[gyom]

The spiders present a puzzle to evolutionary biologists. According to ordinary Darwinian natural selection, only the fittest individuals will pass on their genes. But if that’s the case, why do tangle-web spiders act in ways that might conflict with an individual’s drive to outcompete its neighbors? A spider that defends the nest might put itself at personal risk, jeopardizing its chances of producing offspring. And a spider that rears the young might have to wait to eat until the hunters are sated, so it might go hungry. These are not behaviors that would be expected to enhance an individual’s fitness.

Which is the whole point of "The Selfish Gene". It's not about individuals; it's about their genes. Genes win when their bearer helps out other individuals sharing the same genes.

[pesenti]

No, this article does go against the theories in "The Selfish Gene" (which only predict kin selection). See the whole row between Dawkins and Wilson over multi-level selection: http://www.theguardian.com/science/2012/jun/24/battle-of-the...

[scott_s]

To me, the key point of "The Selfish Gene" was that the base unit of evolution is the gene. I don't see anything in this article which disagrees with this notion, and they did quote a biologist who said something similar:

Pruitt and Goodnight don’t propose a mechanism by which the colonies boomeranged back to their original state. And without such a mechanism, some researchers argue that the results could be due to ordinary selection acting on individuals. “I think they over-interpret [the results] as evidence of group-level selection,” said Andy Gardner, a biologist at the University of Oxford. “Natural selection may factor in the needs of the group, to some extent, when it hones the adaptations of individuals. But group fitness is not the whole story.”

I understand Gardner's point to be that even if the selection pressure starts at the group level, it can still go down to the individual, and then down to the gene. In that sense, we're still selecting the gene.

Actually, upon reading the actual article, the authors say basically the same thing in the first paragraph:

"In societies in which individual fitness is tightly linked with the performance of the group, the theory of group selection predicts that evolution will favour traits in individuals that aid in maximizing their group’s success—which, in turn, are predicted to increase individuals’ long-term evolutionary interests. Here we define group selection as selection caused by the differential extinction or proliferation of groups1. This represents a broad definition that is not in any way adversarial to the importance of kinship selection for social evolution."

Assuming I understand the authors correctly, they mean that the "fitness" of an individual cannot always be considered to be in isolation. In order to determine the fitness of an individual, we may have to consider the combination of the individual, and the group in which they live. I see that as consistent with the notion of a "selfish gene", as I understand it.

[Luc]

> Which is the whole point of "The Selfish Gene".

It's one of my favourite books, but I wonder if it had much of a lasting influence academically. It seems this article doesn't even touch upon the selfish gene viewpoint.

[azakai]

It has, or rather, it reflected the strong trend.

Aside from a few prominent figures like E.O. Wilson, I believe that gene-level selection is accepted as the norm. And group-level selection has been almost completely out of favor for generations - gene-level selection killed it. That is why this article is interesting, it might show a new shift.