[komali2]

If you could find that article, I'd be very interested in reading. I ask because I'm quite skeptical of the concept of humans "evolving" different time-preferences. Mostly because of 2 questions: "How was that selected for," and, "given that very little has changed about humans through evolution in the last 10,000 years, what sort of timeline did the article suggest this 'time-preference' evolution occured over?"

[dmm10]

Boyd Eaton and his students did some of this work. Its touched on in The Paleolithic Prescription (popular book) with more details in papers before and since (search on book's authors' names for research papers. Sorry I don't have better for you now.) As I recall they identified three sleep patterns from anthropological studies of populations not influenced by modern technologies (eg. electric lighting) and from historical documents (ex. Victorian era doctors' notes mentioning what were considered normal sleep patterns at the time.) 1) Early to sleep + early to rise 2) Late to sleep + late to rise 3) Late to sleep + late to rise (with waking in the middle of the night)

A modern cartoon characterization of type three is Dagwood Bumstead from the comic strip Blondie. He would get up naturally for midnight snacks and over-sleep for work or snooze during work. Type three Victorians were noted by their physicials as naturally rising for midnight prayers or sex with spouses.

As to genetic selection, the postulated survival advantage relates to the military three watches of the night. With group members naturally awake throughout the night predators were better protected against, surprise attacks less likely, etc.. Hence, a mixed group consisting of all three types would have a survival advantage.

[kolinko]

I'd love to see the article as well.

Having said that, from a genetic standpoint it's possible to have it so that just certain stable percentage of population has a given trait.

How can it be selected for - be eliminating whole villages/societies that have a wrong ratio.

Imagine you need 5% of population to be crazy risk takers with no survival instinct. Any more and a village collapses, but any less and you won't have anyone take certian risks when necessary.

Now, let's assume that the gene is encoded in mitochondrial dna (carried from mother to child, no influence from father). Also, nobody can really tell who has the gene uless a disaster strikes and such a gene is very helpful for survival or reproduction. So if a generation T has X% of this gene, T+1 will most likely have X% too.

Originally, nobody has the gene. So all the societies kind of manage to go by without it. Then one woman has it by random mutation. Her is now 1%, others are 0%.

But then a disaster strikes. And villages with 0% die off. 1% villages die off too, but not as much. After the disaster they are now the majority.

In one of the villages, a woman with the gene, by pure chance) had more children than usual. In that village the ratio is 5% now.

Disaster strikes again, and only villages with 5% are left.

Villages where the ratio increases above 5% through chance again, die because the chaos ensues. Villages with lover ratios are decimated once in a while.

One day people invent ways to protect themselves from the disaster, and ways to cope with above 5% ratio too. Then the gene will begin slowly drifting in one direction or another. But it will take generations for proportions to change significantly, and it's not certain if the geme will disappear or dominate the population ultimately.

I'm not syaing that this is the case with sleep, but that's the wyanit might work.

[arctangent]

I don't know which article thetli8 is referring to, but this might be a good place to start an investigation:

https://en.wikipedia.org/wiki/Chronotype

> I'm quite skeptical of the concept of humans "evolving" different time-preferences

As described in the wiki link above, the differences in "time-preference" are basically down to natural variance - I don't believe there's any evidence of a bi-modal distribution.

Nature does seem to select for anti-fragile populations (because by definition they are robust) and one way of achieving this is to have a certain amount of variability in phenotypes.

[thetli8]

Not sure where the article is but this ASAPScience video gave a quick run-through relative to what I was talking about:

https://youtu.be/BPJ0729NVjw