un)civil memory wars. Lancet, 364, 20–21.

Mazzoni, G.A.L., Loftus, E.F., Seitz, A., & Lynn, S.J. (1999). Changing

beliefs and memories through dream interpretation. Applied Cognitive Psychology, 13, 125–144.

McNally, R.J. (2003). Remembering trauma. Cambridge, MA: Belknap Press/Harvard University Press.

McNally, R.J., Ristuccia, C.S., & Perlman, C.A. (2005). Forgetting

trauma cues in adults reporting continuous or recovered memo-

ries of childhood sexual abuse. Psychological Science, 16, 336– 340.

Polusny, M.A., & Follette, V.M. (1996). Remembering childhood

sexual abuse: A national survey of psychologists’ clinical prac-

tices, beliefs, and personal experiences. Professional Psychology: Research and Practice, 27, 41–52.

Schacter, D.L. (1995, April). Memory wars. Scientific American, 272, 135–139.

Schooler, J.W. (2001). Discovering memories in the light of meta-

awareness. Journal of Aggression, Maltreatment, and Trauma, 4, 105–136.

Schooler, J.W., Ambadar, Z., & Bendiksen, M.A. (1997). A cognitive

corroborative case study approach for investigating discovered

memories of sexual abuse. In J.D. Read & D.S. Lindsay (Eds.),

Recollections of trauma: Scientific evidence and clinical practice (pp. 379–388). New York: Plenum.

Schooler, J.W., Bendiksen, M.A., & Ambadar, Z. (1997). Taking the

middle line: Can we accommodate both fabricated and recovered

memories of sexual abuse? In M. Conway (Ed.), False and re- covered memories (pp. 251–292). Oxford, England: Oxford Uni- versity Press.

Shobe, K.K., & Schooler, J.W. (2001). Discovering fact and fiction:

Case-based analyses of authentic and fabricated memories of

abuse. In G.M. Davies & T. Dalgleish (Eds.), Recovered memories: Seeking the middle ground (pp. 95–151). Chichester, England: John Wiley & Sons.

Williams, L.M. (1995). Recovered memories of abuse in women with

documented child sexual victimization histories. Journal of Traumatic Stress, 8, 649–674.

(RECEIVED 9/4/06; REVISION ACCEPTED 12/14/06; FINAL MATERIALS RECEIVED 1/4/07)

568 Volume 18—Number 7

The Reality of Recovered Memories

PSCI 1940(B W

U S

P S

C I

19 40

.P D

F 2

5- M

ay -0

7 19

:5 9

11 24

64 B

yt es

5 P

A G

E S

n o

pe ra

to r=

R av

is ha

nk ar

)

Author Query Form

_______________________________________________________

_______________________________________________________

Dear Author,

During the copy-editing of your paper, the following queries arose. Please respond to these by marking up your proofs with the necessary changes/additions. Please write your answers clearly on the query sheet if there is insufficient space on the page proofs. If returning the proof by fax do not write too close to the paper's edge. Please remember that illegible mark-ups may delay publication.

Journal PSCI Article 1940

Query No. Description Author Response

. No Queries