This chapter argues that both indefinites and universals call for a distinction between existential scope and distributive scope. After motivating the distinction it focuses on existential scope; matters of distributivity are taken up in the next chapter.

Picking up the thread from §5.1 we start with the well-known case of indefinites, motivate the existential vs. distributive scope distinction, and explore the choice-functional implementation in some detail. We then go on to argue that every NP-type universals warrant the same distinction, and make several steps towards unifying their treatment with that of indefinites.

Indefinites

No such thing as “the scope” of an indefinite

The well-known claim (e.g. May 1977) that quantifier scope is clausebounded is based on examples like the following:

1. (1) A colleague believes that every paper of mine contains an error.

2. #‘for every paper of mine there is a potentially different colleague who believes that it contains an error’

As was mentioned in §5.1, Fodor and Sag (1981) noticed that the scope of singular indefinites is not clause-bounded, see (2); it even escapes islands for movement, such as a Complex DP Island, see (3):

1. (2) Each colleague believes that a paper of mine contains an error. ok ‘there is a paper of mine such that each colleague believes it contains an error’

2. Each colleague overheard the rumor that a paper of mine contains an error.

3. ok ‘there is a paper of mine such that each colleague overheard the rumor that it contains an error’

In fact, they proposed that if an indefinite escapes an island it takes maximal scope.