Hacker News Reader

Comment by TangoTrotFox | Hacker News Reader

I think there's a related issue here that's always kind of nagged at me.

We have evolved many characteristics which are handy, but which seem unlikely to have had enough of an impact on survival to have become so ubiquitously propagated by natural selection alone. For instance the ability to digest milk as well as white skin are both relatively new adaptations. But the advantages they offer, for the locations where they evolved, are pretty mild.

Fair skin could have had sexual selection factors in play helping its spread, but being able to digest the milk of lactating animals? It just doesn't seem like a mutation that would provide enough benefit to become ubiquitous through the mechanisms of natural selection alone, yet it did. And both of the above evolutions are relatively recent, making it all the more difficult to reconcile things through random mutations filtered by natural/sexual selection alone.

View on HN

The key is sexual reproduction. You're not selecting for a single trait at a time, you're selecting for all of them. Yes, the ability to digest milk as an adult was unlikely to have contributed to the reproductive success of the very first individual it arose in. But it didn't need to; it only had to not prevent reproductive success. Over a few generations the gene spread to dozens, then hundreds of individuals, sheerly from being along for the ride. At that point, statistics can begin to take over - even a tiny advantage will gain a foothold if given enough people to prove itself over.

Disclosure: I'm fairly dubious that we as a modern discourse fully understand the reality of life, and propagation, shaping of life.

I have a lot of reservations. And I often times find myself rather disgusted by the attitude and approach of many people in the scientific worldview.

That being said, I do believe there are some interesting numbers out there about how much more efficient dairy is than just simply eating your animals. It also produces a relatively large amount of fat, which is is incredibly important (for your brain, and much else) and is one of the more difficult parts of a diet to attain in most ecosystems.

There could possibly also be advantages related to the ease of lactobacillus fermentation.

I'm confused by your viewpoint. How is a fuller ability to extract nutritionally-complete food from your environment in winter not an advantage you'd expect to be selected-for?

People who are malnourished produce poorer sperm. People who are malnourished are not looked on as good mates who can bear children or provide food for the family. People who are malnourished die earlier, perhaps starving to death before reproducing (or reproducing as much).

Think about your usage of the word word people - as opposed to the singular. In a purely random natural selection driven system, propagation relies on the success of a single mutated individual until we get to very large populations where randomness could more regularly yield similar mutations among different people. Because of this, in the purely random natural selection system the magnitude of an effect is extremely important. Let's imagine we can somehow quantify exactly how likely a specimen is to survive in a certain area per year, we'll call that value x. Imagine x is some percent per year. How much would a random specimen suddenly developing lactose tolerance increase that percent?

In the case of something like Darwin's birds, the percent improvement from beak changes would be very substantial. That mutation would open up entirely new, and vast, resources of food available at all times and completely independent of everything else. In the case of milk, you need a lactating animal to start with. That already implies you have multiple sources of food in extremely healthy shape. That little catch there already ensures that this mutation would not provide a substantial benefit in a one-off scenario.

By contrast, imagine our mutations are not entirely random. And that our bodies, somehow, trend towards more desirable mutations over time -- meaning we don't see just a one-off instance of favorable mutations but them tending to occur over and over again. In this case your language of "people" would be entirely appropriate. Over time, even a small survival edge means a lot and, given enough samples, we'd certainly expect this trait (and any trait for that matter) to start to 'stick' in any area where its net effect on survival was basically anything above 0.

I'm not seeing the issue here. As you say, let's talk about this at the individual level: picture yourself as a prehistoric pastoralist raising cattle. They eat plants you can't digest, and from them you get milk and meat. Everybody knows that you can't just drink the milk raw -- the lactose in there will really do a number on your digestive system -- but processed milk products like cheese make up a major part of your diet.

... Except that you can digest the lactose. You don't feel sick after drinking milk, because you have a mutation that breaks the normal mammalian mechanism for reducing lactase production in adulthood. Sounds nice, right? But maybe not something that'll bring a huge reproductive advantage to you and yours; more of a cool party trick to show off to your neighbors.

And then one year there's not nearly enough rain. Famines have always been pretty common, alas, and here's another one. Now consider this: processing the milk so lactose-intolerant people can digest it loses about 40% of the calories. You can drink it straight because of your freaky mutant milk-drinking powers, and your neighbors can't. You get a lot more food at a time when people are dying from lack of food. Is this sounding like an evolutionary advantage? The sort of thing you might pass on to the kids you'll have because you didn't starve?

The ability to digest lactose as an adult is potentially a really big deal if you happen to live in a culture that already uses processed milk products as a major source of calories.

Dairy products would not have made up a substantial portion of a regions diet until lactose tolerance had emerged. Lactose intolerance extends to all dairy products, not just milk. And again, you need lactating animals to produce consistent dairy products. Animals do not naturally lactate unless they're feeding their young. And animals also do not naturally produce the freakish amount of milk that we're able to extract now a days from millennia of artificial selection and, in the US at least, using a lifelong cocktail of all sorts of different drugs.

To be clear, I do agree with you that in certain uncommon scenarios lactose tolerance could provide a tangible, though not enormous, advantage. And that's the problem. A different shaped beak for a bird entails a never-ending constant, and enormous, advantage. Lactose tolerance is something that's pretty irrelevant by comparison, yet it would become ubiquitous.

In a purely random natural selection driven system, propagation relies on the success of a single mutated individual

That is flatly incorrect. Whoever convinced you this is true has a fundamental misunderstanding of natural selection.

Any time environmental pressures change such that a particular phenotypical expression leads to more reproductive success, that is natural selection. A mutation is utterly unnecessary.

You have the context of what you're responding to right above you. The discussion is about mutations. How a single mutation would propagate in cases where mutations were entirely random as opposed to those where mutations were, in some way, less than random.

Reply on news.ycombinator.com