Santa Fe Institute Collaboration Platform

COMPLEX TIME: Adaptation, Aging, & Arrow of Time

Get Involved!
Contact: Amy P. Chen, Program Manager,

What is Sleep?/How does zooming out to look across species help us to then zoom in on sleep function?

From Complex Time

November 18, 2019
2:00 pm - 3:00 pm


Jerome Siegel (UCLA)Alex Herman (Univ. Minnesota)Van Savage (UCLA/SFI)


Jerry Siegel – A popular approach to investigating sleep function is to deprive people, or animals, of sleep and note the changes that emerge.  However, sleep deprivation is necessarily stressful, no matter how gentle the stimuli applied, making it unclear if the observed effects are due to cortisol release or other aspects of the procedure.  Sleep duration in animals varies from 2 to 20 hours, with a proportionate range in REM sleep amounts.  This variation can be correlated with other aspects of physiology to gain an insight into the evolutionary determinants of sleep without the confounds resulting from deprivation.  I will briefly review the animal sleep literature.

Alex Herman, Van Savage - Sleep is a nearly ubiquitous and evolutionarily ancient process experienced by multicellular animals. Despite the enormous diversity of animals that sleep and the huge range in time scale for different aspects of sleep--total time, REM, circadian--systematic patterns emerge that suggest unifying underlying principles when viewed in the correct mathematical and conceptual spaces. Linking these large-scale patterns to theories about their biological and biophysical origins allows us to derive equations that shed insight into how sleep varies across species. Thus, finding the right lens with which to “zoom in” requires first “zooming out”. To that effect we show that the total time spent sleeping and the length of sleep cycles vary as power-laws with body size across about 6 orders of magnitude. Building on prior allometric scaling theory, we are able to predict these power-laws based on a theory for the central role of repair in sleep. Scaling according to these patterns reveals, for instance, that herbivores display significantly less variability in sleep times than carnivores or omnivores, potentially yielding insight into differential evolutionary pressures on sleep function. Moreover, we have recently examined how sleep time and cycle length vary as humans grow in body and brain size, as opposed to the commonly used chronological age. We find two distinct scaling regimes that emerge from the data and are separated by a distinct transition. We are able to predict these scaling regimes and the transition with reasonable accuracy by using a theory--supported by experimental data--that in early life sleep serves primarily to facilitate changes in synaptic density and white matter connectivity, while in later life the process of sleep is dominated by maintenance and repair. Our theory helps lay the groundwork for understanding both individual ontogenetic-growth and across-species variation in brain development and sleep dynamics, yielding a powerful new lens with which to zoom in on the functions of sleep.

Presentation file(s)
Related files