Cumulative MOST

In German, Scandinavian, and Basque, proportional MOST can combine not only with count plural NPs but also with mass NPs, and correlatively allows collective predicates in the nuclear scope. After arguing against Hackl’s and Hoeksema’s superlative-based analyses, we propose that this “cumulative” MOST (MOSTcum) is a quantificational determiner, which nevertheless differs from MOSTdist: whereas MOSTdist compares the cardinalities of two sets, MOSTcum compares the measures of two entities. The analysis is extended to the proportional reading of MORE in Bulgarian. We also examined majority quantifiers that are not morphologically related to the superlative of MANY/MUCH but nevertheless show the distribution of MOSTcum: Japanese hotondo and Chinese dabufen. This chapter is theoretically interesting for at least two reasons: it deals with the under-studied area of mass quantification, and attempts an explanation for the correlation between (in)definiteness and the various readings of MOST (superlative, distributive proportional quantifier, cumulative proportional quantifier).

Introduction

We start by presenting the main empirical result of this book: a crosslinguistic typology of MOST, which distinguishes between two types of proportional MOST (MOSTdist and MOSTcum) that can occur in both non-partitives and partitives and two types of MOST that are specialized for partitives. We proceed by presenting our main syntactic assumptions and the semantic analyses we proposed for each of the MOSTs we identified.The crucial semantic distinction is between a “distributive” MOST, which compares cardinalities of sets of atoms, and a “cumulative” MOST, which involves measuring plural and mass entities with respect to a whole. Then, we discuss the main semantic analyses proposed in the literature. The last section presents the empirical evidence against Hackl’s (2009) superlative analysis of most.

Distributive MOST

Romanian, Hungarian, and Icelandic are languages in which MOST allows the proportional interpretation when combined with count NPs but not when combined with mass NPs. In English, the same generalization is found with non-generic NP restrictors. In §2.2 the generic NP restrictors found with the English most are analyzed as kind-referring DPs. Section 2.3 argues that the observed ban on mass NPs cannot be explained by Hackl’s (2009) superlative analysis, nor by Matthewson’s (2001) entity-restrictor MOST. We propose that (in the relevant languages) proportional MOST is to be analyzed as a distributive quantificational Determiner. We show that the observed ban on mass NPs correlates with a ban on collective quantification. Finally, we provide syntactic evidence for the hypothesis that in Romanian and English the proportional MOST sits in Spec,DP and D°, respectively. It thus appears that the quantificational-determiner semantics of MOST correlates with the highest syntactic level inside the DP.

Majority quantifiers based on nouns

Strings of the type the largest/LARGER PART or (THE) LARGE PART, together with MAJORITY nouns, are the most widespread means of expressing majority judgments. We take this to constitute evidence in favor of a compositional analysis, which builds the majority reading by combining the superlative form of the adjective LARGE (or the comparative or positive forms, in some languages) with the functional noun PART, which introduces an unspecified binary partition. We propose a possible extension of this analysis to abstract nouns of the MAJORITY type. We also discuss a peculiar type of relative superlative reading allowed by MAJORITY nouns (in addition to their majority reading), which is identical to the one observed by Kotek et al. (2011) for most of in English. We finally offer three case studies for the majority quantifiers found in Hindi, Latin, and Syrian Arabic.

MOST in partitives

We explain why MOSTdist maintains in partitives the constraints it obeys in non-partitives, whereas the distribution of MOSTcum is larger in partitives, where quantification over parts of atoms is allowed (in addition to quantification over pluralities and mass entities). We then demonstrate the existence of two MOSTs that are specialized for partitives, MOSTRP and MOSTDP, which respectively take of-DPs (analyzed as Zamparelli’s RPs “residue phrases”) and of-less DPs as complements. The distinction between MOSTRP and MOSTDP allows us to formulate interesting crosslinguistic tendencies: whereas MOST preferentially takes RP complements, ALL shows a strong preference for DP complements. We take this as evidence in favor of the non-quantificational status of ALL, which we analyze (following Križ 2016) as a homogeneity remover. We finally argue that the peculiar superlative reading of partitive MOST observed by Kotek et al. (2011; 2015) does not constitute evidence in favor of a superlative analysis of majority MOST.

Conclusions

This chapter summarizes our main empirical findings. We then make some suggestions regarding deep issues such as: Why is it the case that we have two majority MOSTs that can occur in both non-partitives and partitives (MOSTdist and MOSTcum) as well as two majority MOSTs that can occur only in partitives (MOSTRP and MOSTDP)? For each of these cases we demonstrate that the difference in LF representations stems from syntactic distinctions (NPs vs MeasPs as complements of MOSTdist and MOSTcum, respectively; RP and DP complements for partitive MOST). Comparisons with ALL further refine the proposals. We reiterate one of the main claims of the book: the fact that none of the majority MOSTs can be analyzed as a superlative. We propose that majority MOST evolved from the superlative of MOST via grammaticalization. We finally observe that the majority MOSTs that allow mass quantification invalidate some generalizations regarding “necessarily quantificational” DPs.