Hacker News Reader

Comment by jauco | Hacker News Reader

> It would be the biggest evolutionary mistake if sleep does not serve some critical function," Walker says.

it's a minor quote from the whole article, but sentences like these always annoy me: Evolution does not make mistakes! It's not an It! it doesn't do!

Evolution's simply a term that describes change. You should treat it like the term "moving". As in "Creature a moved to B" <=> "Creature A evolved to B" how silly would it be to say that movement made a mistake.

And even if we forget syntax and focus on what the sentence implies the sentence is overly bold. If we sleep, then somewhere along our ancestry, creatures that slept survived. That's about all you can say. Maybe sleepers where unfit for life if strange virus X hadn't wiped out all non-sleeping superior lifeforms. Or whatever. Don't infer based on the fact that something is as it is, that it is therefore superior to all other possibilities.

View on HN

I agree that the "mistake" phrasing is atrocious.

I don't exactly agree with this, though:

Don't infer based on the fact that something is as it is, that it is therefore superior to all other possibilities.

Actually, we can conclude that there must be a good reason why the average human needs eight hours of sleep, as opposed to four hours or two hours or half an hour. It's not as if we can't get there from here: Some living humans really do need only four hours of sleep. And other creatures, not so different from us, have much niftier tricks: Bears can sleep for four months straight, horses can sleep very lightly while standing up, my parrot wakes up at almost any stimulus, some creatures sleep with half the brain at once, etc. If sleep were a big liability, there seem to be lots of genetic switches that could be thrown to reduce, alter, or eliminate it. The fact that those switches have not been thrown in us suggests that our sleep pattern is an equilibrium point in evolutionary space.

Now, maybe you're trying to say that, although there might be a way to design a brain that didn't need sleep -- perhaps scientists will use futuristic neuron-CAD software to figure that out, someday -- some accident in early evolution sent us down this particular branch, and now we can't get there from here. That's true, and it's really fascinating. It seems that creatures without sleep would have to evolve nearly from scratch -- because even fruit flies sleep (a fact that I didn't know until today) which suggests that you have to go back past our common ancestor with the insects to find a creature that wasn't designed around sleep. And if such a creature with a prototype sleep-free brain tried to start up shop today, odds are it or its descendents would all be eaten by fish before they got very far. It's hard to evict species from an ecological niche, particularly when you're going up against nearly every animal species on earth.

Actually, we can conclude that there must be a good reason why the average human needs eight hours of sleep, as opposed to four hours or two hours or half an hour.

Now, maybe you're trying to say that some accident in early evolution sent us down this particular branch, and now we can't get there from here. That's true, and it's really fascinating.

That was indeed the idea behind my previous post. Rather the same thing happens with humans only changing teeth once a lifetime instead of continuously. There is some evidence that changing once doesn't give us an evolutionary advantage, but is simply a leftover from some ancestor's trait way back. (I'll see if I can find the reference).

I'm not saying that humans might not need their sleep (we die if we are sleep deprived), but that, based on the fact that we sleep, you can't claim that sleeping creatures have an advantage over hypothetical non-sleeping creatures. Maybe it's just something that stuck and doesn't do enough harm to kill us all out.

The sentence may be a little ill-formed, but the general point is true. There are genes like the ones that cause Tay Sachs disease and sickle cell anemia, that have genetic advantages, like resistance to TB or malaria. But in the absence of the selecting force, such as with modern day medicine, the frequency of those genes in populations decrease over generations, which is the definition of evolution. Sleeping seems like a huge disadvantage over a creature that could stay active and hunt 24 hours a day^, but if there were no selective force for creatures that sleep, creatures would evolve that don't require sleep pretty quickly. I'm not sure how fast, but the change would be observable over a couple of generations.

^(This advantage is not necessarily true in all cases. Bears sleep because it's not energy efficient to forage in the winter, humans may have evolved sleep as a way to save energy during the night because electric lighting became popular. However, I really doubt this hypothesis because of the number of critical functions that degrade under sleep, as well as the increased risk of heart problems from people who don't sleep a lot, which is really independent of whether you feel tired or not.)

On an unrelated note, how do you escape asterisks?

My reply to mechanical_fish sort of covers your argument as well I think.

you can type a solitary * by placing a space after it

Well here is evolution making a "mistake":

http://news.ycombinator.com/item?id=143251

Reply on news.ycombinator.com