J.P. Moreland writes: “Causal necessitation grounds the derivation of counterfactuals” [J. P. Moreland, The Argument from Consciousness in the Blackwell Companion to Natural Theology, p.293]
It occurred to me that counterfactuals cannot be straightforwardly true or false according to a Humean view of causation as constant conjunction. If one believes that causal connections are inferred, and only constant conjunction is empirically observed (and also that there is nothing necessary about causal connections), then one cannot believe in some counterfactual’s being straightforwardly true or false. Consider that a counterfactual such as ‘if I had not stopped the bowling ball from rolling into the pins, then the pins would have been knocked over by the bowling ball’ (of course, this is all things being equal, thus we are properly ignoring the possibility of somebody else stopping the bowling ball, etc.). The statement cannot be considered ‘true’ strictly speaking, since we are only projecting an inductive speculation about our uniform experiences, but there is absolutely no necessary connection between what we perceive as cause and what we perceive as effect.
Perhaps the Humean would want to say that they can speak ‘the counterfactual language game’ by projecting inductive speculation, but they cannot ever take counterfactuals to be straightforwardly true in the sense that, given some set of conditions which causally give rise to some event, if those conditions obtain, then the event necessarily follows.