To dream, perchance to remember: individual differences in dream recall

Abstract

Each day for 14 weeks, 193 college students indicated whether or not they remembered a dream from the previous evening; they also recorded their sleep schedules and noted whether they had (1) exercised and (2) consumed alcohol and caffeine. In addition, the students completed measures of dissociation, schizotypy, sleep-related experiences, and the five-factor model of personality (N=169). Analyses of these data indicated that individual differences in dream recall were strongly stable over a 2-month interval. Dream recall was specifically associated with openness and was unrelated to the other Big Five traits. Subsequent analyses indicated that individuals who are prone to absorption, imagination and fantasy are particularly likely to remember their dreams and to report other vivid nocturnal experiences. These results are consistent with a salience model of dream recall and a continuity model of human consciousness.

Introduction

Our understanding of dreams was fundamentally changed with the discovery of rapid eye movement (REM) sleep in 1953 (Aserinsky & Kleitman, 1953). REM sleep regularly follows non-REM (NREM) sleep in a recurring cycle that lasts roughly 90 minutes in humans (e.g. Dement & Vaughan, 1999); thus, over the course of a typical night, the average adult will experience 4–6 REM periods. One of the most remarkable discoveries of modern sleep research is that when people are awakened from REM sleep, they report they were dreaming approximately 80–85% of the time (e.g. Goodenough, 1991, Squier & Domhoff, 1998). Moreover, dreams are recalled from a substantial percentage of NREM awakenings (Goodenough, 1991, Squier & Domhoff, 1998). The implications are striking: Virtually every adult human dreams several times each evening. In fact, Hobson (1988) estimated that the average adult dreams for roughly two hours per night; over the course of a 70-year lifetime, this would translate into approximately 50,000 h (i.e. nearly 6 years) of dreaming.

However, the dream experience is ephemeral and easily forgotten (Cohen, 1974, Goodenough, 1991). Upon awakening in the morning, the average person remembers having a dream only about half the time (Goodenough, 1991, Webb & Kersey, 1967). It is noteworthy, moreover, that there are enormous individual differences on this dimension: Whereas some people remember a dream almost every day, others report virtually no dream recall whatsoever (Cohen, 1974, Fitch & Armitage, 1989, Goodenough, 1991).

What gives rise to these individual differences in dream recall? Much of the research in this area derives from the salience hypothesis. This hypothesis represents a natural extension of the more general cognitive literature on the processes that influence the recall of information (see Cohen, 1974, Goodenough, 1991). Salient stimuli—for example, items that are novel, intense or unusual—are more easily remembered. Dreams also conform to this general pattern; thus, considerable evidence indicates that vivid, intense and unusual dream content is more likely to be remembered (Cohen, 1974, Goodenough, 1991). For instance, Cipolli, Bolzani, Cornoldi, de Beni, and Fagioli (1993) classified dream content as either bizarre or nonbizarre, and found that the rate of delayed recall for the former was approximately twice that of the latter.

Extended into the realm of individual differences, this model posits that people who tend to have more memorable dreams—that is, dreams that are highly vivid, intense, unusual, and interesting—should show better overall recall. This basic idea has received broad support. This evidence comes from two interrelated lines of research. The first line emphasizes the normal and adaptive aspects of dream recall, and is based on a continuity model of human consciousness (Blagrove & Hartnell, 2000, Claridge et al., 1997, Goodenough, 1991) . This model assumes that people who have interesting, vivid and unusual experiences during the day—for instance, those who are highly creative, imaginative, and prone to fantasy—also will have more salient and memorable dreams and, hence, show better dream recall. Consistent with this idea, more frequent dream recall is associated with higher scores on measures of creativity, daydreaming, fantasy proneness, absorption, and hypnotic susceptibility (e.g. Fitch & Armitage, 1989, Hartmann, 1991, Hill et al., 1997, Schredl et al., 1997).

The second line of research focuses more on the bizarre aspects of the dream experience, and reflects a broader tradition that has emphasized the similarities between dreams and psychotic phenomena such as hallucinations (Claridge et al., 1997, Levin, 1998, Levin & Raulin, 1991). In essence, this view argues that people who generally have unusual, even bizarre experiences during the day—for instance, those who receive high scores on measures of schizoptypy and psychosis proneness—should remember more of their dreams. In support of this idea, indicators of schizotypy are associated with higher scores on measures of both (1) general dream recall and (2) the frequency of reported nightmares (Claridge et al., 1997, Levin, 1998, Levin & Raulin, 1991).