Metadiscourse, or discourse about discourse, manifests itself in both spoken and written genres. However, the focus of existing studies of spoken metadiscourse remains mainly on academic speaking (Hyland, 2017), whereas research on metadiscourse in spoken non-academic genres is much less common. The present study thus investigates, both qualitatively and quantitatively, the use of metadiscourse in TED (Technology, Entertainment, Design) talks, which are primarily monologic speeches aiming at knowledge dissemination among a lay audience. The corpus for this exploratory study comprises the top 20 most viewed speeches (49,000 tokens and 286 minutes of presentation time) presented in English at an official TED conference in 2019. This study attempts at shedding light on how TED speakers use metadiscourse from the perspective of the discourse functions by applying Ädel’s (2023) taxonomy of reflexive metadiscourse. The results reveal that TED speakers demonstrate a high level of audience orientation. The findings also indicate a strong focus on discourse organization, which underlines the importance of making this explicit in TED talks, thus providing guidance for the audience. One of the key observations of this study is that the audience in the genre of TED talks is complex, as talks are not only presented live when recorded, but also made available to a global, online audience. The thesis aims to contribute insights to TED talks as a genre, enhance the taxonomy development for metadiscourse, and suggest potential pedagogical implications for teaching public speaking.

Cette thèse présente une étude sémantico-pragmatique des emplois dialogiques de mais, oui, non et de leurs quatre combinaisons : mais oui, mais non, oui mais et non mais. Les enchaînements dialogiques « X – Mais Y » posent la question de l’unicité du sens du connecteur oppositif, étant donné que la relation entre les énoncés X et Y est très différente de celle qui s’établit entre les énoncés dans les configurations monologiques « X mais Y ». Dans cette thèse, les deux types d’enchaînement avec mais sont analysés dans le cadre de la sémantique dynamique, qui propose de décrire le sens des expressions linguistiques à partir de leur capacité à modifier le contexte. L’opération énonciative de mais est spécifiée à partir des contraintes imposées sur le contexte linguistique environnant, que l’examen de divers usages contextuels de mais en monologue et en dialogue a permis de révéler. Cet examen conduit à conclure que la spécificité des enchaînements dialogiques avec mais n’est pas liée à la sémantique du connecteur, mais à la progression du discours, différente en monologue et en dialogue. Les combinaisons de mais avec oui et avec non sont étudiées après une analyse détaillée du marqueur de validation et du marqueur de rejet. La façon dont ces deux marqueurs interagissent avec le contexte gauche témoigne de leur asymétrie sémantique, confirmée par la possibilité de les utiliser en tant que formes concurrentes pour valider un énoncé négatif. Dans les combinaisons de oui et de non avec mais, le second élément détermine le type de relation signalée : oui mais et non mais marquent l’opposition, tandis que mais oui et mais non traduisent la validation et le rejet. Néanmoins, le premier élément a des incidences sur l’interprétation du second. Ainsi, les réactions mais oui et mais non présentent l’information validée ou rejetée comme déjà admise dans le contexte ou, au contraire, comme exclue et donc impossible à envisager. En revanche, dans les réactions oui mais et non mais, le connecteur oppositif enchaîne sur l’information accessible à la suite de l’emploi du premier marqueur. / This dissertation presents a semantic-pragmatic study of dialogic uses of mais, oui, non and their four combinations : mais oui, mais non, oui mais et non mais. The dialogic sequences “X – Mais Y” raise the question of the uniqueness of the oppositive connector, because the relationship between the utterances X and Y is very different from that which is established between both utterances in the “X mais Y” monologic configurations. In this dissertation, the two types of linking with mais are analyzed in the framework of dynamic semantics, which aims to describe the meaning of linguistic expressions from their ability to change the context. The enunciative operation of mais is specified from the constraints on the surrounding linguistic context that the examination of various contextual uses of mais in monologue and dialogue revealed. This examination leads to the conclusion that the specificity of dialogic sequences with mais is not related to the semantics of the connector, but the progress of speech, which is different in monologue and dialogue. Combinations of mais with oui and non are considered after a previous detailed analysis of validation marker and rejection marker. The way these two markers interact with the left context reflects their semantic asymmetry, confirmed by the possibility of using them as competing forms to confirm a negative utterance. In combination of oui and non with mais, the second element determines the type of relationship: oui mais and non mais mark the opposition, while mais oui and mais non reflect validation and rejection. However, the first element has a clear implication on the interpretation of the second one. In that sense mais oui and mais non responses present the validated or rejected information as already accepted in the context or, on the contrary, excluded, and therefore impossible to consider. On the other hand, in reactions with oui mais and non mais the oppositive connector regards the information accessible next to the use of the first marker.

This thesis presents a study of the relationship between test-takers’ first language, listening proficiency and their performance on paired speaking tests. Forty participants from two different L1 backgrounds (20 Urdu and 20 Thai) participated in the study. They took two paired speaking tests: one with a shared L1 partner, and one with a non-shared L1 partner, as well as a listening test and a monologic speaking test to measure their listening ability and individual speaking ability. After each paired speaking test, the participants were also interviewed about their test-taking experience. All speaking tests and interviews were video recorded and transcribed. Raters awarded test-takers analytical speaking test scores (grammar and vocabulary, discourse management, pronunciation and interactive communication) and provided comments to justify their scores. Raters also participated in a stimulated recall session. The mixed-methods approach was utilised in analysing and triangulating different data sources. The data analysed in this study included listening and speaking test scores, raters’ perceptions of the test-takers’ speaking performance gathered from stimulated recalls and test-takers’ stimulated recall interviews, as well as the interactional discourse data in the paired speaking formats. The combination of quantitative analysis, Conversation Analysis (CA) and thematic analysis informed the relationship between test-takers’ listening proficiency, their L1 and their paired speaking performance.

Developing speaking proficiency in English has been highly demanded in the field of English education in Japan; however, teaching speaking in academic settings is difficult because of its complex nature. Many Japanese high school students cannot organize their spoken production coherently because they have not been explicitly taught how to meet the expectation of particular contexts or genres. Research on genre-based instruction has shown its effectiveness in the development of reading, writing, and listening skills; however, investigations of genre-based instruction have not been fully applied to the teaching of speaking. The primary purpose of this study was to investigate the effectiveness of genre-based instruction to develop the academic speaking ability of Japanese high school students through a one-year longitudinal study. The effectiveness of genre-based instruction was assessed by focusing on the development of a macro-genre, academic monologic speech, and three micro-genres—procedure, definition, and causation—for within-group assessment as well as one oral summary of a research project micro-genre for between-group assessment. The research design was a multistage intervention mixed-method approach with qualitative data gathered after the experiment. Three analytical techniques were employed: (a) multi-faceted Rasch measurement (MFRM) was used to assess the extent to which the participants’ performance improved quantitatively, (b) descriptive analyses were used to investigate frequency changes in the use of target lexis, and (c) genre analysis was used to analyze how the discourse structure of the target genres changed qualitatively. The results indicated that genre-based instruction led to improvements in the participants’ speaking ability. The findings from the analysis of the three micro-genres—procedure, definition, causation—revealed statistically significant differences between the pretest and the posttest speeches in the procedure and causation micro-genres. A descriptive analysis also revealed the increases in the use of the target lexis in the micro-genres. A genre analysis of the three micro- genres illustrated how the schematic and rhetorical structure of the participants’ speech changed to meet the genre expectations of the target micro- and macro-genre. The analysis of the oral summary of a research project micro-genre demonstrated the effectiveness of genre-based instruction, as the experimental group outperformed the comparison group. This result was supported by the MFRM results, the descriptive analysis of lexis, and quantitative and qualitative genre analyses. The external validity analysis using the TOEIC Speaking Test also confirmed the effectiveness of genre-based instruction. In sum, the results provide evidence that genre-based instruction can improve Japanese high school students’ speaking ability. / Teaching & Learning