The evolution of neural systems for sleep and dreaming
Despite the importance of sleep and dreams for the understanding of human consciousness, science is yet to achieve a consensus about their functions and intricate phenomenology. This article outlines an evolutionary theory of how sleep and dreams were selected over time due to their effects on cognition. The theory proposes that slow-wave sleep (SWS) co-evolved with thalamocortical loops that underlie sensory disconnection, from mere rest in protoreptiles to an extended and periodic quiescent state able to promote offline memory reverberation in amniotes. A second sleep state, characterized by short duration (seconds) and high cerebral activity, co-evolved in crocodiles, birds, and mammals with a complex set of pontine, midbrain, and forebrain structures. This post-SWS state, called rapid-eye-movement sleep (REM), induces genes linked to the stabilization, strengthening, and propagation of memories. Extended single REM episodes lasting several minutes evolved exclusively in the mammalian lineage. Prolongation of the noisy mnemonic reverberation that characterizes REM likely facilitates memory restructuring ('insight') rather than memory strengthening. Dreams as narratives require the activation of selected cortical and subcortical regions related to mnemonic representation and reward. During dreams, memory fragments are concatenated so as to simulate past events and future expectations. The fitness-enhancing function of dreams is to enact potential solutions for the cognitive challenges facing the dreamer. Though probabilistic, dreams can at times yield accurate predictions of future events, a fact of documented importance throughout human history. Finally, it is possible for extant humans to experience an enhanced REM state in which dream events are under partial or total voluntary control. Such lucid dreams likely require prefrontal activity, and their cognitive potential remains uncharted territory for science. © 2007 Elsevier Inc. All rights reserved.
