Search for a command to run...
Data-driven learning (DDL), introduced in 1990 by Tim Johns, has come in multiple guises ranging from “serendipitous corpus exploration” (Boulton, 2020, p. xiv) to focus on preselected linguistic aspects. When the focus is on pre-selected linguistic aspects, teachers’ guidance may vary for a mere suggestion of what lexical/grammatical items to focus on to a detailed pre-processing of the raw concordance lines combined with a scaffolded pedagogical guidance in the exercises presented to their pupils or students. Thirty years after its inception, DDL has evolved in many ways in terms of the variety of approaches used, the types of corpora used, and the targeted end users of DDL. Examples include Corino and Onesti’s (2019) DDL use in Content and Language Integrated Learning; Friginal and Roberts, this volume, for the use of DDL in language for specific purposes and more specifically aviation English; Crosthwaite (2020) for an opening to less advanced learners; Boulton and Cobb (2017) for an evolution towards more online access. What has not evolved, however, is that DDL is still mainly conceptualised as the analysis of written concordance lines (even in case of transcribed speech) on screen (e.g. electronic handouts, connections to online concordancers) or on paper (e.g. classroom handouts, concordance lines in textbooks). Surprisingly, the numerous advances in digital technology over the last 20 years (e.g. multimodality, mobility, etc.) have largely been absent from DDL research and applications (Hirata, 2020; Coccetta, this volume, for exceptions). It is, thus, high time to reconsider oft-quoted definitions of DDL, such as that proposed by Gilquin and Granger (2010, p. 359) saying that “Data-driven learning (DDL) consists in using the tools and techniques of corpus linguistics for pedagogical purposes” or by Boulton (2011), who states, “the hands-on use of authentic corpus data (concordances) by advanced, sophisticated foreign or second language learners in higher education for inductive, self-directed language learning of advanced usage” (p. 572). Such definitions have become too narrow as they limit the boundaries of DDL to the tools and techniques of corpus linguistics, and somehow exclude an opening to other tools that have strong pedagogical potential for DDL. Another issue is that such definitions seem to target mostly advanced users. Also, the focus in such definitions seems to be on corpus linguistics first and pedagogy second, whilst I believe – in line with Kolb (2017) – that it should be ‘Learning First, Technology Second’, or even pedagogy first, technology second. Pushing the boundaries of DDL would in no way betray the key concepts behind datadriven learning as initially conceptualised by Johns (1990) – i.e. learning driven by data – but it would definitely no longer ignore a number of current affordances that were not available at the inception of DDL; it would also give pedagogy the central position it deserves. In the coming sections, the fundamentals of DDL are presented (Section 23.2) together with the reasons for revamping DDL and scaling it up to include the affordances of current digital technology and pedagogy (Section 23.3). Section 23.4 presents concrete recommendations and examples for practice, and Section 23.5 suggests future directions for research.