Dreams, endocannabinoids and itinerant dynamics in neural networks: re elaborating Crick-Mitchison unlearning hypothesis

144 (short abstract) and 218 (long abstract) SHORT ABSTRACT In this work we reevaluate and elaborate Crick-Mitchison's proposal that REM-sleep corresponds to a self-organized process for unlearning attractors in neural networks. This reformulation is made at the face of recent findings concerning the intense activation of the amygdalar complex during REM-sleep, the involvement of endocannabinoids in synaptic weakening and new neural network models with itinerant associative dynamics. We distinguish between a neurological REM-sleep function and a related evolutionary/behavioral dreaming function. At the neurological level, we propose that REM-sleep regulates excessive plasticity and weakens over stable brain activation patterns, especially in the amygdala, hippocampus and motor systems. At the behavioral level, we propose that dream narrative evolved as exploratory behavior made in a virtual environment promoting " emotional (un)learning " , that is, habituation of emotional responses, anxiety and fear. Several predictions of the unlearning idea that are at variance with the memory consolidation hypothesis are discussed. LONG ABSTRACT In this work we reevaluate and elaborate Crick-Mitchison's proposal that REM-sleep corresponds to a self-organized process for unlearning attractors in neural networks. This reformulation is made at the face of recent findings concerning the intense activation of the amygdalar complex during REM-sleep, the involvement of endocannabinoids in synaptic weakening and new neural network models with itinerant associative dynamics. We distinguish between a neurological REM-sleep function and a related evolutionary/behavioral dreaming function. At the neurological level, we propose that REM-sleep regulates excessive plasticity and weakens over stable brain activation patterns, especially in the amygdala, hippocampus and motor systems by over activating the cannabinoid system. This proposal is motivated by the high density of endocannabinoid receptor CB1 in regions highly active during REM, the high level and efficiency of endocannabinoids during sleep and unlearning-like effects of CB1 receptor activation. At the behavioral level, we propose that dream narrative evolved as exploratory behavior made in a virtual environment promoting " emotional (un)learning " , that is, optimization of emotional responses, fear learning and anxiety level. Several predictions of the unlearning idea that are at variance with the memory consolidation hypothesis are discussed. Our proposal is illustrated by an itinerant associative memory model (a Hopfield model with transition between attractors and self-organized unlearning dynamics) that mimics William James' dynamical ideas about the stream of thought. Taking a purely naturalistic view of the matter, it seems reasonable to suppose that, unless consciousness served some useful purpose, it …