A Problem About Sufficient Conditions

Abstract

(I) ... is a sufficient condition for . Hence one would expect the pitfalls and foggy patches of the concept to be well known and copiously discussed. For the most part this is true. Yet there is an obscurity concerning sufficient conditions among distinct existences which, though it arises quite crucially in standard uses which philosophers make of sufficient conditions, has drawn no attention or comment so far as I know. But, before turning to this obscurity (in section 2), let us sketch the already mapped terrain. I will confine my attention to the very common sort of case where the blanks in (I) are filled not with names of statements, but with singular terms which name (or purport to name) properties; or, with a more cautious and frugal scanning of our ontic commitments, we might fill them with singular terms naming predicates or sentential functions. These fillings would yield the following exemplars of schema (I): (2) Bachelorhood is a sufficient condition for being unmarried. (3) 'x is a mammal' is a sufficient condition for 'x is warm blooded'. The examples are intended to differ in two ways. First, I will take (2) as stating a logically sufficient condition and (3) as stating an empirical one.1 Second, a distinction already suggested, (2) is a second order remark about intensional objects whereas (3) is a metalinguistic remark about entities in an object language; but for the most part I shall gloss over this latter distinction.