Anne ABEILLÉ et Danielle GODARD : « La position de l’adjectif épithète en français : le poids des mots »
We suggest an explanation of the position of the attributive adjective with respect to the Noun in French, based on the notion of syntactic Weight. In our approach, adjectives belonging to certain morphological or semantic subclasses are marked as being “lite weight”, and must be preposed, except if they are modified or conjoined. The notion of liteness also accounts for the impossibility to prepose full adjectival phrases (with complements or phrasal modifiers).
Sabine BOUCHERON : « La ponctuation dans le texte : parenthèses, sujets et linéarité dans l’incipit du Palace de Claude Simon »
We set out to deal with the question of linearity through punctuation, especially through the effects of brackets. In the incipit of Claude Simon’s Palace, it can be noticed that the linearity of language gets double. A first level, subjected to the scriptural norms of the nouveau roman, provides a hyperrealistic description of a pigeon ; a second level, brought about by brackets, reintroduces traces of subjectivity : novel characters, the modalizing subjectivity of a narrator and the thread of writing, which reappears at the heart of the text. This text is thus written along several lines, and makes it as if it was possible to avoid the language linearity which ordinarily constrains the writer.
David HOUGHTON : « Relinearizing phonology : an edge feature account of Shona »
In this paper, a new system of description of register tone is presented wherein the characteristic of the melody which is described is the location of edges between tonal plateaus. This is contrasted with the standard non-linear autosegmental account wherein the characteristic described is the tone expressed on syllable nuclei. The two systems of description are each applied to a set of problems in Shona tonology : Meeussen’s rule, Stevick’s rule, and some of the tonology of the verb stem. The edge feature account is more like linear analyses in that it does not involve multiple linking of tones to syllables. It is formally simpler in the number of stipulations it requires. It is less powerful in that its rules are more restricted in the features they can affect and how and where they can affect them : they all reset the pitch to the default tone at morphological junctures. In this, it better reveals patterns in the tonology of Shona. Moreover, its rules are functionally motivated and applicable on-line, unlike the rules of the autosegmental account.
Patrick SAUZET : « Linéarité et consonnes latentes »
Suppose linearity doesn’t pre-exist to syllabification but is built by it, within morphemes, on the basis of the relative accessibility of segments. On the one hand, applying syllabification to syntacticlike structures allows deriving, via phonology, part of the order prevailing between morphemes. Morphology and words thus appear as effects of phonology. On the other hand, it makes it possible for morphemes to embody a segmental content more connected than traditional concatenated morphemes are, or, on the contrary, presenting discontinuities. Elaborating upon this latter possibility, latent consonants (as are typically present in the phonology of French) are analysed as dissociated segments, viz. dissociated from the bulk of the phonetic specifications (instead of “floating” which, it is claimed, appears to be more problematic than is currently assumed). The analysis advocated here not only allows dealing with invariable latency (“petit, petit-t-ami, petitesse”), but also with variable latency (“grand, grande, grand-t-ami”, “vert, verte, verdure”), suppletion (“beau, bel”) and other word final variability phenomena (“sec, sèche”).
Tobias SCHEER : « Aspects de l’alternance schwa-zéro à la lumière de “CVCV” »
This article discusses the distribution of consonant clusters adjacent to Metropolitan French schwa. Their influence on the possibility of eliding schwa are examined. The French central vowel can be dropped by at least a subgroup of speakers when followed by a sequence of rising sonority (“le søcret”) or preceded by a Coda-Onset cluster (“marguørite”), while preceding branching Onsets enforce its presence (“*vendrødi”) (following Coda-Onset clusters do not occur). This situation casts doubt on previous analyses carried out within the framework of Government Phonology. Namely, the statements “Proper Government may not apply over a governing domain” and “heads of non-nuclear governing domains need to be licensed” are falsified. However, French vowel-zero alternations share crucial properties with similar alternations in many other languages, so that a unified account is called for. I argue that such an analysis may be achieved when assuming a “CVCV” syllable structure, as proposed in Lowenstamm (1996). In this view, constituency boils down to a strict sequence of non-branching Onsets and non-branching Nuclei. It is shown that a unified account may be proposed for both the “marguørite” and the “søcret” case if a theory of consonantal interaction is implemented into the theory. As a by-product, the impossibility of “*vendrødi” falls out naturally.
Philippe SCHLENKER : « La flexion de l’adjectif en allemand : la morphologie de haut en bas »
German adjectives may receive different endings depending on their position in a Noun Phrase. An adjectival ending expresses lots of gender/number/case contrasts when it is not preceded by any other inflected element ; otherwise the ending is impoverished, and most contrasts are neutralized. We suggest that : (1) A single feature bundle is transmitted sequentially from the top to the bottom of an NP. (2) At each step a morpheme is inserted, and the features that it expresses can be deleted from the bundle, so that the following adjectives inherit an impoverished set of features &endash; which leads to the neutralization of most contrasts.
Sophie WAUQUIER-GRAVELINES : « Segmentation lexicale de la parole continue : la linéarité en question »
Speech signal and phonological representations of speech adopted by psycholinguistical models (SPE model) have supported for a long time a strictly left-right algorithmic linear conception of speech treatment. Last experimental results and last psycholinguistical modelisations, as also proposals of multilinear phonology, invite us to consider that speech treatment could be more heuristic and could include both proactive and retroactive processes.
Jung-Sup BAK : « Portée des quantifieurs et ordres des mots en coréen »
This article aims at describing the quantifier scope in various multi-quantifier constructions in Korean. For that purpose, it adopts the theory of Diesing which offers an explaination of many quantification phenomena in different languages. Moreover, it proposes a new analysis of a movement called ” M(iddle)-scrambling ” in languages like Korean. For a long time, it has been thought in the literature that the subjet behaves as an exception with respect to M-scrambling. But this widely admitted idea is rejected in this paper. It is demonstrated here that M-scrambling is a movement toward a position outside vP which concerns all the non predicative phrases. This analysis of M-scrambling, combined with the theory of Diesing, offers a correct explaination of quantification facts in Korean.