Abstract Mkid 67

Although it is widely accepted that mentation can occur during NREM sleep, there is disagreement over whether such mentation is qualitatively equivalent to that occurring during REM sleep. This study approaches the problem from the perspective of narrative structural analysis, assessing sleep mentation reports for evidence of the basic attributes of story narratives. Twenty-four college students, of whom 12 were high and 12 were low frequency habitual recallers of their dreams, contributed mentation reports after experimental awakenings from stage REM and stage 2 sleep. A story grammar tool was used to parse these reports into their constituent components (actions, scenes, characters) and to identify the causal precursors and consequences of the constituent actions. Three measures of story coherence were derived: story constituent recall, story constituent co-occurrence, and episodic progression. For these measures, it was found that (a) more stage REM than stage 2 reports contained at least one story constituent, but only for the late night reports of high frequency recallers; (b) stage REM and stage 2 reports contained at least one instance of story constituent co-occurrence in about equal proportions; and (c) a greater proportion of stage REM than stage 2 reports contained at least one minimal story unit— episodic progression—but again only for high frequency recallers. Results support the notion that stage REM dreams are stories in the specific sense suggested by our microstructural criteria for episodic progression. They also speak to discrepancies among prior attempts to discriminate stage REM from stage 2 mentation reports, specifically, they underline the need for considering time-of-night and habitual dream recall frequency in such analyses.