Memory researchers have shown that remembering almost always in- volves a constructive process—whether the person is remembering a word list (Deese, 1959; Roediger & McDermott, 1995), a story (Bartlett, 1932), a song (Hyman & Rubin, 1990; Rubin, 1995), or a personal experi- ence (Barclay & DeCooke, 1988; Neisser, 1982). Memory construction is guided both by internal schematic knowledge and by external sugges- tions. Using the classic misinformation paradigm, Loftus and her col- leagues (see Loftus, 1991, 1992) showed people an event, provided neutral or misleading postevent information, and observed poorer memory per- formance following the misleading information. Often, the misleading in- formation would affect not only the particular aspect of the event, but also show how the entire memory was reconstructed based on the new under- standing of the event (Loftus & Palmer, 1974).
Applying this research on memory errors to the creation of false child- hood memories is, however, difficult for several reasons. First, research on the misinformation effect demonstrates that people can change an aspect of their memory for an event that happened. Misinformation studies often
involve the participant as an observer, not as an actor. In addition, the events studied in misinformation research are generally not emotional, al- though there are some studies that involve emotional events. One of the most stunning examples of the pliability of memory for highly upsetting events comes from a study of Dutch citizens who tried to recall the crash of an El Al Boeing 747 into an apartment building in Amsterdam (Crom- bag, Wagenaar, & Van Koppen, 1996). Although film crews rushed to the scene and quickly filmed the fire and rescue of survivors from the build- ing, there was no film of the crash. Nonetheless, when people were asked to recall the crash 10 months later and asked leading questions, 66% claimed to have seen TV film of the crashing plane.
Still, most misinformation studies concern memory change of features of unemotional events that do not involve the self as a participant. In thinking about the relation between the misinformation studies and false memories of child sexual abuse, certain factors must be kept in mind. Cre- ating a false memory of child abuse would involve the construction of an entire event rather than changing an aspect of an event. If someone creates a memory of childhood sexual abuse, such an event would be emotional and would involve the self as a central character. These differences could easily prompt one to conclude that the creation of false childhood memo- ries is at least unlikely, if not impossible. Thus the research question be- came: Will people create entire false memories of an emotional event that involves the self?
When studying the creation of false childhood memories, the research- ers’ goal was to extend the exploration of memory errors. Most research- ers generated similar methodologies to suggest to participants a complete memory that involved the self and that was somewhat emotional, al- though never traumatic (Ceci, Huffman, Smith, & Loftus, 1994; Ceci, Lof- tus, Leichtman, & Bruck, 1994; Hyman & Billings, 1998; Hyman, Husband, & Billings, 1995; Hyman & Pentland, 1996; Loftus & Pickrell, 1995). The re- search methodology was an outgrowth of misinformation studies and research investigating the ability of adults to recall early childhood experi- ences (e.g., Usher & Neisser, 1993). In most paradigms, researchers re- quested information from family members about events that occurred during the participant’s childhood. Then, the participant was asked to re- call these true events. In addition, the participant was asked about a false event—an event that the researchers were fairly sure did not happen to the participant. During a series of interviews, the false event was pre- sented as if it were a true event that was obtained from the initial family solicitation. The participants were usually interviewed repeatedly about both the true and false events and told that their memories would im- prove over time. Although there is an interest in memory for the true ex- periences, the most important result is how the participants responded to
the false event—Do the participants come to believe that the event took place sometime during their childhood?
For example, Hyman et al. (1995) used this methodology in a study in- vestigating college students’ childhood memories. First, the researchers visited an introductory psychology class and asked the students for per- mission to mail a questionnaire to their parents. The students were told that the researchers were interested in how well people could recall child- hood experiences and therefore needed information from parents about what had happened. The researchers obtained descriptions of true child- hood events from the parent surveys. When the parents returned the questionnaires, the researchers asked the introductory psychology stu- dents to participate in a series of interviews investigating their memory for early childhood experiences. The students were told that the research- ers were interested in how completely and accurately they could recall childhood. In each of three interviews (separated by 1 day), the students were asked to remember several true events and one false event. For all true events and the false event, the interviewer provided the students with a basic description (including age, event, a few actions, other people involved, and a location) and asked the students what they remembered about the event. One of the false events used was called the punch bowl event: When you were 6 years old, you were at the wedding of a friend of the family, you were running around with some other kids at the recep- tion, and you bumped into the table the punch bowl was sitting on and spilled punch on the parents of the bride. All of the false events used were self-involving and would have been somewhat emotional at the time of the event, although none were traumatic events.
The participants recalled a majority of the true events in the first inter- view and remembered even more of the true events over time. There are at least two ways to explain the increased recall of the true events. First, by thinking about the events over a period of time, the students provided themselves with additional memory cues that led to the recollection of previously unretrieved memories. Second, the participants created, rather than recalled, memories that matched the cues provided in the interviews. We cannot say whether this recovery of memory for the true experience represents actual memories or the creation of memories.
Regarding the false events, none of the participants remembered the false event on its initial presentation. However, by the third interview, 25% of the students remembered the event. Six students reported memo- ries that were clear and included the critical information (such as turning over the punch bowl) as well as consistent elaborations (such as their par- ents being upset). Five of the reports were less clear; the students included little of the critical suggested information although they elaborated in a consistent fashion. Two of the students created clear images, but they
were not sure if they were remembering or simply imagining the events that had been suggested to them. Although Hyman et al. used only col- lege students in their study, other studies using various populations (e.g., preschool children, adults, teenagers) and different false events have found similar results (Ceci, Huffman, et al., 1994; Ceci, Loftus, et al., 1994; Loftus & Pickrell, 1995; Pezdek, Finger, & Hodge, 1997).
The proportion of individuals who create false childhood memories de- pends on the research method used. Using a similar methodology, Hyman and Pentland (1996) asked college students to imagine and de- scribe any events, whether true or false, that they could not remember. They speculated that imagining events would cause people to construct clearer images and narratives that they would later confuse with personal memories. By the third interview, nearly 40% of the college students re- membered spilling the punch bowl.
False childhood memory research has successfully extended the previ- ous work on memory suggestibility. Not only will people alter aspects on an event in response to suggestions, but they will also create entire events. The created events involve the self and are at least mildly emotional.
Nonetheless, questions of generalizability still remain. Spilling a punch bowl at a wedding or being lost in a mall is not the same as being sexually abused. Currently, no researcher has attempted to have participants cre- ate memories of being sexually abused. For ethical reasons, it is unlikely that anyone ever will; if a memory impacts one’s self-concept and family relationships whether it is true or false, then experimentally inducing such memories would be atrocious. This makes generalizing to sexual abuse memories difficult, but it is not impossible. Real-world cases often provide dramatic evidence that individuals can create false memories of a great variety of events. Recovered memory cases of alien abductions and satanic ritual abuse suggest that a variety of false memories can be cre- ated. Because in most of these cases the events cannot be true, but are nonetheless traumatic, evidence for the creation of false traumatic memo- ries is substantiated. Although many examples could be provided, a sin- gle one will suffice to support the point.
In the mid-1980s, Nadean Cool, a nurse’s aide in Wisconsin, sought therapy from a psychiatrist to help her cope with her reaction to a trau- matic event. During therapy, the psychiatrist used hypnosis and other suggestive techniques to reveal buried memories of abuse that Cool had allegedly experienced. In the process, Cool became convinced that she had repressed memories of being in a satanic cult, eating babies, being raped, having sex with animals, and being forced to watch the murder of her 8-year-old friend. She began to believe that she had over 120 personal- ities—children, adults, angels, and even a duck. She was repeatedly urged to believe that these events had occurred. The psychiatrist also performed
exorcisms on her; one lasted for 5 hours and included the sprinkling of holy water and screams for Satan to leave Cool’s body. Such cases demon- strate the extreme experiences that people “remember” in response to re- peated suggestions. How do people believe such unlikely things as being forced to eat babies or watch the murder of a friend?
THREE PROCESSES INVOLVED IN MEMORY