摘 要：葛梯尔问题产生的实质是传统知识三元定义中的三个条件：真、确证、信念之间，尤其是确证与真之间存在着某种程度的独立性。针对葛梯尔问题的实质，丹尼尔（Daniel Howard-Snyder）等人从“可错主义”的角度对葛梯尔问题加以分析和解决，而弗洛里迪（Luciano Floridi）则从“协作攻击”的角度对葛梯尔问题进行了分析与回答。但是，丹尼尔对葛梯尔问题的分析与解答所依据的逻辑命题是有漏洞的，而弗洛里迪的分析与解答则不能解释可错的经验知识的合理性。葛梯尔问题提示我们，确证不等同于真。一个信念的真与理由之间要有适当的关联，但与一个信念的真相比，知识论研究的关键在于为信念寻找理由。因为在绝大多数情况下，信念与真之间总存在一个时间的间距，我们相信一个信念并不是因为它是真的，更多的是因为有好的理由。正是在好的理由的基础上，主体相信一个命题是真的。
关键词： 葛梯尔问题； 可错主义； 协作
On the Solution of Gettier Problem from Two Perspectives
SONG Qun
Abstract: “Gettier Problem” arises from the tripartite prerequisites of the traditional definition of knowledge in which truth, justification and belief, especially the first two, are partly independent. This paper introduces two perspectives of analyses and solutions to the problem, i.e. fallibism (Daniel Howard-Snyder et al.) and “coordinated attack”(Luciano Floridi). Daniel’s logical proposition on which his analysis and solution to “Gettier Problem” based is flawed, and L. Floridi’s can not explain the justifiability of empirical knowledge. “Gettier Problem” indicates that justification is not equal to truth. There should be proper relations between justification for a blelief and the truth of such a belief. The key to epistemological research lies in justification rather than the truth. In most cases, there is always a time-distance between justification and the truth of a belief, and we accept a belief not that it is true, but that we have some good reasons for it.
Key words: Gettier Problem; Infallibism; coordination
