Published online by Cambridge University Press: 01 April 2022
In his recent work, Michael Redhead (1986, 1987, 1989, 1990) has introduced a condition he calls robustness which, he argues, a relation must satisfy in order to be causal. He has used this condition to argue further that EPR-type correlations are neither the result of a direct causal connection between the correlated events, nor the result of a common cause associated with the source of the particle pairs which feature in these events. Andrew Elby (1992) has used this same condition as a premise in an independent argument for the conclusion that EPR-type correlations cannot be causally explained (except, perhaps, by a nonlocal hidden variable theory). I wish to argue here that robustness is itself too fragile a notion to support such conclusions.
I thank Michael Redhead for explaining to me the relation between his earlier and later arguments using robustness; the participants in an informal seminar at the University of Maryland for their constructive reactions to my early thoughts on this matter; Andrew Elby for an illuminating correspondence; Paul Teller for helpful comments on an earlier version; the editors of Philosophy of Science, and especially an anonymous referee, whose detailed comments resulted in significant improvements in the present version.