Is the multiverse hypothesis a legitimate scientific theory? That is, are there regularities that illuminate and reflect underlying laws of nature by testing these laws and making predictions that can be either verified or refuted by experimentation and observation? Generally, these are the guidelines for something to be scientific, can it be verified and falsified? Before I continue, we need to make a distinction in two fundamental philosophies of science: instrumentalism and realism.
Instrumentalism: Scientific theories are not intended to be literally true and accepting a theory requires us to believe only that its observational consequences are true. Observation statements are literally true and science is only about these statements and the observations that verify them. A few strengths of this philosophy is that it doesn’t conflict with common sense realism; we can believe in straightforward observations. Plus, it’s more modest and non-commital than scientific realism. A few weaknesses are that scientists seem to assume the realist view of the world in their “un-thinking” moments. The instrumentalist should be able to draw a clear cut distinction between what is and what is not observable, which creates limitations on what really is observable (i.e. naked eye, magnifying glass, microscope, electron microscope, cloud chamber, etc.). This also raises the question, at what point is the objecting being observed really being observed, and so real, but then one bit smaller is not observable and thus not really existent?
Scientific Realism: Scientific theories are intended to be literally true, and accepting a theory involves believing that it gives a true description of reality, “as it really is.” A few strengths of this is that it makes the aspect of explanatory power superior to instrumentalism because explanation requires real things that cause the chain of causality. Explanation by means of fictitious entity is not explanation at all. Instrumentalism cannot explain the actual success of science, especially science’s making predictions, which are empirically adequate (i.e. Boyles-Charles Law, pv=k).
I’m going to argue that we should adopt the realist position partly because it is common sense and because it means and ends in explanation provide a robust sense of explanatory power that lacks instrumentalism and the metaphysical baggage it may carry is less deleterious than instrumentalism.