Expressing stance in parliamentary debates: A French-Spanish corpus-driven study
(2018) Using Corpora in Contrastive and Translation Studies (5th edition) — Location: Louvain-la-Neuve (12.September.2018)
Files
UCCTS2018.pdf
Open Access - Adobe PDF
- 84.78 KB
Details
- Authors
- Abstract
- This presentation reports on an exploratory, corpus-driven study of French and Spanish stance expressions in debates held at the European Parliament. Stance is here understood as “the speaker’s evaluative, epistemic or affective positioning towards a stance object” (Kärkkäinen 2012: 2195). More specifically, our paper focuses on recurrent multi-word stance expressions, such as French à mon avis ‘in my opinion’, il me semble + ADJ + de ‘it seems + ADJ + to’, il est essentiel de ‘it is essential to’ and Spanish yo creo que ‘I believe that’, me gustaría ‘I would like to’, es necesario ‘it is necessary’, and takes stocks of insights from both contrastive pragmatics (Aijmer 2011) and contrastive phraseology (Ebeling & Oksefjell Ebeling 2013, Granger 2014). Our study relies on comparable Europarl subcorpora of European parliamentary proceedings in French and Spanish (see Koehn 2005), corresponding to approximately 10 years of debates (up to 2010) and 3 million tokens per language. The subcorpora used in this study are restricted to verbatim reports of speeches originally delivered in French and Spanish by Members of Parliament (see Cartoni & Meyer’s directional version of Europarl, which relies on Europarl’s language tag). In other words, they do not contain any translated texts. Even though it must be acknowledged that French and Spanish are occasionally used by non-native speakers in the European Parliament, in the vast majority of cases, the two languages are used by Members of Parliament from France, French-speaking Belgium and Spain, respectively. In this paper, we adopt a corpus-driven approach to stance. The stance data were obtained by automatically extracting n-grams (also called ‘lexical bundles’), which are recurrent sequences of n contiguous words, i.e. “sequences of word forms that commonly go together in natural discourse” (Biber et al. 1999: 90). As pointed out by Granger (2014: 69), n-grams “are a powerful window onto pragmatics and rhetoric. It is undeniably a quick-and-dirty method, but one that has great heuristic power: it generates a multitude of word sequences that have so far received very little interest in the contrastive literature”. In this study, we have extracted 2- to 5-grams with a minimum frequency of 50 occurrences per million words. The automatically retrieved n-grams were then manually analyzed in context so as to identify stance expressions and patterns. Methodologically, the comparative analysis of French and Spanish, which are arguably quite close, being two Romance languages, raises a number of issues. One of them is that Spanish is a pro-drop language, while French is not. As a result, Spanish stance expressions containing a conjugated verb form may either be 1- or 2-grams, depending on the optional overt expression of the subject (e.g. (yo) creo ‘I believe’, (nosotros) pensamos ‘we think’), while in French, corresponding expressions typically contain two words (je crois ‘I believe’, nous pensons ‘we think’). In order to ensure an optimal cross-linguistic comparability of the datasets analyzed, a number of additional single-word stance expressions corresponding to the multi-word sequences identified through the n-gram approach were also extracted (e.g. lamento/je regrette ‘I regret’, desearía/je souhaiterais ‘I would like to’). The presentation of the results will proceed in three steps. First, we will provide a structured, contrastive overview of the hundreds of stance expressions uncovered through the n-gram approach adopted in this paper, paying special attention to well-known French-Spanish morphosyntactic contrasts, such as the compulsory subject expression in French vs. the pro-drop character of Spanish, and the wider use of verbs with an experiencer dative in Spanish (e.g. me gustaría decir ‘I would like to say’; cf. Vázquez Rozas 2016). Second, we will zoom in on the different uses of formally similar or cognate forms, such as croire/creer ‘believe’ and penser/pensar ‘think’. Indeed, as repeatedly shown in studies on French and English political interviews (Fetzer & Johansson 2010), English and Spanish parliamentary enquiries (Marín Arrese 2015) and Catalan and Spanish parliamentary debates (De Cock & Nogué Serrano 2017), the use of cognate forms and seemingly similar patterns may in fact differ significantly across languages, mainly in terms of frequency, distribution and pragmatic/argumentative functioning. Finally, as noted by Goethals & Blancke (2014) in relation to thanking in French, Spanish and Dutch in the European Parliament, speakers of different languages adhere to different discursive conventions, even within the same parliament. In our paper, we will examine if and to what extent multi-word stance expressions can be used to uncover the variation of discursive conventions concerning stance-taking across linguistic communities. In our conclusion, we will sketch our future work, which mainly consists in extending the analysis to other languages (Dutch and English) and additional registers (newspaper editorials and research articles; cf. Neumann 2013, 2014) and in using parallel corpus data to analyze the impact of translation on some of the stance-related discursive and pragmatic traits of parliamentary proceedings.
- Affiliations
Citations
De Cock, B., & Lefer, M.-A. (2018). Expressing stance in parliamentary debates: A French-Spanish corpus-driven study. Using Corpora in Contrastive and Translation Studies (5th edition), Louvain-la-Neuve. https://hdl.handle.net/2078.5/125612