Abstract
The overall motivation of this study arises from two assumptions about science which are quite uncontroversial in today’s philosophical discussions: Firstly, it is assumed that empirical evidence is fallible and many of our present or past evidential beliefs have actually been false. Secondly, it is assumed that our ability to correctly infer the truth or falsity of a hypothesis depends on whether our body of evidence is correct.
Forming beliefs according to confirmation, is this a valuable way of reasoning with false evidence? What different ways of forming beliefs according to confirmation are there and do they differ in their truth-conduciveness? How truth-conducive is it to adjust one’s beliefs to degrees of confirmation? How truth-conducive is it to only accept those hypotheses which are sufficiently confirmed by one’s evidence? These questions trigger a statistical turn and are answered in detail in part one of this thesis.
What is going on in real scientific debates? How do participants change their beliefs such that they finally reach a consensus? Do they shift from one group of a dating hypothesis and evidential beliefs to another such that their degrees of confirmation increase? Do they shift from one group of a dating hypothesis and evidential beliefs to another such that there is no other hypothesis which is better confirmed? Do such shifting always increase similarity with the final consensus? These questions trigger a historic turn and are answered in detail in part two of this thesis.
