Abstract

this paper up front, together with the Peters one. Savitch et al. note that "it can serve as another introductory article to the whole volume" (p.22), but the reader may miss this pointer, and judge the paper solely by its title, which would be a great pity. And while I am on the topic, I would like to call attention to a point made by Wasow, which has only very recently resurfaced in the literature. Many formal languages are obviously too simple to serve as possible human languages. Wasow cites finite languages and infinite mirror-image language_.s as examples, and there are many others, e.g., all languages over one letter. However, linguistic models have always allowed such degenerate cases. Chomsky was no sooner done arguing that human languages were neither finite nor regular than he introduced TGs, which clearly generate all such languages. Moreover, as noted, Wasow's position implies the all-important conclusion that the formal complexity of human languages cannot be measured in terms of the Chomsky hierarchy

Citations