The varieties of confirmation - Mark Siebel
Bayesians agree on the qualitative principle that confirmation is to be equated with increase in probability: E confirms H just in case P(H|E) > P(H). There is, however, a myriad of proposals for corresponding measures of confirmation, such as the difference measure P(H|E) – P(H) or the ratio measure P(HE) / P(H). Recently, it was suggested to drastically narrow this field by drawing upon the idea that confirmation is also a generalisation of deductive entailment. After some more general remarks on probabilistic accounts of confirmation, I try to show that the whole dispute on “the one true measure of confirmation” is misdirected. Firstly there are disparate qualitative conceptions of confirmation, among them increase in plausibility versus generalisation of deductive entailment. Secondly, even if we confine ourselves to increase in plausibility, a quantitative ambiguity remains because there is in general a plurality of proper answers to the question how much a quantity has increased, such as ‘to twice as much’ (ratio) or ‘by 0.8’ (difference).
Mark Siebel is chair in theoretical philosophy at the University of Oldenburg since 2007. His research areas include epistemology and philosophy of science, philosophy of mind and language and the history of analytic philosophy. He made his PhD in Hamburg in 1998 and worked as a postdoc in the Postgraduate Programme for Cognitive Science (Hamburg) and as a scientific assistant in research projects in Leipzig. In 2005 he came back to Hamburg as a scientific assistant, where he habilitated in 2007. He wrote numerous articles and the books Der Begriff der Ableitbarkeit bei Bolzano (1996) and Erinnerung, Wahrnehmung, Wissen (2000).
|Laatst gewijzigd:||30 oktober 2012 20:39|