DESCRIPTION (provided by applicant): The purpose of this R21 proposal is to develop a new research program that will explore potential cognitive processing specializations of REM (rapid eye movement) and NREM (non-REM) sleep states. Identifying such processing specializations would point to hitherto unrecognized correlates of functional specializations for REM and NREM and thus may lead to a breakthrough in our understanding of sleep function. In a recent (McNamara, McLaren, Smith, Brown, & Stickgold, 2005) analysis of over 500 mentation reports from the REM, NREM and wake states, we found that REM exhibited greater numbers of simulations of aggressive social interactions than NREM or waking reports, and NREM exhibited greater numbers of simulations of friendly interactions than REM or wake reports. Indeed, there were absolutely no (zero) simulations of aggressive interactions in NREM reports. We plan to follow-up on and extend these findings by performing overnight polysomnographic recordings along with REM and NREM awakenings in a group of healthy young volunteers. We will explore the following questions in the proposed experiments: 1) Can the McNamara et al. (2005) findings be replicated and extended? Is there any evidence for REM and NREM cognitive processing specializations with respect to a) simulations of social interactions, b) emotions, and c) attributions of intentional states to other dream characters ("Theory of Mind") and 2) Is content of a NREM "dream" related to the content of a REM "dream" across a single night of sleep? Identifying cognitive processing specializations of REM and NREM sleep will yield better quantitative models of a variety of sleep disorders and better treatment of sleep-related mental health disorders.