Icosian Reflections

…a tendency to systematize and a keen sense

that we live in a broken world.

[Polyphasic Sleep] Errata, Food, Reductionism


It's become apparent to me in the past few days that, when I wrote last week's post on the theory of sleep, I didn't know what I was talking about. Or rather, I had an incomplete picture of the subject at hand, and oversimplified a system that was more complicated than I was giving it credit for.

A little more digging has revealed that there is, in fact, evidence that SWS may be important for consolidating declarative memory, which breaks down into episodic memory (events) and semantic memory (facts). By contrast, REM consolidates procedural memory and spatial memory (both of which are more or less exactly what they sound like). Previously, I had dismissed it as "useless". Oops.

Of course, that's not the whole story, either: non-REM sleep is split into periods of "light sleep" and periods of SWS. Altogether, an 8-hour monophasic sleep cycle includes something like 1-2 hours of REM, 1-2 hours of SWS, and 4-5 hours of NREM2 light-sleep. The goal of the E3 schedule, then, isn't so much to absolutely reduce needless SWS, but to restructure the sleep cycles to reduce NREM2, at minimal cost to SWS/REM. (Note that E3 differs from other polyphasic schedules in this respect; notably, Uberman almost entirely eliminates SWS, and packs two hours of REM into just six short naps.)

This is me, not drawing a diagram which may later turn out to be wrong. At least I've learned that lesson.

Now, it makes intuitive sense that, if the brain is using SWS for consolidating certain types of memory, there's an additional "rest state" where it's actually doing nothing but resting. And so, my initial analysis (re: we wear ourselves out less today than in the ancestral environment) seems to apply just as well to cutting out NREM2 light-sleep as it did when I was talking about "SWS resting-sleep". There's still a lot of "resting" that I don't need to do, if I'm going to spend most of my day sitting in various chairs.

Independently, I also believe that it's probably worth giving up some SWS to ensure I'm getting a full complement of REM cycles daily. At the point where I am in life (i.e. formative college years), it seems more important to be learning how to do things rather than amassing a brain-full of facts and for this, I'm going to need more procedural memory than declarative. Note that this has nothing to do with the above understanding of NREM actually being correct; ideally, I'm only cutting NREM2. But it's good to know that, even if I'm wrong and I'm cutting into SWS, it's not the worst thing in the world.


Of course, there's a bigger problem here. It's no small matter that I attempted to describe a complete view of "how sleep works" and failed. "How do you know" asks my inner skeptic, "that you're not horribly wrong this time, too?"

And that little voice has a point. It's certainly true that my philosophy on nutrition is very adamantly anti-reductionist. Why? Well...

A (Short) Parable on Nutrition

People realized that if you didn't eat food, you died very quickly. So they concluded that you needed calories to be healthy.Later, they found that there were different kinds of calories: proteins, fats, carbohydrates. If they were imbalanced, you had problems. These took longer to manifest than starvation, but were definitely bad for you. So they concluded that macronutrients were important.

But then, of course, sailors began to die of scurvy. And, digging deeper, people eventually found micronutrients -- vitamins, minerals, and other things that make it impossible to eat {bread, meat, sugar} and be fully healthy. These health problems took longer still to manifest than macronutrient-imbalace issues, but they were still definitely there. And so they began tracking micronutrients as well.

Since then, we've continued to find more things that your body needs to survive, with longer timeframes on the bad-stuff-happens-without impacts: Omega-3 fatty acids, phenols, antioxidants...Why would we think that right now we understand it all?

In fact, I'm doubtful that, in the next hundred years, we will come close to "fully understanding" nutrition. And I certainly believe that if you mixed together pure versions of everything modern nutritional understanding says is required for health, you'd still be leaving some important things out, and would see negative health effects on the scale of at most 20-30 years. I'd even lay money on it.

So I stay away from reductionist nutrition. I try to eat closer to traditional diets when possible, following the conventional wisdom that variety is good for you, fat/protein/carbs are all essential, vitamins are no substitute for vegetables, and so on. My hope is that this approaches the diet of our evolutionary ancestors, which is a lot closer to what my body actually needs than a nutrition-facts, daily-recommended-values report can describe.


Okay, so why am I accepting (even pursuing) a reductionist model of sleep? If there's something we understand less than the human gut, it's the human brain. And I've certainly been wrong once (publicly) about what my body does and doesn't need in terms of sleep...what's to say I won't be wrong again?

It's really hard to say. A lot of it has a lot to do with the fact that the potential benefits of a polyphasic success are just so much larger than a food-reductionist success. (I maintain that even if you could reduce nutrition to an all-in-one shake, I still would turn it down, because food makes me fundamentally happy. By contrast, 25% more waking-hours and extra REM cycles would be a really big deal for me, and is worth far more to me than sleeping in a few hours longer.) And so I'm more willing to take a risk.

Additionally, I'm more confident in my ability to self-diagnose neurological failure modes than nutritional ones. I've been able to test basic mental function, and observed no significant changes from baseline to E3. (Though I have good reason to believe that these tests are bunk, anyway.) My psychomotor vigilance has remained at baseline, as has my general mindfulness (as far as I can tell). I've been dreaming more vividly, and have not been feeling unusually tired during daytime hours (the 4am-8am stretch still kills me, though). I understand the sensations of sleep-deprivation and SWS-deprivation, and I'm sure I'm experiencing either. In my social interactions with people, they've stopped telling me that I'm acting strangely. In short, many perspectives appear to indicate success.

I'm not a neuroscientist, but I interact regularly with a community of rationalists who are really into (1) properly (or better-than-properly) functioning brains and (2) not dying. I trust them to do their research. And their research is what I'm following, mostly. And furthermore, they/I might well be wrong about any number of things re: sleep neurology and still be coming out ahead (cf. the end of (1) above). Of course there are still dangers, but I believe that they are less likely to be unexpected, invisible, or ignored, and so they bother me less.

I can't imagine collecting a commensurate amount of test data on my nutritional state. I don't know any community of nutritionalists that I trust to think (1) rationally and (2) unconventionally about nutrition. (Someday, I'll explain exactly what I mean by "rationally". Unless you'd describe yourself as a rationalist, it's probably not what you think I mean.) And, as noted above, the benefits of a reductionist diet just aren't all that appealing, so I have little to gain and even less to balance against unexpected costs.

So, if there's a tl;dr to this post, it's this: Reductionism is hard, because sometimes we get it wrong. Applying it to vital aspects of our lives is not something we should do lightly, without serious caution, and without serious potential for large benefit. Here, defense-in-depth is very much your friend. And Ross really likes food. Seriously, if you want to get on his good side, buying him food is always a great bet.