The sorites paradox

The characteristic feature of a vague predicate is its possession of borderline cases. For us to even consider an item to be a borderline case of V, or equivalently, a borderline case of notV, our intuitions on the correct use of V must to a significant degree find plausible the application of V to it counterbalanced by our finding nearly as plausible the application of n o t V to it; otherwise we would consider it to be a definite case of V or a definite case of no t -V. Borderline cases of a vague predicate V are items for which it is rationally acceptable to classify them as V and rationally acceptable to classify them as notV, that is, items for which the difference in warrant for applying V or applying n o t V is too slight to judge that the one term can be acceptably applied but not the other, too slight to enshrine only one of the predicates V, or n o t V as the only correct one to apply to any one of the given items to whose classification there has as yet been no commitment . The borderline cases of V constitute the overlap of the range of the acceptable application of V with the range of the acceptable application of no t -V. Tha t is the overlap construal of borderline cases: the rational acceptabili ty of opposite classifications. To accurately characterize borderline cases more would have to be said. For one thing, one 's finding acceptable both of opposite classifications of an i tem is not to be attribtued to ignorance of certain relevant facts about the classified item or of linguistic usage. For present purposes no more need be said. Far f rom it being impossible to say of a borderline case whether it is V or notV, we can admit that either classification of a borderline case is a rationally acceptable one, though obviously, either term is applied with less warrant than we customarily require or seek to have for the assertion of a s tatement expressing the classification of a nonborderl ine case. There is no reason to disallow this lowering of the standards of warrantedness for classification judgments on borderline cases; such judgments are less like judgments of matters of fact and more of the nature of decisions settling what one 's usage is to be on the particular occasion where the speaker ' s intention and say-so can be definitive. It is