This is the fifth installment of a five-part series.
Either way, if snoozing helps the brain to alter its circuitry, what kinds of rewiring are most important? For instance, are connections between neurons—in some parts of the brain or during some portions of sleep—getting stronger, as the rebuilding of circuitry might suggest? Or are they mostly getting weaker? This latter possibility is central to a new hypothesis from Giulio Tononi, a neuroscientist and veteran sleep researcher at the University of Wisconsin, Madison. Tononi proposes that sleep may enable our brains to be plastic and go on learning each day precisely because it mainly downgrades neuronal connections at night.
When I reached Tononi by phone at home, opera was soaring in the background. Our waking consciousness is not a zero-sum game, he told me, in a rich Italian accent. By day, we mostly strengthen connections between brain cells—based on everything we see, hear, taste, touch, and do. But this process cannot go on indefinitely. Stronger connections (synapses) consume more energy, require more maintenance, and take up more physical space in the brain. By the end of the day, our heads are getting crowded, literally. Our capacity to learn is starting to max out. This may help explain why we feel so tired.
How does sleep renew us? During non-REM, Tononi proposes, our brain's connections overall scale down. By and large, connections that were strengthened during the day remain relatively more prominent at night's end. And those that were weak or went unused could well be erased. By morning, Tononi suggests, the brain is ready to learn more again. "Sleep is the price we pay for plasticity," he says. Sleep allows us to go on taking in new information, learning and adapting throughout our lives. The evolutionary assumption here, again, is that sleep is maladaptive: We need a good reason to do it, given the risks.
So far, the evidence for Tononi's hypothesis is mostly indirect.
And the work is limited in scope. It pertains only to the cortex, and it addresses only non-REM sleep. It's possible that REM is also involved in downscaling connections, albeit by a different mechanism. Or REM could build up connections, while non-REM provides a check by pruning some of the dross. It's too soon to say. But Tononi's hypothesis prompts more and better questions. It is consistent with the idea that sleep enhances memory but is not restricted to that purpose. Instead, it suggests that the fundamental function of sleep is to make it possible to start learning again every day—along the way offering one of the most appealing explanations yet as to why, at day's end, our heads can feel so crowded and jangled.
The scientific world has fallen in love with sleep again at the same time the rest of us are snapping up soporific mood music, room scents, and dessert drinks like Vanilla Van Winkle and Crème de la REM. As paper piles atop paper, it seems increasingly likely that sleep's ultimate function will turn out to be both grander and more basic than memory consolidation alone. The baby-cat experiments suggest that sleep helps rewire the brain after new experience. But we still don't know how sleep may tweak different circuits at different times and how this may affect our ability to think, feel, act, or make meaning.
Over the next 10 years, cellular, genetic, and behavioral work is likely to fill in the picture, telling us a lot more about whether sleep plays the same role in grown-ups as it does in babies, and how REM and non-REM work in tandem in their nightly roundabout. We may come to know whether parts of sleep are a neural rehearsal for waking, whether we are mostly strengthening brain wiring during sleep or mostly cleaning the closets. We may even come to understand why we go mad each night in our dreams.
Ancient myth held that the soul leaves the body during sleep and returns with arousal. Modern science seems to suggest the opposite—that we go deeper into our selves in slumber, only to reconnect to the world upon waking.