对“老证据问题”的解决——从语境论的角度看

Solutions to “the Old Evidence Problem”: From the Perspective of Contextualism

对于"老证据问题",豪森和厄巴赫提出一个基于贝叶斯方法的解决方案。该方案实际上借助于一个"检验语境的原则"即:新证据相对于当前背景,老证据相对于老背景,而且老背景相当于当前背景减去老证据。笔者的方案接受了该方案的贝叶斯方法和语境主义的立场,但对老背景的内容作了修正,即由"当前背景减去老证据"改为"由当前背景减去旧理论"。相比之下,笔者的方案更为符合科学史实,同时印证了库恩的范式理论。

For the old evidence problem, Howson and Urbach propose a solution based on the Bayesian approach. Their solution appeals actually to the principle of testing context, i.e., the new evidence is relative to the present background, while the old evidence is relative to the old background. Besides, the old background amounts to the present background minus the old evidence. The solution proposed in this paper adopts the standpoint of Bayesian approach and contextualism, but makes some modifications about the content of the old background. It manages to replace 'the present background minus the old evidence' with 'the present background minus the old theory'. The author's solution matches historical facts of science better, and confirms Kuhn's theory of paradigm to a great extent.