Volume 37, Issue 1

The standard explanation for Ν Ρ movement in the passive construction has been that the N P must move into the nominative position because no accusative case is available. This paper examines the implications for this view of some double-object constructions in Mandarin Chinese and English that are ungrammatical as active clauses but improve significantly as passives. These facts are unexpected under the standard view of passives, but I suggest that they can be explained if we assume that the second object is not licensed for case in the active versions but is able to check accusative case in the passive version, thus arguing that accusative case is available in passive clauses.

In common with English and German and many other languages, French has an accentual configuration that arises in clash contexts, here referred to as clash resolution. Experimental research has shown that clash resolution is bounded by the phonological phrase. Moreover, the optional application of clash resolution has provided evidence for the restructured phonological phrase. However, the conditions on restructuring that have been proposed in the literature have been found to make the wrong predictions in a number of cases. In contradiction to earlier claims, perceptual and acoustic evidence shows that lexical heads can also be restructured with their branching complements. In addition, an analysis of the phonetic evidence shows that clash resolution is characterized by the absence of a final pitch accent, while the simultaneous realization of an initial pitch accent on the stress-shift item is optional. In our conclusions, we will discuss the implications of these findings for a phonological account of clash resolution.

The paper examines the diachronic implications of asymmetries in the synchronic distribution of syntactic variants, by inquiring whether syntactic alternatives can differentiate historically, following the model of morphological split. A discussion of the general principles underlying diachronic differentiation and syntactic variation is followed by an in-depth analysis, in a large corpus of Goldoni's comedies, of the alternation in Italian of con phrases with first and second person pronouns as object where the preposition is enclitic vs. preposed (meco/con me, teco/con te). The two variants are shown to be nonequivalent, and pragmatically in complementary distribution. The failure of the (older) variant with enclitic -co to survive historically and the order in which it disappeared in the three grammatical persons are both accounted for; it is further shown that the synchronic distribution of the alternatives, especially the relative proportion of cases of the older variant found in more or less "categorical" environments, foreshadows the historically observed outcome.

This article examines the differing degrees to which young male and female speakers are apparently participating in a process of linguistic change currently taking place in southern French. Previous research suggests that younger females in southern France are adopting the treatment of schwa characteristic of the northern standard variety; but no study of linguistic change in southern French has yet been carried out through a direct, variationist comparison of young male and female speakers. We analyze here the linguistic contexts in which schwa is most susceptible to penetration by the northern French norm. Much research on the differential treatment by male and female speakers of variable linguistic forms has shown that young males tend to resist the introduction of supraregional linguistic forms, while females adopt them. As well as discussing new evidence on a Frenchspecific research issue, we also examine the crosslinguistic significance of the French findings. Further, we attempt to shed new light on the behavior of male and female speakers in relation to national and regional linguistic forms, and to reconsider some issues associated with the definition of standard languages.

The article deals with the notion of category and the linguists operations for delimiting linguistic categories. A threefold organization is suggested that subdivides categories into features and features into values. Every word of a natural language can be categorically described via a matrix of values that represent the implementation of features that, in their turn, are categorial properties. Since there exist many nonclearcut cases, that is, items that may paradigmatically belong to more than one category, it is necessary to use both a functional and a formal approach in order to get a categorial definition of the items. The traditional parts of speech still seem to be the best categorization, in spite of the fact that typology has become acquainted with languages that show very different morphosyntactic structures.