Investigations in the Theory of Confirmation and Inductive Inference

证实理论与归纳推理研究

• 批准号: 9818150
• 负责人: John Norton
• 金额: $ 5.88万
• 依托单位: University of Pittsburgh
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 1999
• 资助国家: 美国
• 起止时间: 1999-09-01 至 2000-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9818150&HistoricalAwards=false
• 关键词: Investigations; Theory; Confirmation; Inductive; Inference

This project comprises a series of studies into confirmation theory and the nature of inductive inference. The first part is designed to deepen our understanding of a scheme of inference known as demonstrative inference. In it, an inductive inference from evidence to hypothesis is rendered deductive (demonstrative) by the addition of suitable general assumptions to the premises. This turns out to be a strategy used quite frequently in the history of science as a way of controlling inductive risk. The principal investigator will seek further examples of these inferences and also attempt to understand how it is possible for a special class of these inferences to be self-validating in the sense that the success of the inference also supplies support for the general assumptions. In the second part, the principal investigator will seek to understand the weaknesses and strengths of the Bayesian treatment of confirmation theory by developing an axiom system for probability theory. The axiom system will not be designed to be as parsimonious as possible, but rather to be based on intuitively distinct commitments concerning the nature of confirmation. This will enable various of the appealing properties of the Bayesian scheme to be distinguished from others. Finally the principal investigator will briefly explore several heresies pertaining to inductive inference.

这个项目包括对确认理论和归纳推理本质的一系列研究。第一部分旨在加深我们对一种称为证明性推理的推理模式的理解。在它中，从证据到假设的归纳推理通过在前提中添加适当的一般假设而变得演绎(演示)。事实证明，这是科学史上经常使用的一种策略，作为控制归纳风险的一种方式。首席调查员将寻求这些推论的进一步例子，并试图理解如何可能有一类特殊的推论是自我验证的，因为推论的成功也为一般假设提供了支持。在第二部分中，首席研究者将通过发展概率论的公理系统来试图理解确认理论的贝叶斯处理的弱点和优点。公理系统将不会被设计得尽可能节俭，而是基于关于确认性质的直观地不同的承诺。这将使贝叶斯方案的各种吸引人的特性与其他方案区分开来。最后，首席研究员将简要探讨与归纳推理有关的几种异端邪说。