- 著者
-
細川 雄一郎
- 出版者
- 科学基礎論学会
- 雑誌
- 科学基礎論研究 (ISSN:00227668)
- 巻号頁・発行日
- vol.47, no.1, pp.15-34, 2019
<p>By applying the extended system of modal logic developed in [32], we present a logical analysis of the Gettier problem. Based on the result, in particular, we clarify the following point. In the literature on the Gettier problem, most authors seem to share the opinion that in general 'belief' precedes 'knowledge': we believe a proposition while we do not necessarily believe ourselves to have known it to be true. Interestingly, our analysis suggests that the story is the other way round in a sense: in general, for some proposition <i>p</i>, we <i>believe</i> that we <i>have known</i> that <i>p</i>, <i>then</i> we <i>believe</i> that <i>p</i>. Accordingly, even if given some reason for wanting to believe that <i>p</i>, we usually do <i>not</i> have the bare belief that <i>p</i>. Instead, in such a situation, more deliberately we think that <i>it might be that</i> <i>p</i>, or <i>it can be hypothesized that p</i>. Then, we can say, what we have at the start is <i>not</i> the bare belief that <i>p</i>, <i>but "might possibility"</i> or "hypothetical possibility" that <i>p</i>, which presumably involves <i>abductive reasoning</i>.</p>