Probabilistic Causality

From MaRDI portal





This is a work in the tradition of \textit{P. Suppes} [A probabilistic theory of causality (North Holland, 1970)], \textit{Good} [Pac. Philos. Q. 61, 197-217 (1980)] and \textit{Cartwright} [Nous 13, 419-437 (1979)]. The concern is not that of discovering causal relations in statistical data, but that of understanding what an assertion of probabilistic causal relation ``means. This is not an easy question to answer, and the argument of the book consists primarily of artificial examples designed to tease from us the right intuitions. The argument is careful and detailed; if it is not entirely persuasive, it is because the concept being explicated, that of probabilistic causality, is one that is not clear to start with. The general idea is that causes raise the probability of their effects, other things being equal. But we do not want to hold everything equal. So given an empirical interpretation of probability [much like that of \textit{R. Braithwaite} in ``Scientific explanation (Cambridge, 1953)], we want to say that \(C\) causes \(E\) when the probability of \(E\) given \(C\) is greater than the probability of \(E\), in populations in which factors of one kind are held fixed, and in which factors of another kind are varied. Chapters 3 and 4 discuss the two kinds of factors. Chapter 5 discusses the role of temporal asymmetry, and Chapter 6 opens up a new discussion: that of probabilistic ``token causality -- that is, causality that concerns a single unique event, rather than a population. Those concerned with the discovery of causal patterns in statistical data will not find much of interest to them here; those concerned with understanding the meaning of causal assertions where universal relations are lacking will find a lot to intrigue them.




Cited in
(34)








This page was built for publication: Probabilistic Causality

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4023238)