AUTOHYPONYMY: IMPLICATURE IN LEXICAL SEMANTICS, WORD FORMATION, AND GRAMMAR

Abstract

The word meanings given in dictionaries normally are “pragmatic meanings,” that is, utterance meanings in prototypical situations. Those glosses are perfectly adequate for the needs of the normal dictionary user but they do not meet scientific standards. As H. P. Grice and L. Horn have shown in various publications, the apparent meanings of many words are in fact combinations of their lexical meanings proper and some superimposed conversational implicatures. Many contradictions can only be avoided by keeping these two parts separate. For example, the use of the adjective rectangular often (but not always) is interpreted as a significant non-use of the (stronger) alternative square. The adjective rectangular thus acquires a second, pragmatic reading ‘rectangular but not square’ that is hyponymous to its semantics; the word appears to be hyponymous to itself, in other words, to be an “autohyponym” (Horn 1984c:110). The purpose of this paper is to show that autohyponymy is pervasive in the lexicon, occurring with nouns, verbs, (scalar/degree) adjectives, conjunctions (e.g., the conditional if); it also occurs in word formation (the German feminine suffix -in), and grammar, for example, the absolute comparative (ein älterer Mann). The data are largely taken from German.This paper is dedicated to Ekkehard König on the occasion of his 60th birthday, January 15, 2001. I am indebted to the anonymous readers of this paper for many very helpful comments.