Does a Lucky True Belief Count as Knowledge?

You enter a café, see that the wall clock reads 4.35, and believe that it is 4.35. The belief happens to be true, but the clock actually stopped three days ago. You appear to have an ordinary reason, you believe the claim, and the claim is true. Yet we resist saying that you knew the time, because the correct display was a coincidence.

This is a familiar Gettier-style case, not one of the examples in Gettier's 1963 paper. His target was a classical analysis according to which knowledge is justified true belief. His counterexamples suggested that all three conditions can be met while the belief is true only through epistemic luck. Later proposals have added conditions such as no false premises, a reliable process or sensitivity to nearby possibilities, but no simple account has secured universal agreement.

This differs from asking whether knowledge requires absolute certainty. A judgement may remain fallible while being supported by reliable evidence; conversely, intense confidence may merely be lucky. My view is that knowledge does not require eliminating every possibility of error. It requires a non-accidental connection between the truth and the grounds on which it is believed. We should ask not only whether the conclusion is correct, but whether the same grounds would still guide us towards the truth in nearby circumstances.

https://plato.stanford.edu/entries/knowledge-analysis/
https://doi.org/10.1093/analys/23.6.121