Replies to Discussants

REBUTTAL Replies to Discussants Anna Wierzbicka X am very grateful to the discussants for their thoughtful comments . For reasons of space, I will not be able to take up all the issues that they have raised. I think the most fruitful approach is to focus on specific lexicological points (bearing in mind the general issues that they illustrate). 1. JURIJ APRESJAN Apresjan's paper presents the author's own lexicographic theory , which I regard as one of the most important contributions to modern lexicography. I will confine my comments to his remarks directly concerning my own paper. I cannot accept the idea that "semantic primitives are language-specific, that is, not cross-culturally valid." If they were language-specific, we would not be able to compare meanings across cultures, and I have tried to show in a number of publications that in fact they can be compared (see in particular Wierzbicka 1992a; also Goddard and Wierzbicka (ed.), forthcoming). For bilingual lexicography , a shared set of primitives is essential, for how could culturespecific concepts of one language be made intelligible to speakers of another unless in terms of shared conceptual elements? On the other hand, the suggestion that universal semantic primitives should be called "near primitives" is quite acceptable to me: since every linguistic system is unique, and since in any language the indefinable words are embedded in a different system, there can be no "absolute" equivalents of elements across language boundaries (see Wierzbicka 1989a and 1989b). Nonetheless, I believe that there is no identifiable semantic difference between, for example, Russian se- 140Anna Wierzbicka mantic primitives such as znat', skazat', orxotet' and their English counterparts know, say, and want (for discussion and justification, see Wierzbicka 1989a and 1991b). Apresjan says that: the English verb want, presumably a primitive, shares a common semantic component with wuh. Yet this common component cannot be verbalized in English. So want may be considered a primitive only in the sense that it cannot be further decomposed . To me, however, this is precisely what defines a primitive: that it cannot be further decomposed. As for the shared part of want and wish, in my view it can be verbalized, because wish, which is semantically complex, can be defined in terms of want, which is simple (cf. Wierzbicka 1987). Roughly: / wuh it would happen. = I say: I want it to happen if it could happen because of this it would be good I don't say: it will happen because of this Furthermore, I cannot agree with the statement that "want suggests the idea of need whereas xotet' [its closest Russian equivalent, A.W.] suggests the idea of intention." In my view, there is no identifiable semantic difference between the English and the Russian sentences below: I want to work/rest/play. Ja xocu rabotat'/otdyxat'/igrat'. The fact that in Russianja xocu ? want' is used more readily with respect to uncontrolled impulses and urges than / want is in English can be accounted for either by positing a second meaning for the Russian verb (as Apresjan himself suggests) or by reference to independently established cultural differences (an alternative that I would favor; see Wierzbicka 1992a). 2. D. A. CRUSE Cruse criticizes my definitions for not fitting neatly into either of the two most popular models—the "classical" model, which assumes a list of necessary and sufficient criteria, and the "prototype" Replies to Discussants141 model. He adds that "the examples offered seem to fall uncomfortably between the two. ..." But I do not accept the assumption that all a semanticist can do is to choose one of these two models. In fact, I regard both of these models as inadequate, and in a number of publications I have advanced and illustrated a third model that distinguishes between necessary conceptual components and necessary features of denotata and that seeks in many cases (wherever appropriate ) to incorporate prototypes into definitions. For example, should a definition of bird refer to flying or shouldn't it? Not all birds can fly, but as I have argued in Lexicography and Conceptual Analysis (LCA, Wierzbicka 1985) 'flying' is a necessary component of the concept 'bird'. A proper definition of bird should accommodate both these facts: that...