Anselm and Actuality

Philosophy abounds in troublesome modal arguments-endlessly debated, perennially plausible, perennially suspect. The standards of validity for modal reasoning have long been unclear; they become clear only when we provide a semantic analysis of modal logic by reference to possible worlds and to possible things therein.2 Thus insofar as we understand modal reasoning at all, we understand it as disguised reasoning about possible beings. But if these are intelligible enough to provide modal logic with foundations, they are intelligible enough to be talked about explicitly. Modal reasoning can be replaced by nonmodal, ordinary reasoning about possible things. Given an obscure modal argument, we can translate it into a nonmodal argument-or into several nonmodal arguments, if the given argument was ambiguous. Once we have a nonmodal argument, we have clear standards of validity; and once we have nonmodal translations of the premises, we can understand them well enough to judge whether they are credible. Foremost among our modal headaches is Anselm's ontological argument. How does it fare under the translation treatment I have prescribed? It turns out to have two principal nonmodal translations. One is valid; the other has credible premises; the difference between the two is subtle. No wonder the argument has never been decisively refuted; no wonder it has never convinced the infidel.