University of Durham

abstract

We offer independent, mathematical, support for Bickerton's thesis that in the evolution of language its functor (or predicate) /argument structure (henceforth: "f/a structure") emerged catastrophically (all in one go) rather than by piecemeal adaptation. Using Saussurean principles, we outline a formal proof that f/a structure is an emergent, necessary and unitary feature in the growth of any system of meaningful combinations of symbols. We also mention how the formal character of combinatorial evolution partially reinforces Bickerton's view that f/a gave expression to the thematic structure of co-operative skills which must already have been in existence.

The proof is part of a wider mathematico-philosophical study of how language, as a combinatorial system, might be understood as having evolved through structural growth, somewhat like a biological system (The inspiration came originally from Piaget, D'Arcy Thompson, and Rene Thom. But its formulation draws on Saussure, categorial logic, Kuratowski and upon Wittgenstein and Heidegger.)

Combination is treated here not as mere concatenation or juxtaposition but, following Saussure, as a richer activity which provides a mode of word differentiation, transcending phonology and generating categories. (In another register: as the construction of "ordered", i.e. well-differentiated, pairs.) We identify a simple procedure for initiating and recursively expanding combination, and trace the growing organisation it yields in a formal model. This study reveals that starting from scratch, without prior linguistic abilities, there is only one kind of way in which a system of meaningful word combinations could grow (though this has many alternative varieties): namely, towards a structure familiar as that of natural language. Like biological growth, combinatorial evolution follows a distinctive structural envelope, or geometry, but one determined by its internal organisational requirements, which are semantically motivated, not by prior structures or genes.

This paper is confined to showing how f/a patterning emerges quite rapidly, at a second recursion, in the metamorphosis from proto-language to a semantically powerful system. A concrete illustration is supplied.

Our investigation also uncovers organisational reasons why language must be social. Any technique of combinatorial differentiation can be exerted in alternative, mutually exclusive, ways. It is arbitrary which is employed, and such an arbitrary selection can only be established by agreement in practice (Lewisian "conventions"). Consequently, language, a system of myriad, cumulative, densely nested, differentiations, could only be assembled, like a dance, among participants already skilled in the recognition and negotiation of co-operative practices. That is, in communities of just the kind which Bickerton, for separate reasons, envisages.

 Conference site: http://www.infres.enst.fr/confs/evolang/