Lexicography: Theory and Practice

Lexicography: Theory and Practice1 Patrick Hanks At is an exhilarating time for lexicography. Issues raised over 200 years ago by Samuel Johnson are at last being debated. A resurgence ofempiricism in linguistics and a rehabilitation ofthe statistical study of language, after attacks by Chomsky and others, are engaging some of the best minds in universities and research institutions throughout the world. Questions are now being addressed that, as it happens, are of the greatest importance to lexicographers. How exactly do words work? How are they fitted together to create meanings ? Why is it so difficult to fit the details of verbal behavior neatly into some of the syntactic models that have been proposed? What counts as a meaning? Do words have more than one meaning, and if so, how are we to distinguish one meaning from another? Is word meaning variable, and if so, what is meant by variability? Above all, the development over the last ten years of large computerized corpora , with rapid means ofsearching and sorting, has yielded an abundance of evidence that Samuel Johnson, Noah Webster, and James Murray could only dream of, but—horror!—it turns out that some of the assertions made in our major dictionaries are at variance with the facts of everyday usage. Among those who are at the forefront of the academic study of the lexicon is Anna Wierzbicka. She draws on a great wealth of learning in the European tradition, reaching back through Leibniz to Aristotle . In previous writings (particularly Wierzbicka 1985 and 1987), she proposes definitions for ordinary words that are so bizarre that one's first reaction is that she must be doing something right. Her work raises profound questions about the nature of word meaning and of dictionary definition. Is there an "invariant" truth about the meaning ofeach word, which it is the lexicographer's task to capture? Is each definition ofa polysemous word discrete? Or is the whole phe- 98Patrick Hanks nomenon ofword meaning so fluid and variable that it cannot be captured at all, but only hinted at in sequences of half-truths? Some of the biggest headaches for monolingual lexicography lie in the area of accounting accurately for highly variable phenomena . Iftheory could offer a practical solution to this problem, it would earn the undying gratitude of the practitioners. Unfortunately, Wierzbicka's paper, no doubt an attempt to simplify the issues and concentrate on essentials, virtually ignores such problems. Indeed, she goes further: she describes hedges (the lexicographer's informal tools for indicating vagueness: words such as especially, usually, often) as "visible signs of indecision and analytical failure." This is very odd, for elsewhere (Wierzbicka 1985) she urges: An adequate definition ofa vague concept must aim not at precision but at vagueness: it must aim at precisely that level of vagueness which characterizes the concept itself. Apparently, natural-language hedges are not to be used to indicate vagueness. The point is of great importance to lexicography; it is a pity that Wierzbicka does not make more ofit here, in discussing what linguistic (lexical) theory might offer lexicography. Let us now turn to what she actually does say. I shall discuss just seven of Wierzbicka's strictures on dictionaries in general, before concluding with a few remarks on the context in which lexicography takes place. 1. Dictionariesfail to "capture the invariant." This point is at the heart of Wierzbicka's case against conventional dictionaries. It presupposes that there is an invariant to be captured . What might this be? It cannot be something that is stable over time. Linguists since Saussure (and indeedJohnson) have insisted on the arbitrary and unstable nature of the connection between signifier and signified. Almost every page ofa major historical dictionary such as the OED has some improbable (but true) story to tell about changes in word meaning . A superficial interpretation of "capturing the invariant" cannot, therefore, be what Wierzbicka has in mind. Instead, she seems to be alluding to a concept that has attracted a diverse array of philosophers , bishops, psycholinguists, and others from Aristotle to Fodor, by way of Leibniz, Wilkins, and the young Wittgenstein. She associates the invariant with philosophical notions of truth when she writes: Lexicography: Theory and Practice99...