Volume 20, Issue 2

In this paper we discuss the interaction between postverbal subjects in French and interpretations of (in)definite DPs, including relevant pragmatic properties generally stated in terms of topic and focus. We show that postverbal subjects are not homogeneous, neither from the perspective of theories that explain differences among DPs in terms of (in)definiteness (distinctions between ‘high’ and ‘low’ readings) nor from the perspective of theories that distinguish topic-oriented constructions from focalisation. We argue that mapping theories should be kept as ‘clean’ as possible, so that they may shed light on the position of the subject DP, i.e. the syntax of inverted structures. We show that interpretative and pragmatic variation in combination with a simple mapping theory argues in favor of a diversity of syntactic (remnant) movements somewhat richer than has been standardly assumed for French (and somewhat different too). We propose a (partial) hierarchy of functional projections Wh < Topic < Generic < Focus < Existential with three types of remnant movement in embedded clauses and a fourth one in root wh-questions.

This article draws on empirical findings from child language in order to evaluate syntactic analyses of object clitic constructions in adult French, focusing on two long-standing issues in the literature: (i) the argumental status of direct object clitics, and (ii) the site of their first merge. Two sets of acquisition data bearing on these questions are discussed: the persistent occurrence of object (clitic) omission, and the lack of clitic misplacement observed in child French. I argue that the former can only be adequately explained if object clitics are not argumental, and that the latter presents evidence against a first merge of the clitic in the verbal complement. This leads to the conclusion that the developmental data considered here are fully consistent only with syntactic analyses of object clitics as inflectional markers generated outside the complement of the verb, as proposed, for example, in Roberge (1990) and Sportiche (1996).

This paper focuses on opaque Latin vowel deletion and its relevance to the actual theoretical debate of dealing with opacity in OT. Besides derivational OT, three parallel OT approaches to deal with opacity in OT are discussed: Sympathy Theory (McCarthy 1999 and 2003), Comparative Markedness (McCarthy 2002) and OT with Candidate Chains (McCarthy 2007). It is argued that Sympathy Theory, for principled reasons, is unable to deal with opaque Latin vowel deletion. The main reason is that the opacity in this case crucially depends on the prosody and not on faithfulness. Comparative Markedness can account for opaque syncope by splitting up the constraints Parse -σ and Ft - Bin in new ( N Parse -σ and N Ft - Bin ) and old ( O Parse -σ and O Ft - Bin ) constraints. However, the ranking necessary to deal with the cases of opaque deletion leads to incorrect outputs in cases of transparent deletion. Finally, OT with Candidate Chains will be shown to allow for a more restrictive and more principled account of opaque syncope than a serial, level OT approach.

Focusing on Spanish, this article discusses the internal syntax of adjectival projections headed by a gradable adjective. The analysis of the properties of the Spanish term poco ‘little’ in combination with gradable adjectives and degree terms ( poco inteligente , ‘not very smart’; bastante poco inteligente ‘not so much smart’; increíblemente poco inteligente , ‘incredibly not smart’; lo increíblemente poco inteligente , ‘how incredibly not smart’) will serve as the starting point for the proposal that gradable adjectives project a quantifier phrase above AP and below DegP. The quantifier phrase adds a semantic component of scalar interpretation. From the analysis of the extended projection of gradable adjectives proposed throughout this article the following classification of degree terms in the adjectival domain is derived: (a) Quantitative terms, which head the quantifier phrase and convey an upward or downward orientation on a scale: poco ‘little’ and a phonetically null quantitative term equivalent to mucho ‘much’ in Old Spanish. (b) Degree operators, which identify a specific degree on the scale from an upper functional degree phrase. Examples are bastante ‘too’, demasiado ‘too much’, tan ‘so’, más ‘more’, etc. And (c) Elatives, which modify the degree argument of the adjective by adding an appreciative extreme degree reading. Examples are increíblemente , ‘incredibly’, buenísimo ‘tall- superlative ’, enorme ‘enormous’.

Les nombreux travaux centrés sur la construction causative française avec faire ont débattu, entre autres, du placement des pronoms clitiques datifs correspondant aux compléments prépositionnels à SN . Cependant, le placement des pronoms clitiques datifs correspondant aux compléments prépositionnels des adjectifs dans la construction causative avec rendre reste en marge des débats. L'étude conjointe des verbes et des adjectifs bivalents menée dans cet article aboutit à la conclusion que seuls les verbes et les adjectifs bivalents « ergatifs » dont le complément au datif est « affecté » par le procès décrit par le prédicat complexe formé par le verbe causatif et le prédicat enchâssé autorisent que le pronom clitique datif soit attaché au verbe causatif enchâssant. Les faits rassemblés et discutés tendent donc vers la conclusion que les adjectifs français ne sont pas tous « ergatifs » : certains adjectifs sont inergatifs i.e possèdent un argument externe, ce qui implique la décomposition du syntagme adjectival.