Universal Grammar versus linguistic diversity

Abstract

The target article by Evans and Levinson (E&L) concludes that there is little, if any, empirical evidence corresponding to 'immutable' principles of Universal Grammar. It would be surprising indeed if Universal Grammar failed to leave its imprint on languages around the globe, but it is a misunderstanding of the theory to limit the search for linguistic universals to counting up construction-types across languages, as E&L do in the target article. Here is an instructive quote from Chomsky (1965:6). "The grammar of a particular language ... is to be supplemented by a universal grammar that ... expresses the deep-seated regularities which, being universal, are omitted from the grammar itself. Therefore it is quite proper for a grammar to discuss only exceptions and irregularities in detail. It is only when supplemented by a universal grammar that the grammar of a language provides a full account of the speaker-hearer's competence." As this quote makes clear, in providing descriptions of particular languages, linguists may concentrate on what makes a language special, and not on what it has in common with other languages. Universal Grammar does not, however, attempt to account for exceptions and irregularities. Therefore, little is gained in arguing against this theory by pointing out that human languages exhibit a lot of individual differences. Any challenge to Universal Grammar requires more than this. So what is Universal Grammar, if it is not a theory of constructions that appear in the final state of particular human languages? It is a theory about the initial state of the human faculty for language. Universal Grammar (UG) circumscribes the kinds of hypotheses that language learners can formulate. To cite a famous example constructed by Chomsky almost 40 years ago, UG maintains that children can hypothesize structure-dependent operations, but not structure-independent operations. To illustrate how structure-dependent principles differ from structure-independent operations, Chomsky offers, as an illustration, how Yes/No questions could be related to their declarative counterparts. In view of E&L's conclusions that there are no immutable linguistic principles, it is worth reviewing the claim that structure-dependence is one such principle.