Eds: Vandenitte S\u00e9bastien, Lepeut Alysson and Meurant Laurence<\/p>\n\n\n
Jorina Brysbaert & Karen Lahousse<\/strong><\/p>\n<\/div>\n<\/div>\n<\/div>\n This paper is about the influence of text formality on the syntactic position of contrastive adverbs in French. We analyze the distribution of the two contrastive adverbs en revanche and par contre, both conveying a meaning similar to English \u2018on the other hand\u2019, in three written corpora representing different levels of formality. We show that (i) these adverbs occur more often in clause-initial position as the level of formality decreases, and that (ii) the formal character of en revanche and the informal character of par contre explain distributional differences between the two adverbs. We conclude that text formality has a significant influence on the position of these two contrastive adverbs, next to factors such as text type and lexical idiosyncrasy, which had been put forward in the linguistic literature. Our findings also corroborate the idea that French is largely characterized by formality-induced variation, even with respect to word order patterns, which have not received much attention in previous research on formality.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div> Silke Creten, Priscilla Heynderickx, Sylvain Dieltjens<\/strong><\/p>\n In 2015 schatte Alzheimer\u2019s Disease International dat er wereldwijd ongeveer 50 miljoen personen met dementie (PMD) waren, onder wie 135.000 in Vlaanderen. Ondanks hun toenemende aantal krijgen PMD nog altijd te maken met stigmatiserende houdingen, zowel tegenover hun leeftijd als tegenover hun psychische aandoening. Deze studie analyseert de representatie van dementie in subjectieve en informele tekstkanalen, om zo een beter inzicht in het stigma rond dementie in Vlaanderen te krijgen. Daarom analyseren we in detail acht blogberichten op een blogsite over dementie. De focus van de analyse ligt op conceptuele metaforen en op de gevoelswaarde van de teksten. De kwalitatieve analyse wordt op vooraf vastgelegde conceptuele metaforen uitgevoerd met de Conceptual Metaphor Theory (CMT). We analyseren in detail de conceptuele metaforen voor dementie en PMD die door een mantelzorger gebruikt worden, en vergelijken die met bevindingen uit eerder onderzoek. Daarnaast gebruiken we de software van Linguistic Inquiry and Word Count (LIWC) om de teneur van de blogberichten te bepalen. De mantelzorger gebruikt voornamelijk metaforen voor dementie die in eerder onderzoek als \u2018negatief\u2019 voor de representatie van de aandoening werden bestempeld. Toch toont de LIWC-analyse aan dat over het algemeen de blogberichten ook positieve emoties overbrengen. De vergelijking met metaforen voor dementie in eerder onderzoek toont aan dat er in het discours over dementie terugkerende patronen zijn die in toekomstig onderzoek omvat kunnen worden in een framework met vaste categoriee\u0308n.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div> Alysson Lepeut<\/strong><\/p>\n This paper examines manual holds in French Belgian Sign Language (LSFB) conversations. Holds have usually been examined in sign languages in relation to linguistic structures but barely for their relevance in managing signed discourse. Nonetheless, scholars have begun to argue for their non-arbitrary positions as part of interactional practices that signers employ during conversations. Rather than approaching holds in relation to prosody and syntax, this paper investigates their roles as potential meaning-making practices used by LSFB signers in their organization of social interaction. The following question is explored: what are their roles in LSFB conversations? The present paper argues in favour of a paradigm shift in the conception of holds by showing that an absence of manual motion does not equal an absence of meaning. On the contrary, the results reveal that these are not mere insignificant moments of signing excursions but efficient means to achieve pragmatic goals in conversation. Results show that LSFB signers use holds to carry out four main functions in discourse, i.e., to: monitor addressees, plan discourse, hold the floor and suspend one\u2019s turn when the addressee intervenes in the main line of action.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div> Erica Lippert<\/strong><\/p>\n En 2019, d\u2019importants incendies ont eu lieu en Amazonie, e\u0301ve\u0301nement constituant un the\u0300me majeur dans les me\u0301dias. Notre analyse discursive porte sur la communication au sujet des incendies sur le profil Instagram du pre\u0301sident franc\u0327ais, E. Macron. Quelles repre\u0301sentations de la nature sont de\u0301ploye\u0301es ? Quelles rhe\u0301toriques visuelles et discursives sont employe\u0301es ? Nos recherches montrent que le discours s\u2019appuie sur un ethos e\u0301cologique que le pre\u0301sident construit, entre storytelling et position de sauveur.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div> Clara Lombart<\/strong><\/p>\n<\/div>\n<\/div>\n<\/div>\n La composante suprasegmentale de la phonologie \u2014 ou prosodie \u2014 s\u2019organise selon la me\u0302me hie\u0301rarchie de constituants et accomplit des fonctions similaires en langues vocales et en langues des signes. Les diffe\u0301rences formelles e\u0301tablies entre ces deux cate\u0301gories de langues induisent cependant des spe\u0301cificite\u0301s de marquage, qui ont d\u2019abord e\u0301te\u0301 de\u0301crites en conside\u0301rant uniquement la parole et les signes manuels. Avec la reconnaissance de la multimodalite\u0301 des langues vocales s\u2019ope\u0300re une inte\u0301gration des gestes dans le champ des recherches linguistiques. De ceci de\u0301coule une prise de conscience des relations temporelles et structurelles e\u0301tablies entre les gestes et les indices suprasegmentaux, suivie d\u2019une rede\u0301finition de la prosodie des langues vocales comme une composante audiovisuelle. Face a\u0300 ces constats, nous proposons d\u2019e\u0301tudier de concert les marqueurs manuels et non manuels des langues vocales et des langues signe\u0301es utilise\u0301s a\u0300 des fins prosodiques. Cette comparaison, encore ine\u0301dite, pose un certain nombre de questions me\u0301thodologiques auxquelles cet article souhaite re\u0301pondre.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div> Mariana Montes<\/strong><\/p>\n<\/div>\n<\/div>\n<\/div>\n The present study is part of a larger research project with the goal of developing tools for large-scale corpus-based semantic analyses. One such tool involves representing semantic structure with vector space models (VSMs), which currently requires a deeper understanding of its inner workings and how its results relate to cognitive theories of meaning. As a distributional method, it builds on the context of a lexical item to describe it and compare it to others, which raises our main research question: how is that context defined, and how does it vary for different semantic phenomena? We investigate how models based on different parameter settings deal with a range of semantic issues, such as granularity of meaning. For this purpose, a set of 7 polysemous homonyms in Dutch was selected to test how the distance between homonymous usages of a lexical item are represented in relation to polysemous usages within the same homonym. The models were built from a 520MW corpus of contemporary Dutch and Flemish newspapers and the resulting VSMs were evaluated through visual analytics, via scatterplots where more similar tokens appear closer to each other. In addition, colorcoding of manual tags allows us to compare how they were grouped by human annotators and by the computational models in a way that is consistent with the cognitive approach to meaning and categorization. The results indicate that not one set of parameters deals well with granularity in all cases. For example, those that disambiguate stof \u2018substance\/fabric\/topic, dust\u2019 well fail with hoop \u2018heap, hope\u2019 and vice versa. Furthermore, for some nouns some senses may be well grouped while the homonyms are not.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>