Re: Article: Group selection, a theory whose time has come...again

John Harshman wrote:

dkomo wrote:


Cj wrote:



"dkomo" <dkomo871@xxxxxxxxxxx> wrote in message news:AtOdnS7OK6i-28ranZ2dnUVZ_hSdnZ2d@xxxxxxxxxxxxxx



John Harshman wrote:



dkomo wrote:





John Harshman wrote:





dkomo wrote:







John Harshman wrote:



<snip top>

So
cheaters will arise at a much greater rate than will altruistic groups,
for example, making it unlikely for any altruistic group to last long
enough to achieve "fixation".

Not if the individuals of a group are tightly coupled. Then each individual's fitness depends on the fitnesses of the other members of the group. Cheaters will decrease the fitness of other individuals while attempting to enhance their own, so in the end the fitness of the cheater will decrease as well, hence cheaters will be selected against by group selection.


I'm afraid you don't understand natural selection.

I spent a lot of time studying natural selection et al and recently became disillusioned. I concluded that NeoDarwinism just doesn't have the horsepower to explain fully the complexity of life. More powerful theoretical models are needed as an extension to NeoDarwinism.


So you claim. But I have seen no sign of any such need or of any such
models in any of your posts so far.





So don't conclude that because I don't automatically buy into the usual kin selection and reciprocal altruism models contra group selection that I "don't understand natural selection."


No, that's not why I think you don't understand natural selection. It's
because your paragraph above (and the beaver example that used to go
along with it) shows you don't understand natural selection. It's just a
collection of non sequiturs. If cheating is disadvantageous for an
individual, then natural selection explains cooperation without recourse
to group selection. If cheating is advantageous, then natural selection
is expected to promote cheating, and you would have to explain what
prevents that.


The beaver example was taken from Sterelny and Griffiths, _Sex and Death -- An Introduction to the Philosophy of Biology_. In chapter 8 "Organisms, Groups and Superorganisms" there is a fairly long and detailed discussion of group selection and the beaver example is used in various spots. For example,

"What makes an organism an interactor? Wilson and Sober define an interactor as an entity whose parts share a *common fate*, rather than one whose parts form a complex organization. An organism is more than a population of cells, not because those cells form a complex system of organs, but because those cells share a common fate. Their reproductive fate is locked together on a single causal trajectory. Similarly, a group of beavers in a lodge is an interactor if their fitness is linked together on a common causal trajectory. Beaver traits that affect that trajectory for better or worse can be visible to selection through the fate of that beaver collective...Wilson and Sober propose that common fate is defined on a trait-by-trait basis. If the beavers cooperate in the construction and maintenance of their dam and lodge, then that characteristic will have a common effect on all the beavers in the collective."

"Thus D. S. Wilson introduced and defended the idea of trait groups. Trait groups are groups of organisms, each of which feels the influence of the others with respect to some trait. If the trait is dam building, the trait group is the group of beavers that live and shelter behind the dam."



Beavers (Castor canadensis) are a particularly poor choice of animal groups for exploring group selection since all the members of a beaver "colony" are related. The beavers in a group are a mated pair and their offspring. When the offspring are relatively mature the parents drive them out of the lodge and from the beaver pond itself. In fenced experiments young that cannot escape their natal pond are attacked and killed by the adult beavers (their parents). Only after the dispersal or death of the offspring will the beavers breed again. At best a beaver group is representative of kin selection and that is seasonal, when the offspring are fairly well grown there is extraordinary kin aggression. What appears to be a simple model can actually be a game theory nightmare for analysis.

Interesting. I think the use of the beaver example in the quote I provided was only meant to illustrate what a trait group was. The beavers share a common fate through dam building. Lazy beavers (cheaters) who let others build and maintain dams can decrease the fitness of the entire group, and end up less fit themselves.

If the beaver example was presented as *evidence* of actual group selection, then the details you gave would become important. I'm not sure if it was Wilson or Sober who came up with the beaver example, or if it was Sterelny and Griffiths attempting to illustrate Wilson or Sober's ideas.

However, trait groups are general and don't just apply to beaver groups. They are found throughout the plant and animal world.


May be, but the beaver example just turns out to be kin selection, and
so reducible to individual selection. Can you produce a real-world
example that can't be reduced to individual selection?


LOL, Wilson says that kin selection is nothing more than a instance of group selection:

"In Wilson's account of kin selection, kin groups are the interactors, so kin selection is a variant of group selection, not an alternative to it. Wilson and Sober treat reciprocal altruism in the same way."

Sterelny and Griffiths, p. 163

So there you are, if kin groups are the interactors, then they are the basic unit of selection, and not individual organisms.


--dkomo@xxxxxxxx




.

Relevant Pages

  • Re: Article: Group selection, a theory whose time has come...again
    ... I'm afraid you don't understand natural selection. ... kin selection and reciprocal altruism models contra group selection that ... because your paragraph above (and the beaver example that used to go ... "Thus D. S. Wilson introduced and defended the idea of trait groups. ...
    (talk.origins)
  • Re: Article: Group selection, a theory whose time has come...again
    ... Cheaters will decrease the fitness of other individuals ... I'm afraid you don't understand natural selection. ... because your paragraph above (and the beaver example that used to go ... trait groups are general and don't just apply to beaver groups. ...
    (talk.origins)
  • Re: Article: Group selection, a theory whose time has come...again
    ... Cheaters will decrease the fitness of other individuals ... I'm afraid you don't understand natural selection. ... because your paragraph above (and the beaver example that used to go ... May be, but the beaver example just turns out to be kin selection, and ...
    (talk.origins)
  • Re: Article: Group selection, a theory whose time has come...again
    ... Cheaters will decrease the fitness of other individuals while attempting to enhance their own, so in the end the fitness of the cheater will decrease as well, hence cheaters will be selected against by group selection. ... Groups and Superorganisms" there is a fairly long and detailed discussion of group selection and the beaver example is used in various spots. ... Trait groups are groups of organisms, each of which feels the influence of the others with respect to some trait. ...
    (talk.origins)
  • Re: Co-evolution
    ... nor kin selection suffice to discourage 'cheaters'. ... may still be needed, though, to encourage the punishment of cheaters. ... I take a different lesson from the review. ...
    (talk.origins)