UC Davis

This study aimed to investigate the cycles (2nd/4th) and duration-related (5/10 min) variations in the story-like organization of dream experience elaborated during rapid eye movement (REM) sleep. Dream reports were analysed using story grammar rules. Reports were provided by those subjects (14 of 22) capable of reporting a dream after each of the four awakenings provoked in 2 consecutive nights during REM sleep of the 2nd and 4th cycles, after periods of either 5 or 10 min, counterbalanced across the nights. Two researchers who were blind as to the sleep condition scored the dream reports independently. The values of the indicators of report length (measured as value of total word count) and of story-like organization of dream reports were matched taking time-of-night (2nd and 4th cycles) and REM duration (5 versus 10 min) as factors. Two-way analyses of variance showed that report length increased significantly in 4th-cycle REM sleep and nearly significantly for longer REM duration, whereas the number of dream-stories per report did not vary. The indices of sequential (number of statements describing the event structure developed in the story) and hierarchical (number of episodes per story) organization increased significantly only in dream-stories reported after 10 min of 4th-cycle REM sleep. These findings indicate that the characteristics of structural organization of dream-stories vary along with time of night, and suggest that the elaboration of a long and complex dream-story requires a fairly long time and the availability of a great amount of cognitive resources to maintain its continuity and coherence.