语用学onhornscales.pdf

YO MATSUMOTO THE CONVERSATIONAL CONDITION ON HORN SCALES O. INTRODUCTION 1 Gazdar (1979), Horn (1972, 1989) and others have claimed that, given a pair of a semantically stronger expression (S) and a weaker expression (W), if a speaker chooses to make a weaker statement Q that contains W rather than a stronger statement P that contains S, the speaker implicates (as one possibility) that he or she believes or knows that P does not hold. For example, in uttering (la) the speaker implicates (lb) on the basis of the pair hot and warm. (1)a. b. “It is warm in Northern California today.“ I-+ '(The speaker believes) it is not hot in Northern California today.'2 The implicature is defined as a part of the speaker's meaning that is conveyed by virtue of the assumption that the speaker and the hearer are obeying the Cooperative Principle of Conversation, and more specifically, various Conversational Maxims: Maxims of Quantity, Quality, Relation and Manner (Grice 1975). The type of implicature illustrated in (1) is called a Quantity implicature or more precisely a Quantity-1 implicature, 1 I am indebted to Eve Clark, Mary Dalrymple, Makoto Kanazawa, Stephen Levinson, Patrick O'Neill, Ivan Sag, and Shun Tutiya for their helpful comments during various stages of the development of this paper, and to Mary Dalrymple (again), Ki-sun Hong, Michael Inman, and Harry and Elizabeth Owen Bratt for providing me with data. I am also grateful to Orin Gensler and the reviewers for Linguistics and Philosophy, who made a number of valuable suggestions for improvement. None of them are responsible for any remaining insufficiencies in this paper, or for my possible misinterpretation of what they have said. Address for correspondence: Yo Matsumoto, Box 15, Tokyo Christian University, 3-301- 5 Uchino, Inzai, Chiba-ken, 270-13, JAPAN; e-mail: yomatsum@tansei.ec.u-tokyo.ac.jp. 2 In this paper the following notational conventions are used. ~p~ ~p~ P--~Q P4~Q el~o eb Q the utterance P the proposition P P entails Q P does not entail Q the speaker conversationally implicates Q by P the speaker does not conversationally implicate Q by P Linguistics and Philosophy 18: 21-60, 1995. (~) 1995 Kluwer Academic Publishers. Printed in the Netherlands. 22 YO MATSUMOTO since what is crucial in the production of this implicature is the first Gricean Maxim of Quantity - 'make your contribution as informative as is required'. The pairs (or sets) of semantically stronger and weaker expressions that license or provide a basis for Quantity-1 implicatures are called Horn scales (Horn 1972, 1989; Levinson 1983). According to the convention introduced by Horn, such a scale is indicated in angular brack- ets, with the items in the scale ordered from strongest to weakest (from left to right), as in (hot, warm~. The main question addressed in this paper is that of the constraints on Horn scales. Two kinds of constraints have been recognized in the litera- ture. One of them is what I call the informativeness requirement: S must be more informative than W. A precise formulation of this requirement has been a much-discussed topic. Horn (1972) formulated it in terms of logical entailment: S must entail W. Hirschberg (1985) has observed that items related by non-entailment relationships such as rank orderings, spa- tial orderings, and process stages can also form Horn scales, and argues that any items constituting a partially ordered set in which it can be determined whether one item is higher or lower than another can function as a Horn scale. In addition to such an informativeness requirement, some researchers have proposed that S and W must satisfy certain additional conditions in order for S to function as a legitimate alternative to W so that it can be used in producing an implicature. For example, Gazdar (1977) has suggested that W and S must share selectional restrictions and item-induced presuppositions (cf. Gazdar 1979). Atlas in other cases the speaker implicates that s/he believes (or knows) that a stronger statement does not hold of the situation. I will call the former type weak Quantity-1 implicature, and the latter, strong Quantity-1 im- plicature. The two kinds of Quantity implicatures can be understood in terms of 4 This formulation is slightly different from Grice's original formulation 'make your contribu- tion as informative as is required (for the current purposes of the exchange)'. Grice's formulation implies that one is supposed to make his contribution only as informative as is required. This upper bound of information is captured by the Quantity-2 Maxim, 'do not make your contribution more than is required in the context of the exchange'. In order to keep the two maxims apart, I will adopt the version of the Quantity-1 Maxim given in (2), which is in fact congruent with many reformulations of the Quantity-1 Maxim that have appeared in the literature. These include Harnish's (1976) Quantity-Quality Maxim: 'make the strongest relevant claim justifiab。