Why does cosmology need philosophy?

[C C]

Why does cosmology need philosophy?

[...] Cosmologists have developed the multiverse theory for good reasons (see this Plus article). But as Ellis says: "There's absolutely no direct evidence and it never will be possible to get direct evidence. If you can't test the theory you have to say, 'Is this now a scientific proposal or a philosophical proposal?' In my view this is scientifically inspired philosophy, not science, because I think you should draw a hard line."

In view of this untestability some people have suggested that we should weaken the requirements of science. 'If we have a really strong theoretical argument, [people have suggested] we should say, 'It's so good, we no longer need to test it in the way we've taken for granted up to now,'"explains Ellis. "I think that's very dangerous and I think it would allow all sorts of pseudo-sciences to be reclassified as science. I don't think we want to see that happen." [...]

[Yazata]

If you can't test the theory you have to say, 'Is this now a scientific proposal or a philosophical proposal?' In my view this is scientifically inspired philosophy, not science, because I think you should draw a hard line."

I don't think that it's always possible to draw a hard line between science and philosophy. The expectation that scientific assertions must be testable is a philosophical requirement, after all.

Most of us will agree that a scientific theory does have to be testable. But one can argue that the testability requirement needn't suggest that every implication of the theory be subject to testing. It's possible to imagine theories that have been successfully tested in many different ways, still having some additional implications that currently are (and may in principle always be) untestable.

If a theory successfully withstands many different modes of testing, its success may well justify our belief that its as-yet untestable implications are more likely than they would have seemed without the theoretical basis.

But yeah. If no means exist to test whether or not the more exotic implications of otherwise well-founded theories hold true, then those implications will remain speculative.
   

[C C]

Yah, a lot of people still seem to conceive philosophy as just those few classic areas attributed to it, when it's really developed into [or perhaps always has been] the study, critique, and inventing of formal systems in general. Including any framework of pre-conditional assumptions and tenets that methodological naturalism operates under. In that way, it's always going to be around no matter how much budding new science fields may take over ethics, ontological issues, etc.