Functional explanation in linguistics and the origins of language

Abstract

In these remarks, I shall place scare quotes around the terms relating to ‘function’ (‘functional’, ‘functionalist’, etc.), because it is misleading to present this term as in some way opposed to ‘formal’ (etc.), and I do not accept Newmeyer’s lumping strategy in regard to those linguists whose central project is not the formal calculus of ‘an autonomous grammar’. Furthermore, when referring to Newmeyer’s sense of the word ‘grammar’, I shall likewise bracket it with quote marks and shall always supply the indefinite article ‘a’ and the adjective ‘autonomous’ (viz. ‘an autonomous grammar’). I do this because there is a constant danger of silently moving from his understanding of ‘grammar’ as an independent, discrete, mentally represented, innate entity to a more fuzzy conception of grammar as a mass noun, a cover term for a wide and heterogeneous spectrum of regularities in speech as noticed by a linguist. By slipping back and forth between these two radically different conceptions (‘a grammar’ versus ‘grammar’), Newmeyer is sometimes able to make it seem as if there is or should be agreement where there is none and cannot be any. Much of Newmcyer’s position paper is devoted to the speculative hypothesis that ‘an autonomous grammar’ is an outcome of natural selection. Since any number of evolutionary scenarios might be constructed for language with equal validity, each compatible with a radically different set of assumptions about evolution, language, innateness, etc. it is highly unlikely that sufficient agreement about the meaning and use of terms could be reached to even debate this proposition. For example, in his assertion: ‘ . . . studies of ape intclligcnce suggest that prchominids possessed a surprisingly sophisticated level of mental representation’, one is hard put to find a single content-word that is not open to dispute. With such fundamental disagreement over vocabulary, I cannot see that the controversy over ‘an autonomous grammar’ is illuminated in any way by conjecture of this kind. Newmeyer’s version of natural selection offers nothing over Chomsky’s version of creationism. What is most disturbing to me is Newmeyer’s unwillingness to understand that in the eyes of the diverse linguists whom he lumps together as ‘functionalists’ or ‘functional grammarians’ many or all of the tacit assumptions he works with are suspect. In fact, Newmeyer’s view of ‘functionalism’ would render it distinct only in directionality from standard autonomous grammar: in this view, the forms of a language would be isomorphic to functions, and it would not matter much whether you described the functions first and then matched forms to them, or proceeded by developing a calculus of a grammar first and then (if you felt like it) showing how the previously arrived at forms were used in real-life situations. No paradigmatic gulf would separate formalists from ‘functionalists’, only disagreement as to the degree to which statements about functional correlates of