on sleep and consciousness

There are a couple of areas of research in the cognitive sciences that are so mysterious and philosophical that they tend to be the topics of wild speculation when scientists get drunk at the hotel bar at conferences.

One area is of course the nature of consciousness. Consciousness is really really hard to study, since you can't see it or measure it. Lots of the interesting work on consciousness is written by philosophers who navel-gaze until they come up with something new. A few neuroscientists study consciousness, such as Antonio Damasio, looking at how sensation and emotion and consciousness interact in people with brain damage. Avoid any speculation on consciousness by mathematicians or physicists -- big time crackpots.

Another fun area of speculation is sleep and dreaming. We know a bit more about this. For example, there are fairly plausible arguments, with empirical support, that say that dreams are basically a side-effect of a process wherein things that you learned the previous day get moved from shorter-term memory stores in the hippocampus (in the middle of your brain) into your long term memory stores in the cortex (the surface of your brain). But there are lots of ways that we don't understand sleep at all. For example, although the brain is quite active during sleep, we're pretty remarkably unconscious.

A new study published in Science this week, by researchers at the University of Wisconsin (my alma matter, of course!), describes a really straightforward experiment that hints at the nature of sleep, and perhaps even consciousness.

They used two techonologies. One is EEG, which is where they put a bunch of electrodes against your scalp and record the electrical fields from your brain. They can kinda tell where there's electrical activity, although not very accurately. The other technology is called Transcranial Magnetic Stimulation (TMS), which always seemed to me a little scary, although it's supposedly pretty safe... They stick an electromagnet next to your scalp and pulse a magnetic field in such a way that causes electrical signals to be created in a particular spot in the brain. The neurons in your brain communicate primarily with electrical signals, of course, so this technique sort of disrupts everything in a small area for a fraction of a second. Kinda like a tiny seizure...!

The researchers from the UW tried this on people while they were awake, and then later when they were asleep (but not dreaming). When the subjects were awake, the electrical impulse bounced around from the original location of the stimulation, to a few other locations in the brain, for a period of a couple tenths of a second. That is, the parts of the brain were communicating with each other, as they normally do. When the subjects were asleep, however, the impulse didn't spread at all, but spiked right under the electromagnet and then died away quickly. It's as if that section of the cortex wasn't communicating with other areas of the brain. Any signal in that area stayed in that area without being shared.

This is a very cool result. It suggests that sleep isn't so much a process where nothing is happening in your brain, as a state where the bits of the brain are each doing their own thing, not communicating with each other. And the reason you're unconscious when you're asleep is that there's no coherent single "you" that's made up of the integrated signals from all over your brain. When you're awake, the parts talk to each other, sharing information, building up the different experiences that make you feel awake. When you're asleep, the parts are isolated (how?!), doing their own thing (what?!).

Of course, this is just one result, and I'll be very very surprised if we have good understandings of how consciousness or sleep work by the time I retire in 40 years or so. But still, this is to me a really interesting piece of the puzzle.