Recovered Memories: Forgetting What We've Remembered?

One of my near obsessions in cognitive science is the recovered memories debate. Not only has it been one of the most contentious debates in the field over the last 2 decades, but its practical implications are some of the most profound. There are people in jail right now largely as a result of recovered memory testimony, and some of them will be there for a long time. On the legal side, the important issue is how accurate recovered memories are, and to date, the best we can do is say that in some cases they probably are, while in others they probably aren't. We don't really have a good way of determining the accuracy of recovered memories in individual cases, and to date, no one's produced anything like a good theory of how memories might be repressed and later recovered. I've talked about the accuracy and mechanisms issues before, back on the old blog. In this post, I'm going to talk about a recent study that addresses another issue: were recovered memories ever really forgotten? That question probably doesn't matter in legal contexts, but it could have important theoretical and therapeutic implications1.

This issue rose to the fore when Schooler et al.2, in the process of documenting several cases of recovered memories, discovered two interesting cases. People close to two individuals who had recovered memories of childhood sexual abuse reported that those individuals had discussed their abuse prior to "recovering" memories of it. It seems that they may have forgotten that they'd actually remembered the abuse all along. Several studies have since shown that under certain conditions, people do have a difficult time remembering what they've remembered.

For example, Arnold and Lindsay3 presented participants with a list of homographic words (e.g., "palm"; these were the targets), each of which was paired with a context word chosen to suggest one of the homograph's meanings (e.g., "hand" or "tree"). After learning the list, they were given a memory test in which they were given a cue and asked to recall the target word associated with the cue. One third of the cues were not associated with words on the list, one third of the cues were the context words they'd seen with the target words (e.g., "hand"), and one third were the context words not seen in the learning phase (e.g., "tree"). After a delay, participants were given a similar memory test, but in addition to being asked to recall the target words when given a cue (this time all cues were the original context words from the learning phase), they were also asked to say whether they had remembered the target word in the first memory test. During this second memory test, participants were significantly more likely to recall remembering the word in the first test if the cue in the first test had been the original context word. In other words, changing the context in the first memory test made it more difficult to remember recalling the word in the second test.

These and similar studies are interesting, in that they show it's possible to forget that one had previously remembered something, but they also raise an interesting question. If forgetting that one remembered can explain some cases of recovered memories, why do some people forget that they've previously remembered childhood abuse, while others don't? To begin to answer this question, Geraerts et al.4 use a paradigm similar to the one used by Arnold and Lindsay with three types of participants: 57 participants who had recovered memories of childhood abuse, 69 participants who had continuous memories of childhood abuse, and 68 participants with no memories of childhood abuse (the control group). Once again, homographs were first learned with a context word that suggested one meaning, and then participants were given two memory tests, the first of which included either the original or alternative context words as recall cues, and the second of which asked whether they had recalled the word in the first test. All participants were less likely to remember recalling a word in the first test if they had received the alternative cue in that test, but in this condition, participants with recovered memories of abuse were significantly less likely to remember recalling a word in the first test than either control participants or participants with continuous memories of abuse.

In the second experiment, new participants from the same three groups were first asked to remember particular types of childhood experiences (e.g., being home alone or going to the dentist). For each type of experience, participants were told to recall either a positive or negative experience. Here are their examples:

Given the event cue "being home alone as a child," for example, a participant assigned a positive framing for that event might recall enjoying the feeling of freedom of having the house to his/herself; the same participant assigned a negative framing for that event in Session 2 might reminisce about feeling lonely after a while.

Two months after this first phase, participants were asked to remember some of the experiences they'd reported. For half of the experiences, they were cued with the original "emotional frame" (i.e., positive or negative), and for half, with the opposite. Two months after this second phase, the participants were again asked to recall the experiences they'd reported in the first phase (this time with cues that were from the original "emotional frame"), and also to recall whether they'd remembered each experience during the second phase. Once again, in the third phase all participants were less likely to recall that they'd remembered an experience in the second phase if it had been presented in a different context during the first and second phases, and as in the first experiment, participants with recovered memories of childhood abuse were significantly less likely to recall remembering experiences in this condition than the other two groups.

In sum, then, the Geraerts et al. study shows that, while everyone has trouble recalling whether they've remembered information when the contexts in which that information is remembered over time are varied, but individuals with recovered memories of childhood abuse are particularly bad at it. This suggests that individuals with recovered memories of abuse may be more likely to forget abuse that they've remembered all along. Now, this research doesn't say anything about the accuracy of their "recovered" memories, or even that all cases of recovered memories can be explained as instances of people forgetting what they've remembered all along, but it does provide a new route for studying recovered memories. The next questions for researchers, then, are why do individuals with recovered memories do worse in these tasks when contextual information is varied, and then, can this really explain at least some cases of recovered memories?

1One of the more traumatic aspects of recovered memories is that they come as such a surprise to the recoveree. If you've lived a life without reconciling childhood abuse, suddenly having to do it in adulthood can be very difficult.
2Schooler, 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 Research and Clinical Practices (pp. 379-388). New York: Plenum.
3Arnold, M. M., & Lindsay, D. S. (2002). Remembering remembering. Journal of Experimental Psychology: Learning, Memory, & Cognition, 28, 521-529.
4Geraerts, E., Arnold, M. M., Lindsay, D. S., Merchelbach, H., Jelicic, M., & Hauer, B. (2006). Forgetting of prior remembering in persons reporting recovered memories of childhood sexual abuse. Psychological Science, 17, 1002-1008.

More like this

Speaking of studies that make you go "wow," I recently read a very recent one that really surprised me, and thought I'd describe it here. Memory researchers are famous for coming up with different types of memory, sparking years of research and debate just trying to determine whether two particular…
In the past, I've often wondered how journalists pick which studies to write about. The obvious answer is that they pick studies that will get readers or viewers, but given how little their stories correspond with the research they're writing about, it seems to me like they could pick any study and…
One of the more sophisticated theories in embodied cognition is Lawrence Barsalou's perceptual symbol systems theory (which I've talked about before, here). Starting from the premise that cognition is for action, it argues that concepts are represented in the same brain systems that directly serve…
What was your 6th birthday party like? If you successfully retrieved that memory, you may now be ever so slightly less able to remember your other childhood birthdays. A variety of behavioral evidence has shown that such "retrieval induced forgetting" of strongly competing memories is…

Your previous blog post was much more direct; in fact in this one, you seem to suggest that it's even valid to say that people forget that they forget. I personally hate to any comments to the public discourse that can be construed as support for a recovered memory position.
Maybe we ought to talk about these results as retrieval failure, which people might consider to be quite different than forgetting.
The recent work you describe is an offshoot of encoding specificity work by Tulving, right? So in the new work by Arnold and Lindsay and colleagues, we are seeing that people who claim recovered memories of CSA are perhaps more likely to deeply encode the surrounding context of a learning experience. This idea can be tested with the original Craik and Tulving depth-of-processing experiments, and also with the Hasher Burglar/Real Estate buyer schema reevaluation procedure. In the first case, people who claim recovered CSA ought to show enhanced depth or processing efects, and in the second case they ought to be able to bring up fewer details from a non-original schema when asked to re-remember. Do you know if work like this has been done?
Thoughtful post, as usual...

Thank you for posting this research and the link to the older blog. My students asked me about if people can really remember with accuracy sexual abuse after 20 and 30 years after the fact. This research supports our class discussions about remembering long term memories, also flashbulb memories, and the fact that memory of a specific event is not a data file or a DVD stored in our long term memory; rather a flexible file that we can reconstruct or not be able to access with accuracy.

By Zsuzsanna (not verified) on 01 Dec 2006 #permalink

Hi. I found this post very interesting. Do you know of any other pracitcal implications of false memory, besides those related to eyewitness testimonies and law?