Marie Labelle

Anticausative derivations (and other valency alternations) in French

Abstract It is proposed to derive the two distinct French anticausative constructions from the interplay of two functional heads, Voice and v, where non-active Voice dominates the morpheme se, and v is the verbal head introducing a dynamic subevent and assigning the Agent role. The middle anticausative derivation (Le vase se casse ‘The vase breaks’) results from the insertion of se under non-active Voice, coupled with the absence of a vP projection. By contrast, the active anticausative derivation (Le vase casse ‘The vase breaks’) results from the use of active Voice with a v projection lacking a specifier. It is shown how these hypotheses account for the derivation of change of state verbs, verbs of movement, as well as the middle anticausative construction with a typically agentive verb, construire ‘to build’.

The status of post-verbal subjects in French child language

Two different approaches to postverbal subjects in child language are discussed and evaluated: one according to which the postverbal NP is a true left(Deprez and Pierce 1993) or right-branching (Friedemann 1993/1994) VP-internal subject, and the other one according to which the NP is right-dislocated (Ferdinand 1993). We provide arguments against the VP-internal subject hypothesis, and show that the right-dislocation analysis is correct. New evidence, having to do wiih the position of the adverbial tout seul, is provided in support of the right-dislocation analysis. We also suggest that partial results concerning the Intonation patterns ofVS sequences Supports this conclusion.

THE ACQUISITION OF GRAMMATICAL CATEGORIES: THE STATE OF THE ART

Abstract This contribution reviews recent research on the acquisition of grammatical categories, focusing on three aspects: parts of speech, inflection, and subcategories of words. It appears that children are sensitive to the distribution of the various classes of words surprisingly early, and that inflectional categories are mastered by age 3 despite the complexity of their acquisition. Subcategories of words take longer to acquire. To account for the facts, it is generally assumed that children are equipped with a distributional learning mechanism, but given the complexity of the factors that must be taken into account, some constraints must be placed on this mechanism. Various types of possible constraints are discussed, from innate knowledge to processing constraints

Grammatical and Situational Aspect in French: A Developmental Study.

We study the ability of children to provide an appropriate continuation for a stimulus sentence, taking into account the joint demands of situational aspect and grammatical aspect. We hypothesize that the aspectual transitions required by some aspectual combinations play a role in the difficulty of providing an approrpriate continuation for them. We tested 130 French-speaking children of 5;06 to 9;0. In general, the data are consistent with the idea that the ability of children to construe an appropriate continuation for a stimulus clause is a function of both the situational aspect of the clause and the grammatical aspect provided by the verbal morpheme. There is a significant tense x situational aspect interaction in the number of continuations that children are able to provide in answer to the stimuli. Contrary to our expectations, there is no significant tense x situational aspect in the number of appropriate continuations, this being perhaps due to the small number of continuations for each stimulus type, but there are trends in the expected direction, which further studies may be able to confirm.

Examining the contribution of metasyntactic ability to reading comprehension among native and non-native speakers of French

The particular contribution of metasyntactic ability (i.e., the ability to consciously reflect about the syntactic aspects of language and intentionally to control grammatical rules) to second language reading skills is still not clear. While some studies concluded that metasyntactic ability contributes to reading among non-native speakers (NNS), others did not observe any particular contribution of that specific metalinguistic ability among NNS, despite showing a predictive value for their native speaker control group. Methodological aspects might explain these conflicting results, namely the target population, the measurement of metasyntactic ability, and the reading skill examined. The present study was set out to verify whether the particular contribution of metasyntactic ability to French reading comprehension would be the same among native and non-native upper-elementary children. A cross-sectional study was carried out in which 73 children (37 native and 36 non native speakers of French) were given syntactic, metasyntactic, receptive vocabulary, reading comprehension and phonological memory tasks. As in previous studies, results of the multivariate analysis of variance (MANOVA) first revealed that the NNSs of French participants obtained lower MSA results than the native speaker children. However, results from the multiple regressions showed that MSA accounted for a significant part of variation in L2 reading among the native as well as among the NNS children and that language group was not a significant factor. This indicates that the weight of each variable, including metasyntactic ability, did not vary according to language status.

Measuring Metasyntactic Abilities: On a Classification of Metasyntactic Tasks

Researchers working on metasyntactic abilities (i.e., the metalinguistic ability associated with syntax) face the problem of defining and measuring them. Metasyntactic abilities is a multifaceted concept, which encompasses various types of behaviours, from being able to intentionally manipulate syntactic structures to being able to state syntactic rules, and the way in which it is defined and measured varies greatly from one study to another. The present paper proposes a theoretically informed classification of syntax related tasks. The first part presents previous research defining and distinguishing various types of syntactic and metasyntactic abilities and their interrelations. In the second part, commonly used tasks are described and analyzed in terms of the framework presented, with the aim of better pinpointing the type of ability measured by each task. Ultimately, with this analysis of commonly used tasks, we hope to offer criteria for discriminating between the various measures of metasyntactic abilities.

Negative concord in Quebec French

Abstract It is argued that there are two types of asymmetric negative concord languages: in languages like Spanish and Italian, negative concord results from a purely formal agreement relation between the negation and a negative concord item. In Quebec French, in addition to this purely formal licensing, there is a negative dependency relation between both items, which form two segments of a discontinuous negative quantifier. This accounts for the following differences. While Spanish, Italian and Quebec French reject negative concord between a subject negative expression and the negation, in Quebec French, negative concord with the negation becomes possible when the clause contains a postverbal negative expression in addition to a preverbal one. Moreover, in Quebec French, but not in Spanish or Italian, negative concord is blocked across a quantifier meaning almost.

Roots in models of grammar

In the target article, Harley argues, on the basis of suppletion and the existence of meaningless roots, that roots enter the syntax individuated, but unspecified for semantic content and phonological features. In the syntax, roots are pure units of structural computation, referred to by addresses serving as linkage between phonological and encyclopædic instructions. These instructions are accessed after Spell-Out, and determine the pronounciation and meaning of the roots in the context in which they are inserted. She also argues that roots may select complements and that the domain of special meanings is delimited by the first head introducing an external argument. That syntax does not care about phonological or encyclopædic information is relatively uncontroversial, and to argue that the syntactic component manipulates addresses pointing to instructions accessed after Spell-Out is quite reasonable. However open questions remain, among which: 1) whether and how the conceptual-intentional system is involved in the selection of the numeration; 2) the nature of the information attached to roots; and 3) the domain of special meanings. In the present commentary, I would like to reflect on these questions.