The reliability of Inference to the Best Explanation (IBE) is sometimes contested because it may end up recommending the best of a bad lot, which might require a modification of IBE (e.g. contra Schupbach). For instance, some would argue that IBE recommends to believe the best explanation only if it is good enough; others suggest that it recommends tentative inferences only.

I'm not concerned about the exact formulation of such approaches, which is certainly an open issue. But what can they even hope to achieve beyond simply conceding to the bad lot objection?

Sure, they are meant to provide conditions for the reliable application of IBE, though probably only by reflecting the attitude that one would have, in a given situation, towards the best explanation anyway (viz. in light of the bad lot objection) ...

Thanks for input!

  • 1
    Schupbach's response is not very convincing. The matter/form dichotomy makes sense for universally valid logical inferences but not for heuristics like IBE, where material warrants for its application are highly domain specific both in their refined form and plausibility. One surely can not hope to achieve anything in this vague generality of "best" and "good enough", the real progress happens in developing detailed heuristics for generating "good lots" in specific areas. There is no separating material and formal in IBE, it is just a vague meta-principle. – Conifold Dec 22 '20 at 22:00
  • I do not want to stand up for Schupbach here, though I think his goal is not to defuse the bad lot objection but to show that remodeling IBE is an inappropriate response. And that is precisely because the real challenge is to produce "good lots". – Turtur Dec 22 '20 at 22:11
  • 1
    The interesting part is whether IBE (and induction) is vulnerable to the bad lot objection the way MP is not, and I think it is. MP never produces a "bad" conclusion on "good" (close to true, plausible) premises, but IBE may. There is an extra issue of describing what a "good lot" for IBE even is. Therefore, to the extent that vague abstractions are used as guides, it is appropriate to spell out the extra option missing for MP - declaring inconclusive. And we see it exercised in practice. String theory is the best explanation we have for..., but it is not inferred as ripe for acceptance. – Conifold Dec 24 '20 at 6:38

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.