Abstract
It is common practice in formal semantics to assume that the context specifies an assignment of values to variables and that the same variables that receive contextually salient values when they occur free may also be bound by quantifiers and λ s. These assumptions are at work to provide a unified account of free and bound uses of third person pronouns, namely one by which the same lexical item is involved in both uses. One way to pursue this account is to treat quantifiers and λ s as monsters in Kaplan's sense. We argue that this move should be avoided and explore an alternative route based on the idea that there is a variable assignment coordinate in the context and a variable assignment coordinate in the circumstance of evaluation, with the definition of truth in context identifying them. One fundamental challenge that arises in pursuing a unified account is to explain the difference in the way the gender presuppositions of bound and free pronouns project. The proposal that emerges from the attempt to meet this challenge is a non-indexical account of free third person pronouns and a new conception of the role and structure of assignment functions.
Highlights
It is common practice in formal semantics to assume that the context specifies an assignment of values to variables and that the same variables that receive contextually salient values when they occur free may be bound by quantifiers and λ s
While some of the linguistic constructions mentioned above admit of alternative accounts that make no reference to variables of this kind, in the case of third person pronouns a powerful reason for making assumptions (i)-(ii) is the desirability of accounting for their bound and indexical occurrences by supposing that the same lexical item is involved
We saw in the previous section that, if the assignment is only present as a coordinate of the context, we end up with a monstrous semantics for quantifiers and the λ operator
Summary
It is common practice in formal semantics to assume the following:. (i) context specifies, among other pieces of information, an assignment of values to variables. If one analyzes third person pronouns in this way, a natural move (one which, for example, Heim & Kratzer 1998: 243 seem to suggest) is to regard the variable assignment involved in spelling out the interpretations of the quantifiers and the λ -operator as a coordinate of the context of utterance. If this move is made, the semantic clause for universally quantified formulae should have form (3), where cg is the assignment of the context c:. The λ -operator in (4) qualifies as a monster
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have