From the Emergence of Probability to the Erosion of Determinism

Abstract

'Hacking's thesis is that the sceptical problem of induction became a possible problem only as a result of two events. The first was the emergence of a concept of what Hacking calls 'internal evidence'—that is, evidence other than testimony. It was this that enabled the modern concept of probability to emerge, and with it the analytic problem of induction. The appearance of the sceptical problem required one further change. "Once the concept of internal evidence was established by 1660, the final transformation needed for the sceptical problem of induction was this transference of causality from knowledge to opinion" [Hacking 1975, 180] [¶] Hacking's claim that there was no concept of internal evidence in Medieval or Renaissance Europe has been damagingly criticised by a number of writers (Blackburn 1976; Garber and Zabell 1979; Laudan 1981, 72-85). One particular weakness is that Hacking bases his argument on a very implausible (indeed quite unsustainable) claim about natural and conventional signs. [.… ] According to Hacking, therefore, the distinction between arbitrary and natural signs and the concept of internal evidence emerge together around the middle of the seventeenth century. One fundamental objection to this argument is that the distinction between natural and conventional signs, far from being a new discovery of the seventeenth century, was a commonplace of medieval philosophy' (Milton 1987, 65).