University of Siena CISCL Working Papers #3

Proceedings of XXXV Incontro di Grammatica Generativa

Editor: Vincenzo Moscati and Emilio Servidio, Year:2009

Agreement in the production of Italian subject and object wh-questions

MARIA TERESA GUASTI, CHIARA BRANCHINI, FABRIZIO AROSIO University of Milano-Bicocca

[no abstract available]

The syntax of conditional clauses

LILIANE HAEGEMAN, Department of English, University of Ghent

This paper shows that the movement derivation of conditional clauses (Bhatt and Pancheva 2002, 2006, Arsenijevi? 2006, Lecarme 2008) allows us to account for the fact that Main Clause Phenomena are excluded in conditional clauses because this follows from intervention effects. Moreover, the cartographic implementation of the analysis proposed predicts the incompatibility of conditional clauses with the speaker oriented modal expressions as well as the fact that conditional clauses lack the low construal reading which is found in (some) temporal adverbial clauses (cf. Bhatt and Pancheva 2002, 2006). The paper thus reinterprets one of the potential objections against the movement account of conditional clauses into an argument in favour.

Mesoclisis in the Imperative: Phonology, Morphology or Syntax?

M. RITA MANZINI, LEONARDO M. SAVOIA Universit di Firenze

Mesoclisis in the imperative and parasitic plurals in Spanish are currently accounted for either at Morphological Structure (Halle and Marantz 1994) or at the PF interface (Harris and Halle 2005). In previous work (Manzini and Savoia 1999, 2004a, 2004b, 2005, 2007, 2008a, 2008b) we argued that these and similar phenomena in both Romance and Albanian are best accounted for at the syntactic level. Since sub-word constituency is involved, this amounts to saying that syntax subsumes morphology. Here we defend the conclusions of our previous work, including in particular a strictly lexicalist stance on the projection of morphosyntactic structures from the lexicon.

Linearization and the Architecture of Grammar: A view from the Final-over- Final Constraint

THERESA BIBERAUER, ANDERS HOLMBERG & IAN ROBERTS

University of Cambridge, Stellenbosch University, Newcastle University

This paper addresses the issue of the locus of linearization information in the context of a minimalist grammar. Contrary to what is arguably the dominant view in minimalist theorizing today, it is argued that linearization information must in fact be specified Narrow Syntax-internally. The imperative underlying this conclusion is an empirical skewing in the domain of word-order variation, in terms of which head-initial structures associated with a given projection line may only be (harmonically) dominated by head-initial structures, while head-final structures may be dominated either by head-initial or head-final structures in the same context the so-called Final-over-Final Constraint (FOFC). It is argued that attested FOFC effects suggest that linearization information is in fact encoded in such a way namely, by harnessing an already-required movement diacritic in accordance with Relativized Minimality, arguably a third-factor-imposed principle that its NS-internal presence does not violate the Strong Minimalist Thesis of Chomsky (2001 et seq.). We also consider the question syntactic categories and their formal status, against this background.

Bantu verbal extensions: a cartographic approach

GLORIA COCCHI, Universit di Urbino

In this work I am going to discuss the nature, productivity and combinatory possibilities of verbal extensions in Bantu languages, considering some Tshiluba data. I will argue that this phenomenon is best accounted for within a cartographic approach to sentence structure. Though verbal extensions are generally treated as a unitary phenomenon in the literature on Bantu, several important differences can be observed; hence I will argue that extensions should be divided into two main groups, syntactic and lexical extensions, and the latter into two sub-groups. I will assume that syntactic extensions are merged under specialized functional heads in the predicative domain, while truly lexical extensions are merged with the verb stem in the lexicon. Finally, the analysis of the so-called lexical-argumental extensions will lead to the assumption of a second vP projection immediately above VP, in whose head these extensions are merged. Therefore, the two phase heads, C∞ and v∞, will exhibit a similar behaviour, in that they both consist of two separate heads, between which other functional projections are generated.

TopicPs and Relativised Minimality in Mcheno left periphery

FEDERICA COGNOLA, University of Padua

In this work I will take into consideration the high left periphery of the Tyrolean dialect Mcheno, showing that in this language, unexpectedly under the latest version of Relativised Minimality in terms of features (Rizzi 2004), Relativised Minimality effects are found between a topic and a focus, but not between a wh-element and a topic. In my work I will show that this asymmetry inside the Quantificational class is due to the structure of Mcheno left periphery and should not be taken as evidence in favour of the need of a split inside the Quantificational class. Looking at the structure of the topic fields activated by operators in Mcheno, I will propose that in this language Relativised Minimality effects between two XPs belonging to two different featural classes do not occur if two conditions are met: i) the two XPs belong to two different featural classes; ii) TopicPs dedicated to constituent categories (strictly ordered one with respect to the other) are available. In the last part of the work I will bring evidence in favour of the idea that also in Romance (Italian) TopicPs are strictly ordered, even if this is not immediately visible due to topic free order.

Focus Fronting in Sardinian and Sicilian

SILVIO CRUSCHINA, University of Oxford; EVA-MARIA REMBERGER, Universitat Konstanz

This paper analyses the common properties and differences in the interpretation as well as in the syntax of Sardinian and Sicilian Focus Fronting (FF) phenomena. In both varieties, FF is a syntactic device used to mark not only Contrastive Focus (CF), but also Informational Focus (IF). The fronted IF is often associated with a "special interpretation", such as a mirative value or verum. Fronting is not limited to DPs or PPs, but may also involve predicates. However, while Sardinian allows FF for all kinds of predicates, predicate fronting in Sicilian is only possible in copulative constructions, and marginally possible with infinitives. We claim that FF is always XP-fronting and that the difference between Sardinian and Sicilian emerges because in Sicilian the verbal predicate in the form of the active past participle is outside the VP and cannot thus be XP-fronted, whereas participles are always inside the VP in Sardinian.

On bare nominals and argument structure

M.TERESA ESPINAL & JAUME MATEU, Centre de Lingistica Terica, Universitat Autunoma de Barcelona

This paper aims at focusing on a set of phenomena related to the syntax and semantics of bare count nominals: bare count nominals (BNs) in argument position, BNs in predicate position, and the relationship between argument structure and the interpretation of bare nominals. The novelty of this paper is to relate the occurrence of bare nominals with the argument structure position in which they may occur at a syntactic level of representation.

Stylistic Fronting: a comparative analysis

IRENE FRANCO, Universit di Siena - CISCL

Stylistic Fronting (SF) is a syntactic phenomenon present in modern Insular Scandiavian languages, probably as a residue of Old Icelandic word order. Mainland Scandinavian languages have lost SF, but diachronic studies show that Old Scandinavian languages display SF (cf. Falk 1993, Trips 2003). SF is also found in Old Romance varieties, among which is Old Italian (cf. Beninc 2006). Despite the considerable number of proposals, SF has not received a satisfactory account. It is difficult to find a theory of SF compatible with the idea that the left periphery of the clause has general structural properties. In section 1. I give a brief overview of the properties and distribution of SF in contemporary Icelandic. In section 2. I present some comparative data showing that the same phenomenon is attested in Old Italian. In section 3. I focus on the syntactic conditions licensing SF, and identify the target position of fronted items by observing the Old Italian SF distribution with respect to overt subjects. In section 4. the distribution of SF is observed with respect to CP expletives (cf. Poletto 2005) and enclisis/proclisis (cf. Beninc 1993). In section 5. I propose a unifying analysis of SF for Old Romance and Icelandic as a potential strategy to extract/drop the subject, based on an integrated synchronic/diachronic perspective and adopting a derivation of SF in terms of remnant movement to the CP (cf. Franco 2009).

The Adnominal Form in Japanese as a Relativization Strategy

HANAKO FUJINO Instituto Universitario de Investigacion Ortega y Gasset, Universidad Pontificia Comillas de Madrid

This article focuses on two aspects of relative clauses in Japanese: its syntactic structure and the special form ('adnominal form') that the embedded predicate must take. It has been assumed that Japanese relative clauses are basegenerated and are D-IP structures (Murasugi 2000a,b). I will argue on the contrary that they are D-CP structures, that are derived by raising of the head. This argument is supported on three accounts: (i) a reconsideration of reconstruction effects with respect to the reflexive interpretation of zibun; (ii) the manifestation of the weak crossover phenomenon; (iii) the existence of sentential modifiers with CP elements. Then, I will show that Japanese has a requirement on sentential modifiers in general, namely that the embedded predicate must be in a special form called the adnominal form. On the basis of the Clausal Typing Hypothesis (Cheng 1991), I will propose that the role of the adnominal form (and the particle no) is to enable clausal typing of the embedded clause.

Romanian null objects and gender

ION GIURGEA University of Constance & Iorgu Iordan Al. Rosetti Institute of Linguistics, Bucharest

This paper argues that Romanian has anaphoric object pro, which is used for variables bound by a quantifier lacking gender and for propositional objects. It will be shown that the so-called neuter pronouns of Romanian and other Romance languages, which are used for referents that do not fall under a nominal concept, are genderless. This follows from the fact that natural gender in these languages is restricted to humans.

The hybrid complementizer system of Cimbrian

GUENTHER GREWENDORF, University of Frankfurt; CECILIA POLETTO, University of Venice

In this work we examine the syntactic properties of two classes of complementizers in Luserna Cimbrian, an endangered language spoken in the Dolomites, and show that they occupy distinct positions. The first type of complementizer starts out in the Fin∞ position and moves up to Force∞ thereby blocking the whole CP which is not available for the verb to move. The second type is external to the clause itself, which can behave as a main clause as its CP is entirely empty. The tests we use to show that this distinction is necessary are: the position of clitics and of the sentential particle /da/, the position of the inflected verb with respect to the negative marker /net/ and to verbal prefixes, and the distribution of the CP expletive /z/, which is the Cimbrian counterpart of standard German /es/. This analysis has consequences on the one hand on the layering of the CP area and on the other on the V2 properties of Cimbrian.

Head-internal relatives in Japanese/Korean

ALEXANDER GROSU, Tel Aviv University

The internally headed relatives (IHRs) of Japanese and Korean belong to the general class of 'definite' relative constructions, whose CP was characterized in Grosu & Landman (1998) as denoting singleton predicates. Kim (2007), building on Hoshi (1995) and Shimoyama (1999, 2001), defines this CP as denoting a proposition that contains the antecedent of an E-type anaphor. It is argued in this paper that this approach, which undesirably enriches the class of definite relatives, also necessitates the imposition of highly unnatural restrictions on anaphora, which blur the distinction between pragmatics and grammar. The paper proposes an alternative analysis that avoids the conceptual and empirical objections faced by E-type approaches in general and by Kim's in particular, and assigns singleton status to Japanese/Korean IHRs in a straightforward and natural way.

Why and how to distinguish between pro and trace

DALINA KALLULLI, University of Vienna

Starting from a well-known observation, namely that in a language like Hebrew there is no free alternation between traces and (overt) resumptive pronouns, this paper aims to demonstrate that even in languages with seemingly little or no resumption such as English, the distinction between a putatively null resumptive pronoun and trace is equally material. More specifically, I contend that positing a resumptive (i.e. bound variable) pro also in English-like languages is not only theoretically appealing for various reasons (a.o. ideas in Hornstein 1999, 2001, Boeckx & Hornstein 2003, 2004, Kratzer 2009), but also empirically adequate (as conjectured e.g. in Cinque 1990). The central claim of this paper however is that resumption is restricted to (sometimes concealed) relatives. Applying this proposal to languages like English, the distinction drawn between (resumptive or bound variable) pro and trace accounts for phenomena as diverse as lack of superiority effects, lack of weak crossover in appositives, lack of Principle C effects in relative clauses, and so-called ATB movement phenomena.

The processing of N-words in Italian.

DANIELE PANIZZA, Universit di Trento, DiSCoF Rovereto

In this paper I investigate the processing of N(egative)-words such as mai (ever) in Italian, by means of the Event-Related Potentials methodology. N-words pose some questions about their syntactic and semantic behavior, such as: a) what principles underlie their distribution; b) which is the role of syntactic and semantic factors in determining their meaning; c) how the brain processes these linguistic items in grammatical and ungrammatical sentences. In the present work I will provide some theoretical and psychological argumentations to address these questions.

Labelling, Multiple Spell-Out and the Final-over-Final Constraint

MICHELLE SHEEHAN, Newcastle University

This paper explores the relationship between labelling and Kaynes Linear Correspondence Axiom, arguing that the latter requires some version of the former. More specifically, a copy theory of labelling, compatible with inclusiveness, is proposed which renders phrase structure linearizable via the LCA. This labelling mechanism, coupled with Uriagerekas (1999) Multiple Spell-Out, makes a number of interesting predictions about the properties of left-branches, which are shown to have empirical support. Most importantly, an apparently correct distinction is predicted to hold between underlying specifiers and derived specifiers, and more generally between head-final vs. head initial specifiers. These differences, it is proposed, serve to explain a number of seemingly unrelated facts: (i) the restrictions on extraction from derived specifiers in English, (ii) patterns of complement extraposition, (iii) the lack of CED effects in harmonically head-final languages, and (iv) the Final-over-Final Constraint (cf. Holmberg 2000).

Epistemic Modality in Romanian: the role of BE

ELENA SOARE, University of Paris 8

This paper contributes to the view that the epistemic/root distinction in modals across languages relies on structural differences. I will address the behavior of Romanian modals in combination with perfective Aspect, in comparison with other Romance languages (French, Spanish and Italian). It is a wellknown fact that Romance languages, inasmuch they distinguish between perfective and imperfective Aspect, have ambiguous modals in the perfective; they allow either the root or the epistemic reading. In Romanian the root, veridical reading is triggered with perfective on the Modal. I will claim that the explanation for that relies in the fact that Romanian Modals have bi-clausal structures. I will also address the conditions in which the epistemic readings can arise, and provide an explanation of their preference for embedded verbs (or Small Clauses) denoting stative situations, further developing the idea of a structural difference between the relevant readings. I will also propose a possible connection between epistemics and evidentials in Romanian.

Identificational Focusing: Focus Raising and Stress-Focus Correspondence

BALAZS SURANYI, Research Institute for Linguistics, Hungarian Academy of Sciences

This paper revisits the division of labor in the grammar in identificational focus constructions in Hungarian. It is argued that if applied to identificational focus (rather than focus in general), Chomskys (1971, 1976) proposal that focusing involves a syntactic readjustment operation analogous to Quantifier Raising can and should be upheld: identificational focus undergoes movement in order to be interpretable. Since identificational focus is a subcase of focus (defined as involving alternatives), the PF manifestation of identificational focus movement is affected by prosodic constraints on focus, including a StressFocus Correspondence requirement (Cinque 1993, Reinhart 1995). The interaction of this requirement with general principles of economy determines the realization of focus movements at PF. It is shown how the basic structure of the Hungarian clause, without a specialized functional head for focus, accommodates both the semantic and the prosodic needs of identificational focus elements in a variety of focus constructions.

The subject/object relative clause asymmetry in Italian hearing-impaired children: evidence from a comprehension task

FRANCESCA VOLPATO, Universit Ca Foscari di Venezia; FLAVIA ADANI, Laboratoire de Sciences Cognitives et Psycholinguistique, EHESS/CNRS/DEC-ENS, Paris

We investigated the comprehension of subject and object relative clauses in hearing-impaired (HI) children using a cochlear implant compared to that of hearing children, by using an agent selection task. We show that HI children performed significantly poorer than their typically-developing peers. Despite their low performance, HI children show nonetheless a typical gradient of difficulty, with subject relatives (OS) easier to comprehend than object relatives with preverbal subject (OO) and these latter are easier than object relatives with postverbal subject (OOp). These asymmetries are explained in terms of some recent minimalist proposals on locality theory and on the fragility of Agreement occurring with postverbal subjects. A correlation between performance on OOp and digit span tasks was found only in the HI group.