Competing Motivations and Emergence: Explaining Implicational Hierarchies

This paper is part of an attempt to provide a theoretical framework for various kinds of functional-typological explanations that is also broadly compatible with the nativist approach to language acquisition (Kirby 1994b). The approach relies crucially on the idea of selective adaptation: ": : : the sorts of explanations made by typologists are essentially adaptive ones: language structures are the way they are because of their adaptation to the function (s) of language : : : In this respect linguistics also parallels biology. However, the philosophical analogy between linguistic functional explanations and biological adaptation is not always fully worked out in linguistics." (Croft 1993) The framework put forward in this paper is an explicit description of the adaptive process. In other words it is an attempt to solve the problem stated below (also discussed in other forms by Hall 1988 and Bybee 1988): (1) The problem of linkage: How does a universal property of language use give rise to a restriction on the distribution of occurring languages in the space of possible languages? For the purposes of this paper, the linkage in question is that between properties of innate language processing mechanismsand hierarchical universals. These are chains of implicational universals of the form (P 1 ! P 2 )&(P 2 ! P 3 )& : : : &(P n\Gamma1 ! Pn) commonly written P n ? P n\Gamma1 ? : : : ? P 3 ? P 2 ? P 1 , where P i is associated with some binary property whose value can be determined for all languages. Of course, these hierarchies predict that only the following language types will occur (where + indicates that the language has the propert