Via Matthew Yglesias, there is an article in Scientific American on the support for multiverse hypotheses in cosmology. The most striking feature of the article is the claim that the multiverse hypothesis (allegedly) corresponding to Lewisian modal realism has testable consequences. I think this is either a misuse of ‘testable’ or a philosophical mistake akin to the argument that modal realism leads to inductive scepticism. But whatever the causes, it is nice to see Lewisian metaphysics being taken seriously.