Experience, causation, and induction › Doubt without indifference to evidence

A reasonable objection is that Hume has substituted a description of belief for a justification of belief. Knowing why a habit developed does not show that it is trustworthy. The objection identifies a real limit of the skeptical solution. Hume's Section V does not produce the missing noncircular foundation. It explains why that missing foundation does not leave human beings unable to infer or act.

Another objection asks whether demonstrative certainty was the wrong demand from the outset. Perhaps an inference counts as reasonable because it responds well to evidence, even though it cannot justify the whole practice of responding to evidence without using that practice. This is a proposed philosophical response, rather than a claim that Hume explicitly resolves the difficulty in these terms. Its strength is that it distinguishes good local reasoning from a foundation for all empirical reasoning. Its unresolved task is to explain those standards without simply assuming the conclusion under dispute.

Hume's own discussion of testimony shows why his skepticism should not flatten evidential differences. In Section X, he recommends proportioning belief to evidence and distinguishes stronger from weaker assurance. He examines the circumstances that support or undermine testimony, including disagreement between witnesses and their interests in what they report. He then applies the comparison to miracle reports. His argument weighs testimony against experience supporting natural regularities; it is not merely the claim that anything unusual is impossible. E 10.3–13.

That discussion raises an objection of its own: how should an investigator distinguish an inadequately supported report from evidence that a supposed regularity needs revision? Hume's procedure requires a comparison of evidence, and the result depends on the strength attributed to each side. Reporting his criticism of miracle testimony as a definition that automatically excludes every challenge to an accepted law would hide the epistemic question rather than answer it.