A Theory of Vague Expectation

在之前由国家科学基金会支持的工作中，范·弗拉森教授研究了科学认识论与经验主义兼容的可能性。他特别关注贝叶斯与传统认识论模型之间的张力，即理论的改变和证据的适应。他通过对未来(后验)概率分布的先验预期理论找到了一种可能的和解形式。对迄今为止发展起来的观点至关重要的是，通过概率函数对意见进行建模，但以一种允许对意见进行模糊、定性和比较判断的方式，只承认数字概率是一种极限情况。许多相关的定理很容易从数值上的精确转变为模糊的观点，但许多却并非如此。范·弗拉森教授目前正在进行的项目的目的是发展这一不可或缺的子理论。

In prior work supported by the NSF, Professor Van Fraassen has examined the possibility of an epistemology for science compatible with empiricism. He focussed special attention on the tension between Bayesian and traditional epistemological models of theory change and accommodation of evidence. He found a possible form of reconciliation through the theory of a priori expectations of future (posterior) probability distributions. Crucial to the view developed to date the modeling of opinion through probability functions, but in a way that allows for vague, qualitative, and comparative judgments of opinion, admitting numerical probabilities only as a limiting case. Many relevant theorems transpose easily from numerically precise to vague opinion, but many do not. The aim of the project which Professor Van Fraassen is now undertaking is to develop this indispensable subtheory.