Friday, 3 October 2014

Confusion about the topic in the popular press

The popular presentation of the kin selection vs group selection affair still seems to be very confused. For example, consider the coverage of a recent study by some group selection advocates: "Elusive Form of Evolution Seen in Spiders". The article says:

According to one model, known as kin selection, highly related organisms such as bees and ants can develop altruistic behavior — for example, many females forgo reproduction in order to raise the queen’s brood — because they will still pass down their genes indirectly, through the queen. But despite its altruistic appearance, kin selection is selfish — it helps an individual’s genes to survive. Can natural selection promote truly unselfish traits, behaviors that are good for the group, but not necessarily to the benefit of individuals (or their immediate kin)? Some evolutionary models predict that it can [...]

The rest of the article is all about group selection. Can kin selection be selfish while group selection is altruistic? Not according to the modern scientific consensus on the topic - that has kin selection and group selection being different accounting methods that attempt to account for the way in which genes propagate in viscous populations. Modern versions of these theories make the same predictions.

Originally, group selection was widely imagined as taking place between demes. However Williams, Maynard Smith, Dawkins and many others pointed out that this sort of "group selection" didn't work very well - individual-level adaptations swamped the group-level ones. Much later, the "groups" of group selection were re-imagined as involving any kind of social cohesion, including - critically - family groups consisting of mothers and their offspring - or groups of siblings. Individuals are then modeled as being part of an enormous number of partly-overlapping groups - including, critically, close family groups. With this radical readjustment, this new form of group selection - if properly applied - reproduces the predictions of kin selection exactly.

Most group selection advocates have now (finally) come on board with the scientific consensus on kin selection that was established in the 1970s. They recognize they these two theories - if properly applied - produce the same predictions and results. However, you wouldn't guess at this from this popular article. It's promoting the idea of group selection as a novel, revolutionary theory. This conception of group selection is not scientifically accurate.

Another article promoting the study has a different take on it: group selection vs gene selection. Their champions of "gene selection" are Pinker and Dawkins. It should really be group selection vs kin selection. Genes (in the broad sense of heritable information) underlie both.

More coverage:


Update 2014-10-29: Goodnight has weighed in on his spider study on his blog. Kin selection and equivalence are apparently unmentionable topics. To me this seems like a parallel universe.

No comments:

Post a Comment