Object Wide Scope and Semantic Trees

Abstract

UCLA 19Contrary to what is often assumed, object wide scope (OWS)readings are in fact severely restricted. Liu (1990) Two referees read at least three abstracts 11 'there are at least three abstracts such that each of them was read by two (possibly different) referees' d. Two referees read exactly three abstracts * 'there are exactly three abstracts such that each of them was read by two (possibly diff erent) referees' e. Two referees read more than three abstracts * 'there are more than three abstracts such that each of them was read by two (possibly diff erent) referees' f . Two referees read fewer than three abstracts 2We follow Liu in restric ting ourselves to Don-partitive NPs. While no account of OWS is complete without considering partitives, the judgements concerning their scope taking abilities are delicate and more empirical research is needed before any theoretical conclusions can be reached.© 1993 by Dorit Ben-Shalom Utpal Labiri and Adam Wyner (eds.), SALT 111 ,[19][20][21][22][23][24][25][26][27][28][29][30][31][32][33][34][35][36][37] Ithaca, N.Y.: Cornell University. 20 numeral: a, some or a number like three. (1c)-(1£) are examples of NPs whose determiner is a modified numeral: more than three, exactly three, fewer than three, etc. (1g) is an NP whose determiner is no. Crucially, the readings that are missing in (1) are not incoherent. For example, the missing OWS read ing of Two referees read at least three abstracts is the preferred reading of the corresponding passive and negative fronting with subject-aux inversion in (2a) and (2b) (from Szabolcsi and Zwarts(1992)).(2) a.b.Fewer than three abstracts were read by two referees Fewer than three abstracts did two referees readThere are two obvious questions to be answered now. First, what is the positive property that characterizes noun phrases that are able to take OWS? And second, whatever this property turns out to be, what is it about OWS that requires the presence of this property? In this paper we suggest that the answers derive from the following specific properties of the syntax-semantics interface:f.An OWS reading is available only if the object NP is interpreted as a principal filter. The distinctive syntactic property of an OWS reading is that it involves inverse scope: the scope order does not match the order of argument positions (which is roughly the S-structure c-command order of the arguments). Readings involving inverse scope may be derived in either of two ways: (i) Using unary quantifiers, which are applied in an order that reverses their lexical order, or (ii) Using an n-ary quantifier, which applies to the denotation of the verb to yield the denotation of the sentence in a single step. It will be argued that (i), the application of unary quantifiers in a non-lexical order is only possible if lexical argument order is clearly overridden by a syntactic process, e.g. surface movement. This restriction will be motivated empirically, with reference to contrasts like those between (If) and (2b), as well as theoret ically, wi...

How Clause-bounded is the Scope of Universals?

“…An adequate theory of scope has to account for this difference. Third, the contrast between the scope limits on distributive universals and non-distributive quantifiers on the one hand, and that between indefinites and quantificational DPs on the other supports the claim, made most recently in Szabolcsi (1995), that different DPs scope differently, depending on their semantic characteristics (see also Liu (1991), Ben-Shalom (1993).) Finally, the sensitivity of inverse scope to semantic factors of the type discussed above casts serious doubt on the possibility of a purely configurational treatment of scope.…”

Computing Quantifier Scope

“…When the function is increasing, we check to see that every member of the witness is a member of the denotation of the predicate. Similar ideas are proposed by Ben-Shalom 1993 andSzabolcsi 1996.…”

A Derivational Approach to Scope Interaction between Wh-phrases and Quantifiers

“…This is illustrated in (7). Watanabe (2000) further argues that wh-QP scope interaction, and in particular the distributive interpretation, is mediated by the absorption of a Q-feature (Ben-Shalom 1993). Translated into Minimalist terms, Q-feature absorption (henceforth, Q-absorption) depends on the possibility of checking of Q-features.…”