Stella Fillmore-Patrick

Michael Titelbaum argues that there is no objective, three-place evidential favoring relation in cases without entailment: inductive inference, unlike deductive inference, always depends on some subjective feature (2010, 2011). This position is based on the widely accepted view that language dependence violates objectivity, so since inductive logic is language dependent it is not objective. Against this view, I argue that a formal theory of partial entailment is language dependent, objective, and formal. Titelbaum's argument fails, I claim, because he adopts implausible conditions for objectivity: invariance with respect to permutations of predicates and a formal method for selecting a language for induction.

In the middle of the twentieth century, Claude Shannon’sShannon, Claude 1948 paper “A mathematical theory of communication” incited a flurry of philosophical interest in the concept of ‘information.’ Shannon’s work was developed with applications to telecommunication systems in mind. Although Shannon’s theory was developed for a particular application in telecommunication, his use of the term ‘information’ was highly evocative to those working in other fields that employed a concept called ‘information’. Shannon himself considered his theory to be irrelevant to semantics as he only gave an account of amount of information, not content, but others imagined that ‘Shannon information’ could be applied to solve problems in semantics, psychology, and philosophy.