tv.um.es

In recent years, considerable attention has been given to metadiscourse theory and practice in the field of discourse analysis, especially in the study of academic discourse. Metadiscourse here is defined as writers’ reflexive use of language on the current text to make references to the text itself, or to its writers and readers (Hyland, 2005; Ädel, 2006). In Hyland’s (2005) interpersonal model of metadiscourse, which was adopted in our study, there is a distinction between interactive and interactional metadiscourse. By means of interactive metadiscourse, writers can organize a text in a logical and cohesive way and guide the imagined readers through the text; while interactional metadiscourse allows writers to express their points of view, and finally convince readers of their ideas or arguments.

Despite the increasing popularity of this topic, there exist some inadequacies in the growing body of metadiscourse studies. Firstly, most of the prior research has focused on interactional metadiscourse whereas interactive metadiscourse gained less scholarly attention in general (Cao & Hu, 2014; Khedri et al., 2013). Additionally, previous studies seem to show favor to cross-disciplinary comparison of metadiscourse while few studies provide insights into cross-cultural and expertise group comparisons of metadiscourse. Last but not least, there is a dearth of scholarship on metadiscourse of a foreign language other than English. Therefore, this study aims to plug the gap by examining the use of interactive metadiscourse by different writer groups in the context of Spanish academic writing.

To this end, three comparable corpora were built: the first comprised 36 research articles published by native Spanish scholars (around 240,000 words); the second consisted of 22 master’s theses written by Spanish-speaking students (around 420,000 words); and the last one also had 22 master’s theses but produced by Chinese learners of Spanish (around 460,000 words). By drawing on the work by Cao & Hu (2014), Mur-Dueñas (2011), Lee & Casal (2014), and Carrió-Pastor (2016), a fine-grained taxonomy of interactive metadiscourse was used as our analytic framework, and a more comprehensive list of interactive metadiscourse items was created for later analysis. We used MAXQDA 2020 (VERBI Software, 2019) for data management and coding. Both quantitative and qualitative data was analyzed in this study.

Based on the results from a series of one-way ANOVA with post-hoc tests, we found that i) there was a significant difference in the broad category Transitions among the three corpus groups while the other four main categories did not show any significance; ii) for subcategories, significant differences were found in Addition, Consequence, Announcers, and Previews, whereas no strong evidence supported a significantly different use in the remaining metadiscourse subcategories. In addition, we compared the use of individual metadiscourse markers across the three writer groups in order to identify items peculiar to a specific group. The result indeed showed that some items were preferred by one group but not the others, such as por eso, asimismo, así que, en el capítulo. Possible factors that could explain this are the register awareness and expertise of writers, and genre difference. Finally, the study ends up highlighting some discussions and implications and suggesting directions for future work.