The indefinite article – Indefinite? – Article?

Perlmutter (1970) argued that the indefinite article is categorically different from the definite article and proposed that it is a clitic version of the numeral "one". But there are, as Perlmutter himself pointed out, instances of "a" as well as of "one" that don't seem to have the semantics of the numeral. Hence a divorce of "a" (and of "one") from "numeral"-hood is called for. Furthermore, there are instances of what looks like the indefinite article (e.g., German "ein" or its Dutch, etc. counterpart) which occur in contexts from which the indefinite article is supposed to be excluded: with plural nouns, with non-count nouns, in definite noun phrases, etc. This state of affairs was addressed by Bennis et al. (1998), and others since, by reference to a so-called 'spurious article,' homophonous with the traditional indefinite article "een/ein". The goal of the present paper is twofold: First of all, I argue that German "ein" is not always an indefinite article,' and, pursuing the idea that there is only one "ein", it is hence never an indefinite article.' Secondly, I explore some consequences for the structural representation of certain function words which contain "ein" as one of their components, in particular "kein" as well as its English counterpart "no". The discussion promotes a strongly non-lexicalist view, advocating a syntactic derivation of function words, including movement.