Conférence de Joseph Berkovitz : On the role of intuitive thinking in probabilistic reasoning

Joseph Berkovitz (HPST, University of Toronto) donnera une conférence à l'IHPST sur le sujet suivant : On the role of intuitive thinking in probabilistic reasoning.

Résumé : The prevalent epistemic frameworks of probabilistic and statistical reasoning focus on propositional knowledge, conscious inferential reasoning, and objective evidence, and conceive the relations between evidence and probabilities as logical or objective. Probabilistic reasoning is supposed to rely on objective probabilities and logical or objective evidential relations, and subjective probabilities should play a role only to the extent that they are regulated by objective probabilities. In these epistemic frameworks, the role of intuitive thinking is severely restricted, and its output is often considered suspicious, requiring a particular caution and justification.

Bruno de Finetti (1906-1985) is one of the founding fathers of the subjective theory of probability. His theory – which has been influential in the development of Bayesian epistemology and the associated Bayesian statistics – is situated in an epistemic framework that is very different from the prevalent ones. In this framework, intuitive thinking – which is non-inferential and largely unconscious – is central and fundamental to probabilistic reasoning, and propositional knowledge and conscious inferential reasoning play an important but complementary role. De Finetti rejected the idea that subjective probabilities should be regulated by objective probabilities, and he argued that probabilistic and statistical reasoning should be based only on subjective probabilities. 

In this paper, I present the epistemic framework of de Finetti’s theory of probability. I argue that: (i) the objections to this theory are largely unwarranted, as they based on misinterpreting its epistemic framework, overlooking the fundamental role that intuitive thinking plays in it; and (ii) that one of the main objections that de Finetti had to objective probabilities is that they fail to integrate the power and insight of intuitive thinking. I conclude by briefly considering the implications of de Finetti’s epistemic framework for probabilistic and statistical reasoning.