Introduction: Grammar as Historical Accident

Consider the details of how grammar (particularly syntax) operates in contemporary human languages: how phrases and clauses are structured, how they can be combined and subordinated, and how some elements can control or govern others. For many of these details, no off-the-cuff functional explanation suggests itself. For example, why can Algernon be interpreted as coreferential with he or him easily in (1)–(3) but not easily in (4) (Langacker 1969)?

1. While Algernon wasn't looking, Penelope bit him in the leg.

2. While he wasn't looking, Penelope bit Algernon in the leg.

3. Penelope bit Algernon in the leg while he wasn't looking.

4. Penelope bit him in the leg while Algernon wasn't looking.

The complexity of such details, and the lack of immediately obvious explanations for them, is the main reason why all versions of grammatical theory have become such elaborate and often intimidating edifices in recent years.

Alinguist interested in grammatical evolution must confront the issue of adaptation. Has grammar reached its present elaborated state because people who use this sort of grammar have been more efficient at reproducing themselves than have other people? Directly contrary answers have been given. ‘No’, says Chomsky, as cited by Newmeyer (1998): adaptation has little if anything to contribute to expaining why grammar is the way it is. This view is echoed by Lightfoot (1999, this volume) and in part by Bickerton (1990, 1995, 1998, this volume).