Memory for past events in the elderly as a function of the decade in which the events occurred. Based on Rubin et al. (1986).

of information about lifetime periods helps to ensure that most forms of brain damage do not prevent access to such knowledge.


The notion that autobiographical memory is organised in a hierarchical structure is useful. However, it is not clear that the general-event level is as important as was suggested by Conway. Berntsen (1998) distinguished between voluntary and involuntary autobiographical memories. Most research has involved presenting cues to elicit autobiographical memories, and thus focuses on voluntary memories. In contrast, an involuntary autobiographical memory is one that "comes to mind without preceding attempts at retrieving this memory" (Berntsen, 1998, p. 118). Involuntary autobiographical memories were obtained by asking participants to keep a record of them. A much higher percentage of involuntary than of voluntary memories were of specific events (89% vs. 63%, respectively). As Berntsen (1998, p. 136) pointed out, "The results suggest that we maintain a considerable amount of specific episodes in memory which may often be inaccessible for voluntary retrieval, but highly accessible for involuntary recall."

The key implication of Bernsten's findings is that the hierarchical level that seems to be most important depends on the methods used to study autobiographical memory. As Bernsten (1998, p. 138) pointed out:

If autobiographical memory constitutes an hierarchical is an hierarchy with no stable "basic" level. What is basic —in the sense of being most accessible— varies with the retrieval strategy employed. Notably, it seems to vary with whether retrieval is voluntary or involuntary.

Memories across the lifetime

Suppose that we ask 70-year-olds to think of personal memories suggested by cue words (e.g., nouns referring to common objects). From which parts of their lives would most of the memories come? Would they tend to think of recent experiences or the events of childhood or young adulthood? Rubin, Wetzler, and Nebes (1986) provided answers to these questions. There are various features about the findings (see Figure 8.1):

• A retention function for memories up to 20 years old, with the older memories being less likely to be recalled than more recent ones.

• A reminiscence bump, consisting of a surprisingly large number of memories coming from the years between 10 and 30, and especially between 15 and 25.

• Infantile amnesia, shown by the almost total lack of memories from the first five years of life.

The reminiscence bump has not generally been found in people younger than 30 years of age, and has not often been observed in 40-year-olds. However, it is nearly always found among older people. Rubin and Schulkind (1997) used far more cue words than had been used in previous studies. They found "no evidence that any aspect of the distribution of autobiographical memories is affected by having close to 1,000 as opposed to close to 100 memories queried" (p. 863). They also found that the reminiscence bump is not simply due to averaging across individuals. They studied five 70-year-olds, and found evidence for the reminiscence bump in each one of them.

Rubin, Rahhal, and Poon (1998) discussed other evidence that 70-year-olds have especially good memories for early adulthood. This effect was found for the following: particularly memorable books; vivid memories; memories the participants would want included in a book about their lives; names of winners of Academy Awards; and memory for current events.

Theoretical perspectives

How can we interpret these findings? The retention function presumably reflects forgetting over time, but the reasons for the reminiscence bump are less clear. It may be relevant that many new or first-time experiences are associated with adolescence and early adulthood, and such experiences are especially memorable. Cohen and Faulkner (1988) found that 93% of vivid life memories were either of unique events or of first times. Evidence that first-time experiences are very memorable was obtained by Pillemer et al. (1988). Their participants recalled four memories from their first year at college more than 20 years previously, with 41% of them coming from the first month of the course.

Rubin et al. (1998) developed a speculative cognitive theory of the reminiscence bump. According to their theory, "the best situation for memory is the beginning of a period of stability that lasts until retrieval" (pp. 13-14). They argued that most adults have a period of stability starting in early adulthood, because it is then that a sense of adult identity develops. Memories from early adulthood also tend to have the advantage of novelty, in that they are formed shortly after the onset of adult identity. These two factors of novelty and stability produce strong memories for the following reasons:

• Novelty: this causes more effort after meaning.

• Novelty: there is a relative lack of proactive interference (interference from previous learning).

• Novelty: this produces distinctive memories (see Chapter 6).

• Stability: events from a stable period of life are more likely to serve as models for future events.

• Stability: this provides a cognitive structure that serves as a stable organisation to cue events.

What about infantile amnesia? The most convincing explanation was provided by Howe and Courage (1997), who related it to the emergence of the self towards the end of the second year of life. Infants at about 20 months show signs of developing a sense of self in the phenomenon of visual self-recognition, which involves responding to their own image in a mirror with self-touching, shy smiling, and gaze aversion. A few months after that, infants start to use words such as I, me, and you. The crucial theoretical assumption made by Howe and Courage (1997, p. 499) is as follows: "The development of the cognitive self late in the second year of life (as indexed by visual self-recognition) provides a new framework around which memories can be organised. With this cognitive advance in the development of the self, we witness the emergence of autobiographical memory and the end of infantile amnesia."

It follows from this theoretical position that the lower limit for people's earliest autobiographical memories should be about 2 years of age, and that is consistent with the evidence. However, it is hard to show that the emergence of a sense of self is the causal factor. Howe and Courage (1997) also assumed that the processes (e.g., rehearsal) used in learning and memory develop during the years of childhood, and so relatively few autobiographical memories should come from the years 2 to 5. This is also in line with the evidence.

Diary studies

It is often not possible to assess the accuracy of an individual's recollections of the events of his or her own life. Linton (1975) and Wagenaar (1986) resolved this problem by carrying out diary studies, in which they made a daily note of personal events. Both of them later tested their own memory for these events at various retention intervals.

Linton (1975) wrote down brief descriptions of at least two events each day over a six-year period. Every month she selected two of these descriptions at random, and tried to recall as much as possible about the events in question. Forgetting depended substantially on whether or not a given event had been tested before. For example, over 60% of events that had happened years previously were completely forgotten if they had not been tested, compared to under 40% of events of the same age that had been tested once before. This finding indicates the importance of rehearsal in the prevention of forgetting.

One of the main reasons why events were forgotten was because many events were similar to each other. For example, Linton occasionally attended meetings of a distinguished committee in a distant city. The first such meeting was clearly remembered, but most of the subsequent meetings blended into one another. As Linton (1975) expressed it, her semantic memory (or general knowledge) about the meetings increased over time, whereas her episodic memory (or memory for specific events) decreased.

It might be imagined that those events that were regarded at the time as important and high in emotionality would be especially well remembered. In fact, the impact of importance and event emotionality on recallability was only modest, perhaps because rated importance and emotionality at retrieval did not correlate highly with each other. Thus, events that seemed at the time to be important and emotional often no longer seemed so with the benefit of hindsight.

What strategies do we use to remember events from our past? Linton (1975) considered how she set about the task of recalling as many events as possible from a given month in the past. When the month in question was under two years previously, the main strategy was based on working through events in the order in which they had occurred. In contrast, there was more use of recall by category (e.g., sporting events attended; dinner parties given) at longer retention intervals.

Wagenaar (1986) recorded over 2000 events over a six-year period. For each event, he noted down information about who, what, where, and when, together with the rated pleasantness, saliency or rarity, and emotionality of each event. He then tested his memory by using the who, what, where, and when information cues either one at a time or in combination. "What" information provided the most useful retrieval cue, perhaps because our autobiographical memories are organised in categories. "What" information was followed in order of declining usefulness by "where", "who", and "when" information. "When" information on its own was almost totally ineffective. The more cues that were presented, the higher was the resultant probability of recall (see Figure 8.2). However, even with three cues almost half of the events were forgotten over a five-year retention interval. When these forgotten events involved another person, that person was asked to provide further information about the event. In nearly every case, this proved sufficient for Wagenaar to remember the event. This suggests that the great majority of life events may be stored away in long-term memory.

High levels of salience, emotional involvement, and pleasantness were all associated with high levels of recall, especially high salience or rarity. The effects of salience and emotional involvement remained strong over retention intervals ranging from one to five years, whereas the effects of pleasantness decreased over time.

A more complex picture emerged when Wagenaar (1994) carried out a detailed analysis of 120 very pleasant and unpleasant memories from his 1986 study. When someone else played the major role in an event, pleasant events were much better remembered than unpleasant ones. However, the opposite was the case for events in which Wagenaar himself played the major role. Groeger (1997, p. 230) speculated that this latter finding may reflect Wagenaar's personality: "What Wagenaar does not address is the possibility that he may actually have a tendency to be rather self-critical or self-effacing."

The case studies of Linton (1975) and of Wagenaar (1986) are of considerable interest. However, we need to be cautious about assuming that everyone's autobiographical memory system functions in the same way. For example, anxious and depressed individuals recall a disproportionate number of negative events (see Chapter 18), and this recall bias may colour the way in which they remember their own past. What we remember of our own lives depends in part on our personalities.

Dating autobiographical memories

Linton (1975) and Wagenaar (1986) both found they were fairly good at dating the events of their lives. How do we remember when past events happened? People often relate the events of their lives to major lifetime periods (Conway & Bekerian, 1987). In addition, we sometimes draw inferences about when an event happened on the basis of how much information about it we can remember. If we can remember very little about an event, we may assume that it happened a long time ago. This idea was tested by Brown, Rips, and Shevell (1985). People were asked to date several news events over a five-year period (1977 to 1982). On average, those events about which much was known (e.g., the shooting of President Reagan) were dated as too recent by over three months, whereas low-knowledge events were dated as too remote by about three months.

In a follow-up study, Brown, Shevell, and Rips (1986) asked participants to date public events that were either political (e.g., the signing of a major treaty) or non-political (e.g., the eruption of Mount St. Helens). The participants made much use of landmarks, i.e., events whose dates they knew well. For example, someone might be able to date the eruption of Mount St. Helens by relating it to the landmark of becoming engaged shortly beforehand (the start and end of lifetime periods are effective landmarks). Landmarks were used about 70% of the time to aid the dating of public events, with the landmarks being either public or

Stop Anxiety Attacks

Stop Anxiety Attacks

Here's How You Could End Anxiety and Panic Attacks For Good Prevent Anxiety in Your Golden Years Without Harmful Prescription Drugs. If You Give Me 15 minutes, I Will Show You a Breakthrough That Will Change The Way You Think About Anxiety and Panic Attacks Forever! If you are still suffering because your doctor can't help you, here's some great news...!

Get My Free Ebook

Post a comment