Less Wrong is a community blog devoted to refining the art of human rationality. Please visit our About page for more information.

J_Thomas comments on The Tragedy of Group Selectionism - Less Wrong

36 Post author: Eliezer_Yudkowsky 07 November 2007 07:47AM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Comments (88)

Sort By: Old

You are viewing a single comment's thread.

Comment author: J_Thomas 10 November 2007 08:51:14PM 0 points [-]

'Would "innovation" in genetic error correction, or changes to the proteins responsible for allowing greater or fewer mutations in DNA...'

'...would such "meta-changes" (changes to the mechanisms of DNA replication) be the basis for group selection?'

If they can't interbreed, then you get selection like that between two different clones of bacteria. Either the better species survives, or they both survive in their own ecological niches.

If they can interbreed then you might get evolution by group selection but it isn't the way to bet. You'd want a specific case for a particular gene.

'If not this, how did "innovations" to DNA error correction and selection for the different rules about how many mutations to allow in DNA copying even form in the first place?'

Given alternative genes that result in alternative behaviors (either from different enzymes or different regulation of those enzymes or something else) -- the one that works better results in its individuals outcompeting other individuals. That's what natural selection involves. The way you know it's fitter is that the gene frequency increases. We hope that on average things that increase gene frequency also improve survival of the individuals that carry them, and improve survival of the population. But there are examples otherwise.

Sometimes genes can increase because of selection among groups. This can happen but it's a complication that tends not to happen.

How does a mutation in a gene for mutation rate get changed in a population? When it starts out it's outnumbered a hundred million to one. The first favorable mutation that happens will almost certainly be among one of those hundred million and not in the single mutant.

And further, given mutations that are all about as good, the first one to get established gets the lion's share of the results. However, in a large population the mutation that makes a better mutation rate will happen occasionally in individuals that have a favorable mutation, and will spread with them. In the absence of selection against it, it will reach a fluctuating equilibrium for that reason. And this subpopulation will mutate at a better rate, and the result is that it will increase some with each population changeover. With each changeover the better mutation-rate mutants will tend to increase at the expense of the worse mutation-rate variants. Slow but reasonably sure, given a large population.

It's possible for genes to evolve that regulate things like mutation rate. They might increase the rate or decrease it according to whatever cues seem to work well on average. Selection can work to let populations evolve faster, and it's selection on indvidual genes (or combinations of genes) that does it.

It's possible to get genes that improve the survival of groups. But unless they spread they have limited chance to improve group survival.