Logos and episteme
Carin ROBINSON
ABSTRACT: This paper asks the following question: How is deduction non-ampliative? It answers with a conditional: If deduction is non-ampliative then its premises are non-informative. To establish the truth of this conditional, is the task of this paper. I argue that the non-ampliative nature of deduction entails the non-informative epistemic status of the sentences which are its premises. In other words, by non-informative I do not mean there is no more information in the conclusion than in the premises. This construal of ‘non-informative’ is already what we mean by ‘non-ampliative’, and is typically not a controversial claim. By non-informative I mean that the premises themselves are non-informative sentences. Such an account of deduction I call an analytic account of deduction. Three conditions are required to support an analytic account of deduction: 1. epistemic closure, 2. uniformity and 3. intensional semantics. If an analytic account of deduction is correct then it poses problems for philosophical positions that deny the possibility of analytic truth and a priori knowledge. But it also poses problems for philosophical positions that accept analytic truth and a priori knowledge, but nevertheless maintain that deduction is more informative than a mere rearrangement of meanings or ideas along inferential lines.
