An Analysis of Probabilistic Criteria in Inference to the Best Explanation

According to a favored version of inference to the best explanation (IBE), the hypothesis that provides the best explanation for a group of observed phenomena is probably true. One of the key objections against this line of thought is Voltaire’s objection. According to this objection, we have no reason to believe that the explanatory virtues, as the criteria for selecting the best explanation, provide the likeliest explanation. In this paper, we address probabilistic criteria which are subject to Voltaire’s objection in IBE. To be exact, we scrutinize this question: considering proposed probabilistic criteria for assessing the explanatory hypotheses as the criteria for selecting the best explanation, is it justified to think of the best explanation as the likeliest one? Dealing with this question, the probabilistic criteria will be analyzed in three group: criteria based on Bayes’ theorem, criteria based on confirmation theory and criteria based on explanatory virtues. We argue that none of these criteria can determine the best explanation which provides the likeliest one.