Interview in Le Figaro (one of the top three national newspapers in France) around the publication of my book "Savons-nous vraiment parler? Du contrat linguistique comme contrat social" in April 2023 (Paris: Armand Colin/Dunod).

La charge fonctionnelle d'un contraste phonémique renvoie au rôle informationnel joué par ce contraste dans une langue donnée. Plusieurs méthodes existent pour estimer la charge fonctionnelle (pour des revues de littérature, voir Oh et al., 2015 ; Wedel et al., 2013b), dont les plus communément utilisées sont (a) la quantification de la perte informationnelle résultant d'une fusion des deux phonèmes considérés, appréhendée par des mesures d'entropie (ex. Eychenne & Jang, 2018), et (b) le compte de paires minimales générées par le contraste (ex. Eychenne & Jang, 2018 ; Wedel et al., 2013a). Les mesures de charge fonctionnelle peuvent être utiles pour l'élaboration de programmes de correction phonétique destinés à des apprenants de langue étrangère. Elles permettent en particulier de cibler, parmi les contrastes posant des difficultés aux apprenants, ceux qui jouent un rôle prépondérant pour la distinction des unités linguistiques dans la langue cible, et qui, s'ils sont mal maîtrisés, ont par conséquent plus de chances d'aboutir à des problèmes de communication. La présente étude se situe dans la perspective du français langue étrangère (FLE), et a pour objectif de vérifier l'utilisabilité de la base de données Lexique (New et al., 2001; 2004 ; 2007) pour estimer la charge fonctionnelle des contrastes consonantiques et vocaliques du français. Plus précisément, la méthode utilisée a consisté à (i) modifier les transcriptions phonétiques de la base Lexique 3.83 afin qu'elles soient représentatives du système phonético-phonologique du français hexagonal contemporain de référence, et (ii) à repérer et effectuer le décompte des paires minimales générées par chacun des contrastes consonantiques et vocaliques du français. La méthode des paires minimales a été préférée à celle des calculs d'entropie car, dans le contexte de l'enseignement/apprentissage des langues étrangères, elle permet non seulement de quantifier la charge fonctionnelle des contrastes phonémiques, mais aussi de lister les paires de mots pouvant être utilisées pour créer des supports pédagogiques. Durant cette communication, nous présenterons en premier lieu les divers traitements automatiques effectués sur la base Lexique, en justifiant nos choix quant aux contrastes phonétiques considérés ainsi qu'à la façon de gérer les questions de la réalisation des schwas et du traitement de la polysémie. Nous présenterons ensuite les charges fonctionnelles obtenues pour les contrastes phonémiques du français, tout en tenant compte de facteurs pouvant nuancer l'interprétation des résultats, comme la fréquence d'occurrence des unités linguistiques considérées, ou encore le fait de ne traiter que les unités partageant une même catégorie grammaticale. Nous terminerons cette présentation par des perspectives d'application des résultats au cas des apprenants japonais de FLE.

Récemment, dans le domaine des études linguistiques, deux événements fort distincts ont semblé affecter la didactique des langues : l’un mineur, l’autre majeur. Le premier asséna une volée d’incertitudes à l’esprit de certains jeunes chercheurs ; le second promit des miracles aux béotiens. Bien qu’indépendamment mis en mouvement, ils reflètent tous deux les dangers d’une évolution conceptuelle ou technologique fulgurante insuffisamment ancrée dans l’histoire d’un domaine ou l’avenir d’une société. Le premier renvoie au rejet théorique de la notion de « locuteur natif » par une génération de jeunes expérimentalistes épris de plurilinguisme d’une part et de considérations politico-idéologiques d’autre part ; le deuxième à l’ouverture au grand public d’outils reposant sur des technologies dites d’Intelligence Artificielle Générative dans le traitement, la traduction et la production de textes en diverses langues parlées. Dans cette communication, sur la base de travaux menés dans le domaine de la phonologie de corpus et de son utilisation dans la didactique du français, nous proposerons quelques éléments de réflexion relatifs au statut du locuteur natif dans la didactique de la prononciation en français, tenant compte tout aussi bien de la diversité des usages dans l’espace francophone que des développements en cours dans le domaine de la parole technologiquement produite ou modifiée. Références Detey, S. (2023a). Savons-nous vraiment parler ? Du contrat linguistique comme contrat social. Paris : Armand Colin. Detey, S. (2023b). Entre théorie de l’unité linguistique et expérience de la variation langagière : quand la transcription des corpus oraux nous rappelle l’importance de la didactique de la prononciation. Revue Repères-Dorif 28, 1-16. Detey, S., Durand, J., Laks, B. & Lyche, C. (éds), Y. Kawaguchi, N. Yazu, H. Akihiro & K. Sugiyama (tr.) (2019). フランコフォンの世界. コーパスが明かすフランス語の多様性. Tokyo : Sanseido Press. Detey, S., Racine, I., Kawaguchi, Y. & Eychenne, J. (éds) (2016). La prononciation du français dans le monde : du natif à l’apprenant. Paris : CLE international. Detey, S. & Nespoulous, J.-L. (2023). De la psycholinguistique et des corpus oraux en didactique de la parole en langue étrangère : à travers les générations… In R. Renard, M. Billières, B. Harmegnies, J. Murillo Puyal, J. Hardane & B. Pavelin Lesic (éds), Paul Rivenc. Un précurseur de la didactique des langues et des cultures. Mons : CIPA, 65-80.

https://www.reseau-canope.fr/plurilinguisme-en-classe/mercredi-27-mars-2024.html

Be it “hate speech”, “abusive language”, “inappropriate language”, “verbal misbehavior” or overall “verbal miscommunication”, the underlying question at stake is the following: does anyone really know how to speak? While children are taught to read and write, while specialists are hired to rate the verbal appropriateness of movie dialogues or to “moderate” SNS content, and while legal actors and policy makers determine whether such or such verbal expression falls under such rule or such law, the debates around the importance of speech education in social communication are next to none. We know why: the value of freedom of speech, the fear of dictatorial censorship, the uncertain impact of neopuritanism (whether on one or the other side of current ideological spectra), the profitability and unfettered democratization of speech-mediating technologies, the overall lack of metalinguistic awareness, the fallacy of global communication, worldwide educational discrepancies… Yet, with AI-based speech on its way to pervading most digital spheres through automatic textual production, it seems more urgent than ever to reexamine social communication issues, including “hate speech” and “abusive language”, through the lenses of language awareness and global linguistic education (from sociophonetic indexicality to untranslatability), with a specific focus on the distinction between private and public speech, as well as national and international communication. Ultimately, this may lead to the necessity of setting up an explicit “linguistic contract” without which our “social contract” may be jeopardized, as is explained in “Savons-nous vraiment parler? Du contrat linguistique comme contrat social” (Armand Colin, 2023). Organized at the IEE-ULB by the Waseda Brussels Office, this interdisciplinary workshop will bring together dozens of scholars from both Europe and Japan working on “Hate Speech” and “Abusive Language”. Entitled “Hate Speech and Abusive Language: The Investigation of Verbal Violence, Regulation, and Education“, the workshop is co-organized by the School of International Studies (SILS) at Waseda, and the Institut d’études européennes (IEE) of the Université libre de Bruxelles (ULB), with Professor Sylvain Detey (Waseda University – SILS) serving as its’ main academic chair, alongside professor paul Bacon (Waseda – GSAPS) and Norimasa Morita (Waseda – SILS). Supported financially by Waseda and the Institut d’études européennes, the workshop aims, through a multiplicity of angles, to shed some light on the complexity of contemporary “verbal violence” and, maybe, to offer some suggestions as to how to look for a solution to the problems it creates. It is set to welcome two dozen European and Japanese researchers from 5 disciplinary panels. Each of these panels will unpack current trends surrounding “Hate Speech and Abusive Language” from a specific disciplinary perspective, including: (i) linguistic and education studies, (ii) political science, (iii) sociology, (iv) film studies, (v) media studies and international relations. Each of the 5 panels will see 2-to-4 scholars present their work followed by a Q&A with the attending audience. Contributors to the panels hail both from Europe and Japan; represent different disciplines, and are at varying stages of their respective research careers.

En 2023, année de la Cité internationale de la langue française, des États généraux de l’information et de l’irruption de l’IA générative dans l’arène informationnelle et éducative du grand public, un dialogue entre journalistes, linguistes, didacticiens, philosophes et politiques semble nécessaire : afin de renouer le dialogue national, de recadrer la communication internationale, de penser la formation à la communication mondiale et numérique de demain entre Paris, Tokyo, et le reste du monde. C’est à ce dialogue qu’invite cette rencontre, c’est à cette attention portée à nos comportements communicatifs qu’il doit mener : à l’heure où tout un chacun peut s’ériger en modèle de parole publique et influencer les comportements sociaux des uns et des autres, la société s’interroge sur leur bénévolence et leur optimalité dans la quête collective de bonheur et de liberté. « Les normes ne sont pas les ennemis de la libre communication » écrit le linguiste Jean-Louis Chiss dans la recension de l’ouvrage « Savons-nous vraiment parler ? Du contrat linguistique comme contrat social » : quelles en sont les conséquences pour les enseignants et les journalistes ? Les modèles de parole qu’ils offrent à leurs élèves et à leurs auditoires dans leurs pratiques sont-ils à la hauteur des enjeux ? Qu’en est-il des « influenceurs » et de leurs « suiveurs » ? Parlons-nous toujours « bien » ou « à bon escient » ? Et si non, pourquoi ne pas nous taire ? Des cours de récréation aux bancs de l’Assemblée en passant par les plateaux de télévision, la question est donc posée : savons-nous vraiment parler ? À l’heure d’internet et de la parole mondialisée, n’est-il pas temps d’expliciter un contrat linguistique raisonné sans lequel notre contrat social risque de s’effriter ? Conférence : Savons-nous vraiment parler ? Du contrat linguistique comme contrat social. Table Ronde : Quelle éducation à la parole publique, à la communication sociale et au traitement de l’information plurilingue internationale face à la violence verbale, aux tensions sociétales et à la mésinformation mondiale ? Informations: http://ffj.ehess.fr/upload/Actualites/Events/2023/2023-11-21_SDetey_Programme.pdf https://etats-generaux-information.fr/actualites/l-evenement-de-l-ehess-savons-nous-parler-s-organise-dans-le-cadre-des-egi Références : Antheaume, A. (2013). Le journalisme numérique. Paris : Presses de Sciences Po. Atlan, M. et R.-P. Droit (2023). Quand la parole détruit. Paris : Éditions de l’Observatoire. Bonnissent, J.-C. et P. de Sinety (coord.) (2021). Pour des sciences en français et en d’autres langues. Paris : Honoré Champion. Cerquiglini, B. (2018). Une langue orpheline. Paris : Les éditions de Minuit. Detey, S. (2023). Savons-nous vraiment parler ? Du contrat linguistique comme contrat social. Paris : Armand Colin. Develey, A. et J. Pruvost (2020). Les mots qui ont totalement changé de sens. Paris : Le Figaro. Durand, J. et C. Lyche (2024). Paul Passy, un linguiste révolutionnaire : réforme de l'orthographe, enseignement des langues et alphabet phonétique international. Limoges : Lambert-Lucas.

On November 21, 2023, Professor Detey Co-organized a Round Table in Paris at Maison Suger with the France-Japan Foundation of the EHESS around his book “Do we really know how to speak. The Linguistic contract as a social contract” with key figures in the field of linguistic policies and education, communication, and journalism. The program of the event was the following one: -Introductory remarks by Prof. Sébastien Lechevalier, Director of the EHESS France-Japan Foundation -Open Lecture by Prof. Sylvain Detey “Do we really know how to speak?” -Round Table moderated by Dorian Grelier (journalist at Le Figaro) with:  Alice Antheaume, Director of the School of journalism of Sciences Po Paris  Monique Atlan, journalist, and essayist  Bernard Cerquiglini, linguist (U. of Paris), former Rector of the Agence Universitaire de la Francophonie  Paul de Sinety, Délégué Général à la Langue Française et aux Langues de France (DGLFLF, Ministry of Culture)  Sylvain Detey, linguist (Waseda University)  Alice Develey, journalist (Le Figaro)  Roger-Pol Droit, philosopher, and columnist (former researcher at CNRS) -Comments by Prof. Jacques Durand, linguist, Emeritus Member of the IUF -Closing remarks by Christophe Deloire, journalist, Head of Reporters without Borders, General Secretary of Etats généraux de l’information. Almost 200 people registered for the event online and on-site, and it was included in the Etats généraux de l’information. Part of the event was sponsored by JSPS Kaken (B) n°20H01291 (S. Detey).

The presence of a Foreign Accent (FA) in the communicative process may interfere with the correct transmission of a message. This interference can arise from various traits, such as rhythm, intonation and speech rate. This can affect perception by both L1 and L2 listeners. Therefore, in this study we focused on the segments of Japanese FA in English as perceived by French listeners. Seven consonants ([l, ɹ, th, kh, θ, f, v]) were extracted from the beginning of CVC English words. Using sound manipulation techniques, they were transformed into their most expected Japanese realisation ([ɾ, ɾ, t, k, s, ɸ, b] respectively) to generate a new set of words in which only the first segment was accented. The French cohort showed clear differences in perceived intelligibility of rhotic consonants and the [f/ɸ] contrast compared with previously analysed groups of Japanese and American listeners.

This study investigates the possibility to use automatic, multi-level features for the prediction of L2 speech proficiency. The method was applied on a corpus containing audio recordings and transcripts for 38 Japanese learners of French who participated in a semi-spontaneous oral production task. Each learner's speech proficiency level was assessed by three experienced French teachers. Audio recordings were processed to extract features related to the pronunciation skills and phonetic fluency of the learners, while the transcripts were used to measure their lexical, syntactic, and discursive abilities in French. A Lasso regression using a leave-one-out cross-validation procedure was used to select relevant features and to accurately predict speech proficiency scores. The results show that five features related to the phonetic fluency (speech rate), lexical abilities (lexical density), discourse planning and elaboration skills (number of hesitation and false starts, mean utterance length) of the learners can be used to predict speech proficiency ratings (r=0.71, mean absolute error on a 5-point scale: 0.53).

With the development of corpus linguistics and data-based applied linguistics in the 1990s, the number of learner corpora worldwide has risen steadily over the past two decades, with a strong focus on lexical and grammatical constructions. It is only recently that the movement bridging corpus linguistics and second language phonology started to gain momentum (Gut 2014), especially from an educational perspective (Biber 2019, Detey 2020). In the field of French corpus phonology, the PFC project (Phonologie du Français Contemporain: usages, variétés et structure), which was launched more than twenty years ago (Durand, Laks, Lyche 2002), stands as a rather unique research program. Originally designed to refresh L1 French phonology data through a Labovian network of comparable surveys across the French-speaking world, it has led to both educational (PFC-EF) and SLA research-oriented (IPFC) (Detey & Kawaguchi 2008; Racine et al. 2012) offshoots over the past 15 years. Adjusting the PFC research protocol and processing guidelines (Racine et al. 2011) from native to non-native data proved to be both challenging and inspiring vis-à-vis the objectives and the multidisciplinary nature of the L2 branch of the project. In this presentation, we will describe how these adjustments took place, with a particular focus on the tasks involved in our recording protocol as well as on the value of data comparability across different populations of native and non-native speakers, which we will illustrate with data from both the PFC and the IPFC corpora. References Biber, D. (2019). Corpus analysis of spoken discourse. Research findings, prospects, implications for teaching. Plenary talk. PSLLT2019. Flagstaff: Northern Arizona University. Detey, S. (2020). Teaching and language corpora: what of pronunciation? Insights from French. Plenary Talk. 14th Teaching and Language Corpora conference (TaLC2020). Perpignan: University of Perpignan. Detey, S. & Kawaguchi, Y. (2008). Interphonologie du Français Contemporain (IPFC) : récolte automatisée des données et apprenants japonais. Journées PFC : Phonologie du français contemporain : variation, interfaces, cognition. Paris : MSH. Durand, J., Laks, B., and Lyche, C. (2002). La phonologie du français contemporain : usages, variétés et structure. In: C. Pusch, and W. Raible (eds), Romance Corpus Linguistics – Corpora and Spoken Language. Tübingen: Gunter Narr Verlag, 93-106. Gut, U. (2014). Corpus phonology and second language acquisition. In: J. Durand, U. Gut, and G. Kristoffersen (eds), The Oxford Handbook of Corpus Phonology. Oxford: Oxford University Press, 286-301. Racine, I., Detey, S., Zay, F., and Kawaguchi, Y. (2012). Des atouts d'un corpus multitâches pour l'étude de la phonologie en L2 : l'exemple du projet « Interphonologie du français contemporain » (IPFC). In: A. Kamber, and C. Skupiens (eds), Recherches récentes en FLE. Bern: Peter Lang, 1-19. Racine, I., Zay, F., Detey, S. and Kawaguchi, Y. (2011). De la transcription de corpus à l'analyse interphonologique : enjeux méthodologiques en FLE. In : G. Col, and S. N. Osu (eds), Transcrire, écrire, Formaliser. Rennes : Presses Universitaires de Rennes. Travaux Linguistiques du CerLiCO 24 : 13-30.

This study focuses on the automatic evaluation of the speech of Japanese learners of French. Our data are semi-spontaneous oral productions of learners from the CLIJAF corpus. We evaluate different linguistic levels (pronunciation, lexicon, syntax, discourse) using automatic assessment tools. Our first objective is to study whether these measures are in line with the general level of proficiency of the learners, as defined by the CFRL (A2 to C2). Our second objective is to explore the possibility of predicting this level using the most relevant linguistic parameters, with an unsupervised machine learning algorithm. The results confirm the relevance of our objectives and show that automatically extracted parameters such as lexical diversity, the average length of speech turns (syntactic level), the proportion of connectives in the discourse and several parameters related to fluency can contribute to the prediction of learners' level.

Although Japan has often been considered a prototypical monolingual society for a long time, it has been steadily evolving into a more diverse country over the past decades, especially in large urban centres like Tokyo, due to the influx of foreign students, trainees, workers and tourists, as has been witnessed by the evolution of its multilingual linguistic landscape (Backhaus 2005, 2006). Tokyo, the country's capital city where one out of five foreign residents of Japan live, with a total population of about 14 million inhabitants including around 550 000 holders of foreign passports, has been gradually acknowledging the reality of its multilingualism and increasing its emphasis on the positive value of plurilingualism, for economic, educational, cultural, and political reasons, against the demographic backdrop of a rapidly graying population. However, much of the emphasis was initially set on English, depicted as a “global language”. It is only recently that the potential impact of other languages has started to receive some renewed attention in the public sphere, especially through the concept of “glocalisation”, which was introduced in the 1980s to tackle the heterogeneity effects linked to globalisation (Robertson 1994), for example when global services and local interests are mutually adjusted. Given that such developments cannot rely exclusively on top-down public linguistic policies, and that the general public's opinion matters at least as much, especially in a country like Japan where social trends are closely scrutinized by the overall consensus-seeking social community, it is important to assess how the next generation of Japanese and non-Japanese Tokyoite citizens perceive the value of plurilingualism in Tokyo, beyond the “all-English” approach. In this presentation, after a brief introduction to Japan's and Tokyo's linguistic history (Shibatani 1992; Maher & Yashiro 1995), we present the result of a pilot survey (N=402) carried out in 2020 in several universities in Tokyo. This survey explores the representations and attitudes of Japanese and non-Japanese students towards some of the main languages found in Tokyo (e.g. English and its variation; mandarin Chinese, Korean, Filipino, Portuguese, Vietnamese, Thai, Hindi and Urdu, French, German, Italian, Russian and Spanish), in connection with their historical and contemporary relation to Japan, in terms of economic, social and cultural dimensions. Our conclusion aims at highlighting potential tracks for educational developments (Haida 2010; Tajiri & Otsu 2010; Ostheider 2012; Nishiyama 2017; Vollmer & Thanyawatpokin 2019; Detey to appear) which could target a higher degree of plurilingualism in Tokyo beyond English, while at the same time preserving the social value of a national linguistic cohesion which has been remarkably cultivated in Japan since Tokyo became the capital city of the country and “standard Japanese” its national language (Inoue 2011).

The Japanese higher education system is undergoing profound changes due to Japan’s demographic decline and worldwide globalization of higher education standards (Hiroshima University, 2020). Universities are increasingly looking for international students and teachers, using English as a lingua academica and for internationalizing their syllabi and their academic reputation, with the support of the Ministry of Education, Culture, Sports, Science and Technology (MEXT) (Vickers, 2018). This evolution gradually turned the concept of global citizenship education (GCE) (Tawil, 2013) into an increasingly popular notion, especially through the revival of liberal arts departments and the proliferation of English-taught degree programs across the country. However, this situation highlights a gap existing between the inherent characteristics of the Japanese higher education system and GCE. For example, Japanese students may excel in mastering knowledge, but find it difficult to put into practice what they have learned, especially in terms of oral interaction in a foreign language. The GCE agenda could, in this vein, be used in Japan for promoting among students the development of adaptability and confidence in such interactions. In 2014, the MEXT created a Top Global University project in order to support universities’ internationalization, in a country traditionally depicted as strongly monolingual and monocultural, and in which authorities tend to see globalization as a monolingual English-centered process. Our paper shows how such initiatives can help integrate the values of multicultural perspectives and plurilingual abilities into an education framework through which students can acquire a form of global competence. The School of International Liberal Studies (SILS) of Waseda University illustrates the tension between a monolingual and a plurilingual understanding of globalization. Created in 2004, it rapidly became a leading department in the promotion of GCE in Japan. At first its main message was to promote—to an essentially domestic Japanese audience—the value of global understanding through higher education in English. Gradually, the school smoothly and successfully evolved from monolingualism to plurilingualism, thanks to new possibilities offered by English as a common tool for education and research. Our analysis focuses on the French Studies component of the Area Studies and Plurilingual-Multicultural education program (APM), created in SILS in 2017, as part of the Top Global University project. More specifically we will analyse its Content and Language Integrated Learning (CLIL) courses (compare with Davies, 2017) dedicated to French political science, which aimed at creating conditions for all students to participate in the co-construction of meaning and overall understanding of course content. This paper discusses outcomes of the APM program including supporting students to think together about comparable objects in a comparative perspective; classroom interactional competence in content and language integrated learning (Escobar Urmeneta & Walsh, 2017); understanding their responsibility towards the classroom community; and building up a local public opinion on transnational normative issues (compare with Torres & Bosio, 2020). The described APM program serves an example of democratic education in line with GCE goals in its efforts to further student engagement in classroom contexts with questions that pertain to social justice and are central to the public sphere function (Habermas, 2015) that education serves in Japan.

Despite the seminal contributions of 19th century European phoneticians to foreign language (L2) education methodology (e.g. Paul Passy and the founders of the International Phonetic Association, cf. Durand, 2017; Galazzi, 2002), the ‘pronunciation’ component in L2 education has almost always been overshadowed by its lexical and grammatical counterparts in the fields of Second Language Acquisition (SLA) and L2 educational methods. However, thanks to the work accomplished over the past 10 years in L2 phonetics and phonology, pronunciation has turned – according to some prominent scholars in the field – from being the “Cinderella” of SLA to being the “Belle of the Ball” (Derwing, 2019: 27), aview which must still find its way into the subfield of applied corpus linguistics and data-driven learning (DDL) and teaching. In fact “pronunciation” is not included in the language aspects (vocabulary, lexicogrammar, grammar, discourse) examined in the meta-analysis of corpus use in language learning carried out by Boulton and Cobb (2017: 380); it is absent from the introductory article in a special issue of Language Learning and Technology dedicated to corpora in language learning and teaching (Vyatkina & Boulton, 2017); and it seems to be only significantly mentioned once (Meunier, 2019) in a recent volume dedicated to data-driven learning “for the next generation” (Crosthwaite 2019). Likewise, in the TaLC2018 book of abstracts, only one paper (out of 173) examines pronunciation issues in an English for Specific Purposes context (Staples 2018), while another one simply mentions the pronunciation dimension in the description of its spoken corpus data collection protocol (Abe, Kondo, Kobayashi, Murakami & Fujiwara, 2018). Therefore, five years after the publication of the Oxford Handbook of Corpus Phonology (Durand, Gut & Kristoffersen, 2014) – which includes one chapter on corpus phonology and SLA (Gut 2014) – it seems to be the right time for a plenary talk on the topic ofPronunciation and Teaching and Language Corpora – especially as, in the meantime, international pronunciation-focused conferences such as Pronunciation in Second Language Learning and Teachingwelcome, as plenary speakers, famous corpus linguists such as Douglas Biber (whose “former life” actually included articles on accent in Somali and contour tones, Biber 2019). The core of this talk – adding an extra French flavor to the Perpignan-based TaLC2020 – will rely on the output of the international and collaborative work carried out within the framework of the PFC project (phonology of contemporary French, Durand, Laks & Lyche, 2002), especially its education-oriented applied offshoots, namely PFC-EF (PFC for French education, Detey, Durand, Laks & Lyche, 2016) and IPFC (Interphonology of Contemporary French, Detey, Racine, Kawaguchi & Eychenne, 2016). After a short presentation of the historical foundations of the project and its resulting corpus (https://www.projet-pfc.net/), the talk will tackle three main issues: 1) the relevance of spoken corpora and DDL for L2 pronunciation teaching and learning; 2) the relevance of spoken corpora and DDL for language awareness education both in L1 and L2 at the pronunciation level and beyond; 3) the challenges embedded in the design of a spoken corpus-based L2 pronunciation education programme in the light of contemporary corpus-based and DDL-oriented pedagogy (Boulton & Tyne, 2014, Tyne et al. 2014). While Biber (2019) tried to convince applied phonologists/phoneticians at PSLLT2019 that they needed to know grammar to teach pronunciation, this talk will aim at highlighting the fact that corpus linguistics and pronunciation teaching are not so disconnected, especially when it is looked at from apsycholinguistic (hybrid models of speech processing, Nguyen, Wauquier & Tuller, 2009) and a 10 sociolinguistic (variationist approach of speech, Durand, Laks & Lyche, 2016) perspectives (Detey, 2009, 2010). This, in turn, should stress the importance of pronunciation skills (both in perception and in production) in the (initial) acquisition of the lexicogrammatical component of the L2 at stake, andtherefore possibly convince applied corpus linguists that they cannot fully overlook pronunciation to teach grammar...

This study compared automatic assessments of L2 phonetic fluency of Japanese learners of English in read speech, with ratings by native and non-native English assessors, and considers whether the assessors’ first language affects the results. Speech data of 183 Japanese and 25 native English speakers’ readings of “the North Wind and the Sun” were assessed for phonetic fluency by 16 trained assessors with different first languages (four American English, four Japanese and eight other languages). They rated segmental accuracy, prosody, fluency and nativelikeness. A subset of 97 of the speakers’ data (the 25 native English speakers and 72 randomly selected Japanese speakers) was also used to develop an automatic fluency assessment system. The 97 speakers’ data were re-assessed by four different trained American raters. The correlation between the automatic evaluation and the four raters was 0.83. When the automatic system was then tested on the remaining original 111 speakers’ data and the original 16 assessors’ scores, it showed correlations of 0.62-0.67 for the American, Japanese and other language raters. The results suggest that the automatic assessment system can assess phonetic fluency of Japanese-accented English quite reliably, and that native and non-native evaluators used different phonetic cues to evaluate fluency.

For the past fifteen years, CLIL has become a trendy acronym to insert and display in foreign language education curricula worldwide along with bilingual education programs. However, despite the growing availability of training and teaching resources and material, putting CLIL into practice - as well as understanding its function and potential scope – remains challenging, not only for language teaching specialists but also for teachers of academic subjects other than languages. Having a comprehensive perspective on the CLIL approach bridging the fields of language teaching methodology on the one hand (especially in the case of Language for Specific Purposes, LSP) and plurilingualism in higher education on the other hand (especially in a Japanese context) might shed some light on the value of CLIL implementation in Liberal Studies programs such as SILS, and this is what we will be discussing during the presentation.

This research work investigates the possibility of using automatic acoustic measures to assess speech fluency in the context of second language (L2) acquisition. To this end, three experts rated speech recordings of Japanese learners of French who were instructed to read aloud a 21-sentence-long text. A Forward-Backward Divergence Segmentation (FBDS) algorithm was used to segment speech recordings (sentences) into acoustically homogeneous units at a subphonemic scale. The FBDS processing results were used — along with more classic measures such as raw percentage of speech and length/standard deviation of silent pauses — to estimate speech rate and regularity of speech rate, while a formant tracking algorithm was used to estimate speech fluidity (i.e., quality of coarticulation). A step-by-step multiple linear regression was finally computed to predict the experts’ mean fluency ratings. Results show that FBDS-derived measures, raw percentage of speech and standard deviation of the first formant curve derivative can be combined together to calculate accurate estimates of speakers’ fluency scores (R = .92; P < .001). As only low-level signal features were used in the study, the method could also be relevant for the assessment of speakers of other target languages, as well as for the assessment of disordered speech.

The effect of studying abroad on the acquisition of schwa in advanced Swiss German learners of French. Mastering ‘schwa’ is one of the major challenges that non-native speakers face when learning French as a foreign language. The present study, set within the framework of the IPFC project (“Interphonologie du Français Contemporain”), describes the schwa-coding system applied to the IPFC corpus data and is illustrated here with data from Alemannic-speaking Swiss learners of French (two tasks: text reading and guided conversation). Our results show that, even though studying abroad in a French-speaking area alone does not have a direct significant impact on schwa’s behavior, it does interact with two variables: the type of task (more schwa deletion after studying abroad observed in the conversation but not in the reading task) and the type of word (more schwa deletion after studying abroad observed in monosyllabic words but not in polysyllabic words, both in word-initial or word-medial syllabic positions). In this respect, studying abroad seems to help learners achieve native-like sociolinguistic proficiency. These first results pave the way for the development of in-class pedagogical material which could help learners achieve such a level of proficiency, whether or not they have the opportunity to study abroad in a French-speaking area.

Dans cette intervention, nous offrirons une introduction aux variétés non natives du français en nous concentrons sur leur dimension phonétique et phonologique. Après une brève introduction relative à l’intérêt d’étudier les systèmes de production des locuteurs non natifs du français, ainsi qu’à l’apport des corpus oraux en la matière, la présentation sera structurée sous forme de présentation d’un ouvrage récemment paru sur le sujet, contenant non seulement des descriptions mais également des échantillons sonores de ces variétés. Notre conclusion s’orientera vers l’analyse de ces variétés à d’autres niveaux de description linguistique.

Les variétés de français parlé hors de France (Belgique, Suisse, Canada, Afrique) se caractérisent par la relation ambivalente qu’elles entretiennent avec le français pratiqué en France, et plus particulièrement à Paris (Moreau et al. 2007). Le bon usage – et donc la norme – est ainsi généralement situé hors de leur propre communauté par les locuteurs de la périphérie. Ce rapport centre/périphérie se retrouve à l’intérieur même de la périphérie (Singy 1996), faisant ainsi émerger la notion de norme endogène (voir Racine et al. 2013 pour la Suisse romande). Au Québec, le débat entre normes endogène et exogène est particulièrement vigoureux. Au niveau de la prononciation, un « français québécois standard » semble se développer, face au modèle parisien, perçu comme prétentieux par les Québécois (Remysen 2004). Dans ce travail, nous nous pencherons sur le rapport centre/périphérie au Canada francophone. Si certains chercheurs ont cherché à identifier les traits de prononciation qui permettent de différencier le français parlé au Québec du français hexagonal (Brasseur & Ménard 2009), très peu de travaux se sont intéressés à la description et à la perception des variétés à l’intérieur du domaine canadien (Côté 2014). Remysen (2014) fait exception, en montrant que les auditeurs québécois peuvent identifier l’origine montréalaise des locutrices. Toutefois, son test de perception se limite à la variété montréalaise et l’auteur lui-même souligne la nécessité d’élargir les recherches à d’autres variétés. Dans l’étude présentée ici, 96 auditeurs francophones issus de 4 régions du Canada (Ontario, ouest du Québec, région de la ville de Québec, est du Québec) et d’une région française (Paris) ont jugé les productions d’une phrase lue, réalisée par 6 locuteurs (3 hommes et 3 femmes, âgés de 20 à 40 ans, issus de 3 régions : ville de Québec, souvent perçue comme la norme au Canada, Hearst en Ontario et Paris). Ces données sont tirées d’une étude plus large comportant également des stimuli belges, suisses et africains, tous extraits de la base de données du projet PFC (Durand et al. 2009). Dans la première partie, les auditeurs ont dû indiquer, pour chaque stimulus, si la prononciation était identique à leur propre prononciation. Après une deuxième écoute, ils ont dû estimer si cette prononciation était celle du français internationalement reconnu comme « standard » et, enfin, dans la troisième partie, si cette prononciation était celle du français reconnu comme « standard » dans le cadre de leur vie quotidienne. Pour chaque questions, les auditeurs disposaient d’une échelle allant de 1 (= Absolument pas) à 5 (= Absolument). Des analyses statistiques permettront de déterminer l’existence d’une éventuelle norme endogène en perception pour le Canada et d’examiner le statut, au Canada, de la variété de type « parisienne » reconnue comme « standard » dans des études précédentes par des auditeurs français et suisses romands (Detey & Le Gac 2008; Racine et al. 2013).

This article presents the results of an AXB discrimination task testing the ability of Japanese beginner learners to distinguish three French nasal vowels /ɑ̃ ɔ̃ ɛ̃/ and compares them with the results of a corpus-based analysis of their production of the same vowels in two tasks (wordlist repetition and wordlist reading). This perception-production study is embedded in an ongoing 2-year longitudinal study in the framework of an international multi-task corpus-based research programme about the phonetic-phonological systems of non-native varieties of French. Our main results indicate first that the /ɑ̃/-/ɛ̃/ contrast is better perceived than the /ɑ̃/-/ɔ̃/ contrast and second that /ɛ̃/ obtains a lower discrimination rate when it is contrasted with /ɑ̃/, whereas no difference is found between /ɔ̃/ and /ɑ̃/. These findings are discussed in light of the production results and support the principle according to which perception precedes production in the acquisition of a new phonological contrast.

In this project, we developed our original survey questionnaires for Portuguese, Japanese and Turkish based on the French survey protocol cultivated through the International Project PFC (Phonology of Contemporary French). The survey is composed of six different tasks: (1) Repetition task of word list, (2) Reading task of word list, (3) Text reading task, (4) Interview task, (5) Free conversation task, (6) Composition task. We have collected learner language data for each language over four years. The learner language corpus constructed in this way was analyzed by contrasting it with the spontaneous conversation data of native speakers. The research was conducted from several viewpoints, 1) phonetic and prosodic features, 2) lexical features, and 3) discourse markers.

2012年度は、録音と知覚テストを2回に分けて実施し、並行してデータの分析と処理を開始した。まず、2012年6月にはIPFCプロトコルに基づいた手法を用いて、39人分の日本語話者を対象として、Moodleソフトを用いて録音を行った。また、フランス語の鼻母音 /an, in, on/、流音 /l, r/、および唇子音/b, v/の知覚についてAXBタスクを使った知覚テストを行い、また被験者がフランス語の音節数を数え上げるタスクを実施した。さらに、カタカナに変換したフランス語のテクストの読み上げるタスクを加えた。
2回目の録音およびテストは、2013年2月に行った。上記と同様のプロトコルに基づいた録音および知覚テストを実施したが、フランス語テクストをカタカナ変換したものを読み上げるタスクは二回目は行わず、その代わりとしてフランス語母音の/u-y/の弁別ができるかどうかの知覚テストを実施した。27人の学生がインフォーマントとして参加した。
フランス語データのコーディングのポイントは大きく分けて、（1）フランス語のターゲット音声要素、（２）要素の左右の音韻環境、（３）非母語話者発話音声の特徴（ターゲットの音声に近いか否か、母語話者と同じ音声特徴の有無など）、の抽出である。Praatソフトウェアを用い、各音声ファイルごとに、Textgridファイルに音声波形に一致する単語列を記述した。そして、フランス語母語話者が統一コード基準に沿って、音を聞き聴覚的に音素を判断し、アルファベットと数字の記号を用いて、手動でコードを記述した。
またこの科研研究のため、コーディングされたデータの自動分析処理及び統計処理のために、Julien Eychenne氏に依頼し、Dolmenソフトウェアを書いてもらった。

This study investigated speech rhythm produced by native and non-native Japanese speakers, using durational variables which are used to categorize languages by their speech rhythm. We collected Japanese read speech data of native speakers of Chinese, Vietnamese, Korean and French. We have accumulated 343 Japanese words containing tokushu-haku, which are geminate consonants, moraic nasals, long vowels, /CjV/ sequences, and devoiced vowels. The speakers' Japanese utterances were segmented and analyzed by durational variables: (1) durational ratios, (2) standard deviation, (3) coefficient of variation, (4) pairwise variability index, and (5) average of normalized voice onset asynchrony.
The analyses of these durational variables showed that individual variables were not effective enough to differentiate between native and non-native Japanese speakers' speech rhythm. At least two or often more than three variables were required to discriminate between the native and non-native groups.