Alan Author strikes again: More on confirming conjunctions of disconfirmed hypotheses.

The so-called Alan Author Effect is a surprising phenomenon in Bayesian Confirmation Theory. It occurs when a piece of evidence e confirms the conjunction of two hypotheses h 1 ∧ h 2 but at the same time disconfirms each hypothesis h 1 and h 2 individually. In this paper, I present a new and prima f...

Descripción completa

Detalles Bibliográficos
Publicado en:Analysis Vol. 85; no. 2; pp. 359 - 370
Autor principal: Koscholke, Jakob
Formato: Artículo
Publicado: Oxford University Press / USA Apr2025
Materias:
Acceso en línea:Ver este registro en EBSCOhost
Descripción
Sumario:The so-called Alan Author Effect is a surprising phenomenon in Bayesian Confirmation Theory. It occurs when a piece of evidence e confirms the conjunction of two hypotheses h 1 ∧ h 2 but at the same time disconfirms each hypothesis h 1 and h 2 individually. In this paper, I present a new and prima facie stronger version of this effect where additionally, the evidence e confirms the conjunction of the negated hypotheses ¬ h 1 ∧ ¬ h 2 ⁠. I say 'prima facie' because it can be shown that this seemingly stronger effect and the original effect are actually coextensional. I use this insight to formulate a new sufficient (and also necessary) condition for the two equivalent effects. I also examine how likely the two effects are to occur with the help of Monte Carlo simulation methods.