Can counterfactuals be rescued?
I’ve criticized the unserious way in which ordinary language handles counterfactuals. I denied the counterfactual “If the ball hadn’t been caught, then it would have fallen” on the grounds that the ball wouldn’t have existed in the first place if it hadn’t been caught. Of course I accept the obvious inductive truth that uncaught balls fall, but we shouldn’t express that obvious truth as a counterfactual that isn’t obviously true.
One might try to rescue the aforesaid counterfactual by claiming that it contains an unspoken rider: “If the ball hadn’t been caught and still had existed, then it would have fallen.” Isn’t that what ordinary speakers implicitly mean anyway? A problem arises, however. Because in fact the ball was caught, the situation in which it wasn’t caught and still exists is a thermodynamic miracle, like ice that’s left out at room temperature without melting. [For discussion of the thermodynamics, see Douglas N. Kutach, “The Entropy Theory of Counterfactuals,” Philosophy of Science 69 (2002).]
So revising the “If” part of the counterfactual in the suggested way requires accepting a thermodynamic miracle. In that case, we have no principled reason for insisting that the uncaught miracle ball falls to the ground. All it takes for the ball to float upward instead is another thermodynamic miracle of the kind we already accepted. We can see, then, that adding an unspoken rider to everyday counterfactuals doesn’t rescue them.
