Abstract

This study purports to answer several questions through the analysis of logical structure of might Why doesn't might pattern with the other modals of possibility under negation? Why does might resist realization conjunctions---conjunctions in which the second conjunct asserts what is entertained as a possibility in the first conjunct? and more generally, Why are forms which express the O corner of Aristotle's square of oppositions dispreferred syntactically, morphologically and semantically? ;The evidence of realization conjunctions indicates that might expresses uncertainty, not just a kind of possibility, but fundamentally distinct from possibility or contingency in its entailments. Nor does it appear to differ merely in conversational implicature. Realization conjunctions confirm Grice's and Horn's observations that possibility, because it is on a scale with necessity, excludes necessity, thereby including possibly not in the use of possibly, and providing the implicature of contingency. These implicatures are predictably cancellable. Such cancellation does not occur with might. ;The distinction between might and the other modals of possibility must therefore be reflected in its semantics, in its logical structure, and it is the logical structure that answers the other questions addressed here as well. Taking the logical structure of might to contain a conjunction of at least an assertion of possibility and a denial of knowledge, it is possible to identify the reason why wide scope of negation, typical of auxiliaries expressing possibility, does not characterize might under negation. The negation of a conjunction is ambiguous. This in turn suggests that the reason negations of necessity and universality are underrepresented in language is that they are, like the negation of might, informationally weak, either vague or, as in the case of might, ambiguous