Marianne Freiberger reports on a discussion she had with Bernard Carr on whether or not multiverse theories are science. He has a suggestion for how we should classify these theories.
With the possibility for indirect evidence in the future, maybe we shouldn’t dismiss the multiverse as mere speculation, especially since it has many features that are theoretically attractive. So attractive that some have even suggested we change the criteria of science in order to accommodate it. “The key question is: how crucial is testability?,” says Carr. “My view is that it is crucial; you do have to be able to test a theory to make it science.” He advocates classifying ideas like the multiverse in a special category he calls meta-cosmology: outside the present boundary of science, but not on the far end of fiction. “It’s a sort of intermediate state, a state of purgatory, before you’ve decided whether [something] is proper science or not.”
“Meta-cosmology” seems like an obvious dance around the term “metaphysics”, a term physicists seem to hate having applied to any theories they discuss. But it seems like that label makes sense for speculations about unseen and untestable realms. Of course, accepting it means accepting that physicists engage in philosophy, as least to some extent. We should remember that many of today’s scientific concepts, such as atomism, began as metaphysical speculation.
Personally, if it makes cosmologists happier, I don’t see a problem with referring to multiverse theories as speculative science, provided that the “speculative” isn’t dropped. Along the lines of Carr’s reference to calls some have made to change the criteria for established science, I think doing that would, at a minimum, do damage to cosmology’s credibility.