Azar y explicación. Algunas observaciones

Abstract

In the first part of this paper I examine the Hempelian model for the probabilistic explanation of particular events: the inductive-statistical model. Here I focus on an examination of the notion of expectability and the implied requirement of high probability. I intend to show that expectability and high probability, in turn, answer to a deep rooted intuition concerning explanation: given an event E, the same sort of circumstances cannot explain E and -E. This intuition is called here “the basic principle”.
 This basic principle is also the ground for the Hempelian thesis that to explain is always to explain why. I think this thesis is wrong and I propose to distinguish between explaining why and explaining how possibly . I discuss the difficulties confronting the statistical-inductive model in the light of the assumptions here examined and suggest possible solutions.
 In the second part, I explore those possible solutions, in particular I examine the symmetry principle proposed by Salmon: given an stochastic process, the highly probable results are as well understood as the improbable results. This principle implies the rejection of the basic principle. I claim that the symmetry principle must be restricted to explanations how . Later I discuss the DNP model proposed by Railton and I make some criticisms to it. I try to show that if we make explicit the distinction between explaining why and explaining how in this model, we can restrict clearly the range of applicability of the symmetry principle and the basic principle, I conclude that Railton’s model constitutes an adequate basis to establish the scope and limits of probabilistic explanations of genuinely random events.