Verified by 
Most evolutionary neuroscientists now accept the fact that a significant contributor to the disproportionate growth in brain size among the hominin line leading to us anatomically modern humans was that our ancestors began to get less aggressive and more social. We became more cooperative, more groupish, and more social. At some point these innovations in group living gave rise to human culture and we were off to the races with culture selecting for ever greater brain complexity and brain complexity supporting ever more complex social relationships and cultural innovation.
Although there are many hypotheses, what sparked the turn toward less aggression and greater social cooperation remains a mystery. That the “turn” occurred however cannot be denied. I now want to ask what the consequences for the sleeping brain were of this evolutionary turn away from aggression and toward the social. My answer will be that the “turn” re-shaped both NREM and REM sleep properties with NREM sleep supporting the social turn and REM sleep preserving pre-turn brain properties.
The trend toward less aggression and pro-sociality is most prominent among apes with bonobos and humans being outliers in that regard. What is striking about sleep among the apes is a gradual differentiation of NREM sleep into several subphases with humans exhibiting three such subphases and the other apes at least one or two.
Now what NREM sleep does is broadly restorative. Among other things, it restores sharpness of brain function. NREM is associated in its deepest phases with appearance of delta waves in those areas of the brain that were most in need of restoration, i.e., those most used during the waking period. For humans this is almost invariably the frontal lobes. The regular more global changes in delta wave activity that occurs each night during NREM appears to be more strongly related to use or engagement of particular regions of the frontal lobes and its interconnected regions than it is to other areas of the brain.
Why is NREM sleep so concerned with the frontal lobes? The frontal lobes are absolutely crucial if you want to reduce aggressivity and increase pro-sociality. I presented evidence in my 2019 “Neuroscience of sleep and dreams” that the “social brain” network, with the frontal lobes as a key functional site, was a particular target of sleep restorative processes in humans. The brain network that is typically called “the social brain network” is that set of interconnected brain regions that handle or mediate all of the thinking and emotional work we have to do to keep track of and regulate our social interactions with others. The network of structures that make up the social brain include key structures of the default mode network and the amygdala, the prefrontal cortex, the cingulate cortex, and the temporal-parietal junction. NREM sleep stages appear to function at least in part to restore functionality of this network of these structures. For example, in the first NREM episode of the night, there is relatively greater delta indexed slow-wave activity (SWA) in frontal than in parietal and occipital regions. Synchronization of slow-wave activity spreads progressively from frontal lobes to posterior and subcortical regions.
Interestingly, SWA can be expressed in one brain hemisphere at a time, again dependent on use during waking activities. This raises the question of whether that is the case also for REM. To my knowledge REM can only be expressed bihemispherically.
If the turn away from aggressivity and toward pro-sociality re-shaped NREM sleep into its current three differentiated stages that support restorative functionality for the frontal lobes, what has the “turn” done to REM sleep?
Well, interestingly REM is characterized by a kind of opposite profile to that of NREM. There is reduced activation in the dorsal prefrontal regions and increased activation among amygdalar networks. Given that amygdalar networks mediate various forms of emotion and aggressivity, the combination of reduced prefrontal inhibitory pressure and increased amygdalar reactivity during daytime activity is generally associated with impulsive or reactive forms of aggression.
It appears then that REM sleep not only does not support restoration of prefrontal inhibitory functions, it actively opposes it via down regulation of frontal activity and upregulation of amygdalar activity. In short, REM appears to oppose the evolutionary turn toward greater sociality.
Dream content studies support this general characterization of NREM and REM functions. NREM dreams contain zero instances of dreamer-initiated aggressions, while conversely REM dreams are filled with dreamer-initiated aggressions; and conversely NREM dreams are filled with friendly social interactions between dream characters while REM dreams have greater aggressive than friendly interactions.
Can any current evolutionary hypothesis of human sleep explain these differing REM vs. NREM sleep functional characteristics? The genetic conflict hypothesis I described in previous work can, but in addition, the human self-domestication hypothesis (HSD) could. HSD posits that selection against aggression operated in human evolution and produced morphological, physiological, behavioral, and psychological side effects similar to those reported as characteristic of domesticated animals. More on this in a future post.
