Elsevier

Lingua

Volume 181, October 2016, Pages 58-80
Lingua

The syntax of the Italian indefinite determiner dei

https://doi.org/10.1016/j.lingua.2016.05.001Get rights and content

Highlights

  • Italian dei-nominals do not contain an internal partitive structure.

  • dei is not a quantifier in spite of its meaning shared with alcuni ‘some’.

  • Italian dei-nominals are simple DPs.

  • -i in D realises number/gender features and is not kind-denoting.

  • de- in specDP provides indefinite meaning.

Abstract

This paper provides a syntactic analysis of the indefinite plural determiner dei in Italian, also called ‘partitive article’, as in Ho visto dei ragazzi (I saw de.art boys). It argues that dei is neither parallel to the inflected preposition dei in: Ho visto alcuni dei ragazzi (I saw some of-the boys), nor to the quantifier alcuni in: Ho visto alcuni ragazzi (I saw some boys), contra previous literature. We support a simple DP-analysis which takes dei as the plural counterpart of the indefinite article un, and claim that dei-nominals do not have a QPstructure, as proposed for quantified nominals in Cardinaletti and Giusti, 1992, Cardinaletti and Giusti, 2006, Cardinaletti and Giusti, in press. We propose that de- is an uninflected determiner in SpecDP, which cooccurs with an overt morpheme (-i) in D, realizing Gender and Number. This accounts for the apparent morphological similarity of this concord morpheme with the definite article, without the need to attribute referential or kind-denoting interpretation to it. The paper also investigates differences and similarities between dei-nominals and bare plurals.

Introduction

The main goal of this paper is to provide a syntactic analysis of a long standing problem, namely whether the apparent forms ‘of’+article like dei in (1a) and (1b) have the same structure or not:The possibility that dei ragazzi in (1a) has the same structure as dei ragazzi in (1b) is at first sight suggested by the observation that dei ragazzi appears in both structures and that in (1a), it can be paraphrased as alcuni ragazzi ‘some boys’. Previous analyses have approached the syntactic question from the semantic perspective, however reaching opposite conclusions.

On the one hand, Chierchia (1998) capitalises on the similar interpretation of pairs of sentences like (1a) and (1b) to claim that dei in (1a) starts in the position of dei in (1b) and ends up in the position of alcuni in (1b), as depicted in (2a), which is minimally different from (2b):

On the other hand, Storto (2003) shows that in some contexts, the nominal expressions in (1a) and (1b) have different interpretations, in particular (1a), opposite to (1b), does not imply the existence of a larger set, which is a defining property of the partitive interpretation, see Section 4.2 for further discussion. This fact has led Storto to suggest that dei in (1a) is in the same position as alcuni in (1b). Storto does not commit himself on the details of such structure.

Zamparelli (2008) rescues Chierchia's unified analysis in (2) by postulating that the partitive PP in (2a) necessarily embeds a kind-denoting DP, unlike the partitive PP in (2b), which embeds a definite DP. The difference between the two proposals resides in the relationship between the upper NP and the PP/RP (see (3)): in Chierchia's analysis (2), di is the head of the PP complement of the higher N, while according to Zamparelli (2008:319), di is an operator P in (3a) or R in (3b) which selects two arguments, NP in specPP in (3a) or in specRP in (3b) and DP in its complement (see (3)). Such an operator “returns the denotation of its specifier [NP] minus the denotation of its complement [DP] […]. The result of the subtraction is the set of all pluralities minus the largest one, which is the desired semantics for a partitive (see Barker, 1998)”:1

A common feature of the analyses by Chierchia, Storto and Zamparelli is to assume a biunivocal correspondence of structure and interpretation. Our contribution, instead, approaches the problem from the morpho-syntactic perspective. We agree with Zamparelli (2008) that indefinite dei in (1a) behaves differently from both partitive dei and indefinite alcuni in (1b) and provide further evidence for this. We also keep his proposal that the article -i in (1a) is not a definite determiner but is void of semantic import, similar to the article found in kind-denoting nominals in Italian. We will however show that this property can be captured in a simple-DP structure and that the much more complex raising-analysis in (3a) is unnecessary and is in fact excluded on both theory-internal and empirical reasons.

We propose that indefinite dei in (1a) is the determiner of an indefinite DP, as represented in (4a), parallel to a bare plural (4b). This immediately captures the differences with the PP appearing in a partitive construction as in (4d). The difference between indefinite dei and alcuni are captured by suggesting, following Cardinaletti and Giusti, 1992, Cardinaletti and Giusti, 2006, Cardinaletti and Giusti, in press, that alcuni is the Q head of a QP, selecting an indefinite bare plural DP as in (4c) and optionally a partitive PP as its second argument, as in (4d):2

Chierchia, Storto and Zamparelli claim that their proposal of indefinite dei should extend to mass indefinites such as del vino (‘wine’) in (5a), but they only discuss plural count nouns. In our view, this perfect parallel does not hold, for syntactic and semantic reasons. A parallel with (1) shows, for example, that mass nouns can occur with a quantifier such as poco, but not with alcuno in (5b):3

On the semantic side, the interpretation of singular/mass del is not the same as plural dei in negative existential clauses. The sentence in (1) is ambiguous between a wide scope interpretation of dei, compatible with the continuation in (6a) (which states that there are some boys that I did not see but there are others that I did see) and a narrow scope interpretation of dei, compatible with the continuation in (6b):The sentence in (7), with a singular mass noun, can instead only have a narrow scope reading of del, and the continuation in (7a) is unacceptable:4

An anonymous reviewer comments that in (7a), the continuation is incoherent because bare Prosecco and Cabernet may refer to instances of types of wine, thereby triggering the [quantity --> type] shift, which is one way of turning a mass denotation (like that of vino ‘wine’) into a countable one (cf. i vini della Francia ‘the wines of France’). This is exactly our point. In order to ensure the mass interpretation in (7a), we have inserted the option de.art, which does not improve its acceptability as a continuation of (7).

When the [quantity --> type] shift occurs, as in (8), the plural forms dei vini behaves like the plural count dei ragazzi in (6), and the continuation in (8a) is coherent with (8):Note that there is nothing in singular del vino that prevents it from taking scope over negation, as in (9). When del vino is extracted from the basic position, it can be interpreted as an unspecified quantity, as in (9a), or as a type, as in (9b), as witnessed by the continuation:

Whatever the semantic account of the different properties of (singular) unspecified quantity of matter in (7) and (plural) types in (8), our observation on the different scope taking properties of singular del and plural dei in object position seems to hold and to urge for an explanation.

On the morpho-syntactic side, it must be noted that in some central Italian varieties, such as Anconetano in (10), only plural indefinite dei is present, while singular indefinite del is missing:Note that in this variety, bare nouns are ungrammatical:

Narrow scope interpretation obtains with insertion of the definite article, which is then ambiguous between the definite and indefinite interpretation. This is also possible with plural count (12a) and is not limited to singular mass (12b):5

The semantic and syntactic evidence provided above suggests that no analysis of indefinite plural dei can directly extend to singular mass del. For this reason, in this work we limit our analysis to plural indefinite dei, leaving singular del to separate work.

The structure of the paper is as follows. Section 2 presents the raising analysis as rephrased by Zamparelli (2008) and points out two empirical problems with it. Section 3 presents our proposal, which claims that there is no need to assume a complex structure with internal raising. The rest of the paper shows how the many observations made in previous literature are accounted for by our simple-DP analysis. Section 4 gives a number of reasons supporting the claim that indefinite dei is different from P+art. Section 5 supports the claim that dei is not a quantifier. Section 6 presents differences and similarities with indefinite singular uno of which dei is the plural counterpart (cf. Renzi, 1982). Section 7 discusses the nature of the -i morpheme on dei and proposes that it is just concord features: it does not carry referential interpretation, nor does it require that the internal NP be interpreted as kind-denoting. Section 8 compares dei-nominals with bare plurals. Section 9 draws the conclusions.

Section snippets

The raising analysis

Chierchia (1998) and Zamparelli (2008) claim that the nominal expressions in (1a) and (1b) have the same structure, as in (13). We focus here on the common core of the two proposals, namely that dei is first formed inside a PP by incorporating the article i in the lower D into the preposition de and then moving the complex head dei into the higher D:

(13)a.[DP dei[NP dei[PP[P’ dei [DP i [NP ragazzi]]]](Chierchia, 1998:90)
b.[DP dei[NumP dei[PP [NP ragazzi][P’ dei [DP i [NP ragazzi]]]]](

The proposal

In this section, we spell out our structural analysis of indefinite dei. We first make a parallel between the article-like morphology of dei and the peculiar morphology that appears on the demonstrative quei ‘those’ and prenominal adjective bei ‘nice’. Then, we discuss the optionality of -i in Anconetano which sets indefinite dei on a par with quei and bei and apart from the articulated preposition dei found in genitive or partitive PPs. Finally, we argue that the definite determiner dei is

Indefinite dei is not a (partitive) P+art

In the preceding section, we have presented phono-syntactic evidence that indefinite dei is different from P+art and similar to determiners and modifiers of the noun. In this section, we discuss further morpho-syntactic and semantic evidence to show that dei-nominals are not PPs and are therefore different from the partitive PP in (1b).

dei is not a Q

Storto's proposal that indefinite dei is parallel to the quantifier alcuni does not predict the ungrammaticality of plural alcuni in (47a) and the parallelism between indefinite dei in (46a) and the singular indefinite un in (47b):

(47)a.*Mi piacerebbe trovare alcuni Dodo ma so che oramai sono estinti.
I would like to find some Dodos but [I] know that nowadays [they] are extinct
b.Mi piacerebbe trovare un Dodo ma so che oramai sono estinti.
I would like to find a Dodo but [I] know that nowadays

dei phrases are plural indefinite DPs

In Section 6.1, we see how our structural analysis can also derive a number of syntactic phenomena discussed by Zamparelli (2008). In Section 6.2, we follow a suggestion by Renzi (2001) to show that the alleged kind-denoting interpretation of dei-nominals can be derived by a hidden modality in the clause. In Section 7, we will further argue that the -i in dei is not the Italian kind-denoting determiner.

On the alleged kind-denoting interpretation of -i in dei

As we said in Section 3, the morpheme -i in dei is parallel to the -i found with the distal demonstrative quei and the prenominal adjective bei. Our analysis shares with Zamparelli's (2008) the assumption that -i does not have definite interpretation; but it is not bound to assume that -i has kind-denoting interpretation. In Section 3.2, we have already shown that this is a welcome result in Anconetano. We now show that the kind-denoting interpretation of -i is not needed to explain the Italian

Indefinite dei-nominals vs. bare plurals

In this section, we compare indefinite dei-nominals with so-called bare plurals. According to our proposal, repeated here as (81), both have DP status:

(81)a.Ho visto [DP dei [NPragazzi]].
[I] have seen de.artboys
b.Ho visto [DP Ø [NPragazzi]].
[I] have seenboys
From the fact that they are DPs and not QPs, it follows that bare plurals, parallel to dei-nominals, cannot select a partitive PP (82a), can combine with a universal quantifier (82b), and can occur as predicates (82c):

(82)a.*Ho letto dei / Ø

Conclusions

In this paper, we have argued that an attempt to unify (1a) and (1b) is neither supported by empirical evidence nor is it theoretically welcome.

Dei-nominals are not the same as the partitive PP headed by dei in (1b), nor are they to be assimilated to the larger phrase headed by the quantifier in (1b). Dei is neither an articulated preposition nor a quantifier, but an indefinite determiner, the plural counterpart of the indefinite article un. This has been supported by both morpho-syntactic and

References (36)

  • C. Barker

    Partitives, double genitives, and anti-uniqueness

    Nat. Lang. Linguist. Theory

    (1998)
  • A. Belletti

    The case of unaccusatives

    Linguist. Inq.

    (1988)
  • P. Benincà

    Nomi senza articolo

    Rivista di Grammatica Generativa

    (1980)
  • Richard Campbell

    Specificity operators in specDP

    Stud. Linguist.

    (1996)
  • A. Cardinaletti et al.

    Partitive “ne” and the QP-hypothesis: a case study

  • A. Cardinaletti et al.

    The syntax of quantified phrases and quantitative clitics

  • A. Cardinaletti et al.

    L’opzionalità alle interfacce sintassi – morfologia – fonologia

  • Anna Cardinaletti et al.

    Cartography and feature realization in the nominal expression

  • A. Cardinaletti et al.

    Quantified expressions and quantitative clitics

  • G. Chierchia

    Partitives, reference to kinds and semantic variation

  • G. Cinque

    Types of A’-Dependencies

    (1990)
  • D. Delfitto

    A propos du statut lexical de l’article partitif en français: Quelques hypotheses sur l’interaction entre morphologie et forme logique

  • M. Diesing

    Indefinites

    (1992)
  • Murvet Enç

    The semantics of specificity

    Linguist. Inq.

    (1991)
  • G. Giusti

    The categorial status of quantified nominals

    Linguistische Berichte

    (1991)
  • G. Giusti

    The functional structure of noun phrases: A bare phrase structure approach

  • B. Heine et al.

    World Lexicon of Grammaticalization

    (2002)
  • C. Heycock et al.

    Friends and colleagues: plurality, coordination, and the structure of DP

    Nat. Lang. Semant.

    (2005)
  • Cited by (0)

    The paper was originally solicited by Michelangelo Falco and Roberto Zamparelli for a special issue on partitives. We thank both of them and three anonymous Lingua reviewers for challenging comments and constructive criticism on earlier versions. The research has been funded by the PRIN 2012 project: Theory, experimentation, applications: Long distance dependencies in forms of linguistic diversity (2014–2017). Parts of this paper have been presented at the Workshop Partitivity in Romance and beyond (Zurich, December 11–13, 2014), at the PRIN project meeting (Milan, January 12–13, 2015), and at the CIDSM X (Leiden, June 22–24, 2015). We thank the audiences of these events and in particular Roberta D’Alessandro, Denis Delfitto, Luigi Rizzi, and Elizabeth Stark.

    View full text