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

Wiseman 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: Wiseman 07 November 2007 06:42:24PM -2 points [-]

No group selection? I believe the math in Eliezer's post is wrong. Here is how a hypothetical fox/rabbit population could evolve restrained breeding through group selection.

Picture a geographically isolated fox/rabbit population. At some level, this is guaranteed, simply because there's not an infinite amount of land on this planet to inhabit. Even if the entire planet was one continent with just rabbits and foxes, then that's the isolation geography. So at some point there won't be other foxes getting to eat the un-eated rabbits from the restrained fox population.

Start with a balance of rabbits and foxes. Perhaps this is because foxes are newly migrated to the area. Whatever. The foxes feast on the rabbits, because they are so easy to catch. The rabbit population drops to something that can metabolically support only 5% of the current fox population. As the foxes die out, any fox that has genes to restrain its breeding, is going to do better than foxes that don't, because it will spend less energy developing fox fetuses that won't survive anyway because there isn't enough food to go around. THIS IS THE KEY. If we assume all foxes are roughly equal at catching prey, then any fox family with constrained breeding will have more viable offspring because their mothers didn't die out trying to give birth to 6 foxes with food for only 2. Or the baby foxes, will go up stronger because food for 2 foxes is spread for just those 2, and not 6.

Now after this initial die off of foxes, the rabbit population will rebound. So won't the non-restrained breeders just take over again? No. As soon as the non-restrained breeders get large enough to diminish the rabbit population, the restrained breeders will have the same advantage they had the last time around. And even MORE of the unstrained breeders will die out removing even more of their genes from the gene pool.

Eventually, the non-restrained breeding genes become so rare it's as if they never existed. Only when they randomly pop up due to mutations would the cycle start again.

And here's where something almost magical happens. Every time those unrestrained breeders go crazy and eat all the rabbits, it does, to a certain extent, harm the survivability of the restrained breeders. Not as much as the unrestrained foxes, but enough. That means that any gene that will suppress the initial growth of an unrestrained fox population, will spread itself throughout the fox population. Perhaps a gene will arise that builds multiple chemical/hormonal systems in the fox to specifically restrain breeding, making it exceedingly difficult for any one mutation that un-restrains breeding to actually CAUSE unrestrained breeding.

Group selection. Tada.

Eliezer, just because the raw mechanics of evolution are very simple, doesn't mean bizarre and conceptually complicated things can't happen in the real world mechanics of evolution. Even if they SEEM counter-intuitive to the principles of evolution.

Comment author: waveman 15 December 2014 10:48:09AM *  5 points [-]

Something like this occurs with kangaroos, and some other species, which keep foetuses on the ready waiting for good times. They even re-absorb them when times are bad enough. But they breed very rapidly in good times and plagues regularly occur.

What seems to have evolved is an adaptive reproduction strategy, not group selected forbearance.

Comment author: RichardKennaway 15 December 2014 12:06:38PM 1 point [-]

Firstly, can you write all that in mathematics that behaves the way the words say? Words can be made to say anything, but mathematics is a more unyielding medium.

Secondly, there is no group selection here. You have described individual selection: individual foxes making decisions that give them individually a better chance of transmitting their genes to the next generation. That a particular (hypothetical) collective result is produced, that other people have invoked group selection to explain, does not make this group selection.