Inductivism and Parmenidean epistemology: Kyburg's way (Q432961)

The paper discusses two different views on the standard for serious possibility, the view of Henry Kyburg, and the author's view. What is at stake is whether the standard for serious possibility, that is, the distinction between serious possibilities consistent with full belief and impossibilities inconsistent with full beliefs, can be modified. The answer is yes for the author, and no for Henry Kyburg, who is committed to a Parmenidean epistemology where the distinction between candidates for belief that are possibly true and candidates whose truth is ruled out as impossible never changes. An inductive argument can be used to justify or expand the standard for serious possibility according to the author, but not according to Henry Kyburg. At any given time, the inquirer's evidence commits him or her to a standard for serious possibility. For the author, accepting a hypothesis \(\mathbb H\) by induction makes the inquirer fully believe in \(\mathbb H\), which is then added to the evidential corpus of formulas, giving \(\mathbb H\) the status of knowledge that is not risky. For Henry Kyburg, \(\mathbb H\) can only be supported by the evidential corpus and join the practical corpus of formulas; both the evidential corpus and the practical corpus consist of risky knowledge, with all knowledge in the practical corpus being more risky than the knowledge in the evidential corpus. The author discusses how his views have changed over the years. He also sketches some of the notions which he and Henry Kyburg have developed in order to formally support their respective views, using algebras for the former, and modal logic for the latter. A key aspect of the frameworks is that in Kyburg's model, \(\mathbb H\) is accepted in the practical corpus iff its probability reaches a specified level (threshold for acceptance), whereas in the author's model, there is a threshold for rejection, and unrejected elements together with their logical consequences join the evidential corpus.