The thirteen articles gathered in this volume are revised versions of papers delivered at the Indefinites and Weak Quantifiers conference, 6–8 January 2005, Brussels. The collection is excellent and the work provided by the editors must be commended, including their very comprehensive introduction, even though it is misleading in its claim that the discussion of noun incorporation in incorporating languages begins with Mithun (Reference Mithun1984), on the one hand, and Sadock (Reference Sadock1980) and Baker (Reference Baker1988), on the other. As is well known, noun incorporation has a rich tradition in Amerindian linguistics: even the famous debate between Kroeber and Sapir, nearly 100 years old, was anteceded by other papers on this topic. In short, the discussion on noun incorporation began a long time before the 1980s.
The volume may appear oddly titled for two reasons. First, although I understand that the authors intend ‘non-definiteness’ to cover both standard indefinites and bare objects (the latter behaving rather differently from the former, so much so that they can hardly be called indefinites), I nevertheless find the term rather clumsy. Second, since the book is clearly not just about plurals, it is rather odd that ‘plurality’ has made it to the title of the entire volume at the expense of other relevant terms discussed in the book (such as aspect and numeral indefinites).
The papers are organized in three main parts. Part I, ‘Non-definiteness, plurality, and incorporation’, contains five chapters. The first paper, Greg Carlson's ‘The meaningful bounds of incorporation’, discusses the stable semantic properties associated with semantic incorporation: (i) the incorporated nominal is an indefinite rather than a definite or quantified phrase; (ii) the incorporated nominal is non-specific; (iii) the incorporated nominal receives only narrow scope in relation with other logical operators in the same sentence; (iv) the incorporated nominal is interpreted as an existential and not as a generic indefinite; (v) the verbs that allow for incorporation are stage-level verbs, never individual-level predicates; and (vi) the incorporated nominal is number-neutral. Carlson explores the possibility that a further criterion for setting the outer limits of what counts as semantically incorporated should be considered, namely ‘restrictiveness’ or in other words typicality. Examples such as Mark attended class and Mary took the train to Brussels, which involve typical activities, thus qualify as instances of semantic incorporation, but the famous case of bare plurals in languages like English, as in John ate apples, might not after all be a case of semantic incorporation, since bare plural sentences are not restricted to typical activities.
In ‘Bare nouns, number and types of incorporation’, Carmen Dobrovie-Sorin, Tonia Bleam & M. Teresa Espinal build on the first author's previous work in arguing that bare plurals (and mass nouns) are numeral phrases (NumPs), while bare (count) singulars are simply noun phrases (NPs). In Romanian and Spanish, bare singulars in object position undergo syntactic pseudo-incorporation, whereas bare plurals undergo semantic incorporation. Yet both kinds of nominals are property-denoting, of type <e,t>.
Donka F. Farkas in her paper, ‘The unmarked determiner’, puts forward an interesting hypothesis according to which various determiners impose different interpretation constraints on the variable associated with the Determiner Phrase (DP) that they head. These constraints are treated as a set of privative features, among which the features [Pl], [Def] and [Part] should be included. Determiners and the NP that they select are taken to vary cross-linguistically according to whether they have all, some or none of these features. Farkas claims that the English determiner a(n), French un(e), Romanian un/o, and Hungarian egy are unmarked with respect to all of these determiner features. The unmarked nature of these determiners explains their free distribution and varying interpretation.
In their paper entitled ‘Bare objects in Korean: (Pseudo-)incorporation and (in)definiteness’, Song-Nim Kwon & Anne Zribi-Hertz study Korean object phrases that appear without the accusative case marker (l)eul. It is shown that these bare objects have all the characteristics of semantically incorporated nominals, even though they are fully referential. In order to reconcile the apparent contradiction that we face when dealing with these objects (incorporated nominals are not usually referential), the authors propose that semantic deficiency – leading to semantic incorporation – is or may be independent from referentiality.
Part I closes with ‘Bare plurals in object position: Which verbs fail to give existential readings, and why?’ by Sheila Glasbey. The author closely examines the predicates that fail to allow existential readings for their bare plural objects and claims that they constitute a narrower set of verbs than the class of individual-level predicates. She identifies this subset as ‘psychological verbs with experiencer subjects’ (133), distinguishing between verbal predicates that have an eventuality argument and those that do not. On this account, an existential reading is made possible by the presence of a localising situation, which may be provided either by the eventuality argument of the verb or by an appropriate context. Glasbey further proposes that the psychological verbs with experiencer subjects are distinctive among verbs in not possessing an eventuality argument. This is said to account for the absence of existential readings for their bare plural objects.
Part II, ‘Alternatives to the incorporation approach: Relating non-definiteness and plurality to aspect’, which contains three papers, opens with Henriëtte de Swart's article entitled ‘Aspectual implications of the semantics of plural indefinites’. The paper shows that, whereas both indefinite singular NPs (un N) and definite singular and plural NPs (le N/les N) lead to telicity or terminativity, French plural indefinites (des N) are aspectually characterized as atelic or durative, and are the only NPs that allow bare habitual readings, as in Paul répare des bicyclettes ‘Paul repairs bikes’. De Swart treats des NPs as plural indefinites that lack determined reference and weak quantificational force. This combination, it is claimed, leads to atelicity because there is nothing in the semantics of des NPs that can bound the mapping from individuals to events. According to de Swart, this combination also allows bare habitual readings because there is nothing that blocks the combination of a cumulative interpretation with a one-to-one relation between individuals and events.
The introduction as well as the title of part II suggests that the editors treat aspect as a serious alternative to semantic incorporation. It is not clear, however, that the aspectual alternative is a real alternative, since it does not explain all the properties associated with bare nominals. While it may account for the scope properties of bare arguments rather straightforwardly, it has nothing to say about properties such as typicality (as discussed by Carlson) or the presence/absence of referentiality. It appears that the semantic incorporation account and the aspectual analysis are in complementary distribution. This is the direction pursued by Borer's (Reference Borer2005a, Reference Borerb) influential work, references to which are surprisingly missing from most, if not all, the contributions in this volume.
The next paper in part II is Brenda Laca's ‘Indefinites, quantifiers and pluractionals: What scope effects tell us about event pluralities’, which examines the way in which two Spanish aspectual periphrases that contribute pluractional aspectual operators interact with indefinite, quantified and plural noun phrases. Laca argues that temporal pluractional operators of the sort proposed by Van Geenhoven (2004) are suitable for capturing the temporal structure of the derived eventuality descriptions, but that ‘they should not be allowed to enter into scopal interactions with nominal arguments’ (191).
Part II closes with Jacqueline Guéron's ‘Generic sentences and bare plurals’, which claims that generic sentences differ from non-generic sentences only by the absence of a positively indexed Tense node. A predicative sentence with a positive Tense index denotes a property of an individual presupposed as existing in the discourse world. A predicative sentence with a zero Tense index denotes a property not of an individual but of the mental concept associated with a lexical item. Guéron further describes bare plurals in aspectual terms, i.e. as unbounded or imperfective nominals parallel to imperfective verbs in the sentential domain. These nominals are viewed as associated with an extension in time and space but lack spatio-temporal boundaries.
Part III, ‘Numerical non-definites’, which contains five papers, opens with Fred Landman's article entitled ‘Indefinite time phrases, in situ scope, and dual-perspective intensionality’. Landman argues that indefinite time expressions such as three times, as in Dafna jumped three times, enter the semantic interpretation directly as modifiers. This analysis treats indefinite time phrases on a par with intensional adverbial modifiers. At first sight this approach may seem problematic since these time expressions do not show the intensionality of modal adverbials, but Landman argues that a deeper form of intensionality, namely dual-perspective intensionality, is involved.
In ‘Identity of the domain of quantification for numerals’, Mana Kobuchi-Philip claims that numeral quantification in a non-classifier language such as English actually involves phonetically null classifiers. It is argued that this analysis provides two main advantages over prior analyses of English sentences containing numeral quantifiers and classifiers such as flock within the framework of a contemporary, lattice-theoretical approach to plurality. The proposal is also claimed to achieve a unified analysis of numeral quantification in overt-classifier languages such as Japanese and null-classifier languages such as English.
In ‘Scope shift with numeral indefinites: Syntax or processing?’, Tanya Reinhart argues that, although it is sometimes difficult for numeral indefinite plurals to scope out (e.g. for Four guests sleep in two rooms to be interpreted as involving eight guests), in some cases scoping out of the numeral indefinite plural is easier (as in Four guests sleep together in two rooms). This shows that the restrictions on the scoping of numerals are not absolute, but rather depend on the relations of the given derivation to other possible derivations or interpretations. Building on her previous work, Reinhart argues that the size of the reference set is important for the computational cost involved in comparing derivations. When two plural numerals are in the same sentence, the reference set will have five members due to the availability of a distributive interpretation of the subject, combined with a plural-numeral internal argument of the choice-function type. Since a reference set with five members appears to be beyond what adults can hold in working memory and thus the computation required to license the scope shift derivation cannot be completed, only three members are activated in the case where the subject cannot be interpreted distributively (as in Four guests sleep together in two rooms). In sum, Reinhart argues that what appears to be a syntactic restriction on scope shift with numerals is better explained in terms of a processing failure.
In ‘Take “five”: The meaning and use of a number word’, Bart Geurts compares the neo-Gricean view of five, according to which five means ‘five or more’, with the so-called naive view where five just means ‘five’. According to the author, the primary meaning of five is that of an ‘exact’ quantifier, from which an ‘exact’ predicate meaning and an ‘at least’ quantifier meaning are derivable by standard type-shifting rules.
Finally, Francis Corblin's ‘The semantics of paranumerals’ argues that, although their truth-conditions are often the same, numerals (n), numerical comparatives (more/less than n), and set comparators (at least/at most) should be distinguished on the basis of some of their dynamic properties, namely anaphora and apposition. Three claims are made: (i) bare numerals introduce into the representation a set of exactly n individuals; (ii) numerical comparatives introduce only the maximal set of individuals Σx satisfying the conjunction of the NP and VP constraints, and compare the cardinality of this set to n; (iii) set comparators introduce two sets into the representation, Σx and a ‘witness set’, the existence of which is asserted and which is constrained as a set of n Xs, X being the descriptive content of the NP.
The volume closes with an excellent and very detailed index. I recommend this book to anyone interested in the recent advances in the description and theory of bare plurals, bare singulars, and numeral phrases and in the relevance of aspect in relation to these expressions.