Memory sources of REM and NREM dreams.

Sixteen male volunteers slept 4 nonconsecutive nights each in a sleep laboratory. They were awakened for one dream report per night. Awakenings were made, in counterbalanced order, from early-night and late-night rapid-eye movement (REM) and non-REM (NREM) sleep. Following dream reporting, subjects were asked to identify memory sources of their dream imagery. Two independent judges reliably rated mentation reports for temporal units and categorized memory sources as autobiographical episodes, abstract self-references, or semantic knowledge. We replicated earlier findings that semantic knowledge is more frequently mentioned as a dream source for REM than for NREM reports. However, with controls for length of reports, the REM-NREM difference disappeared, indicating that the stage difference in memory sources was not independent of stage difference in report lengths. There was a significant effect of time of night on source class, but only in REM sleep: Both without and with controls for report length, more semantic sources were cited for late than for early REM dreams.

Right hemispheric mediation of dream visualization: a case study.

Our paper addresses, as did a recent study by Greenwood et al. (1977), the question of differential right vs. left hemispheric involvement in REM-sleep dreaming. A male subject with extensive lesioning of the right hemisphere and consequent contralateral homonymous hemianopia was studied for 2 nights in a sleep laboratory where a dream report was collected following each of 6 awakenings from REM sleep. The subject's dream reports were typically sequential and narrative in form, but were lacking specifically in continuous visual imagery. The occasional visual images that appeared in the subject's dreams were described as fragmentary and static in nature. The subject also was unable to perform waking tasks requiring the kinematic representation of extrapersonal space. These observations support the hypotheses that dream narratives may derive from different neurocognitive sources than do the specifically visual realizations of those narratives, and that the visual mediation of dream experience may depend on right-hemispheric processing. Differences between our findings and those of Greenwood et al. may be attributable to differences in dream interview technique.