The learner corpus as a pedagogic corpus

While the reference corpora of native-speaker language use such as the Brown Corpus or the LOB Corpus, which were created from the mid twentieth century onwards, were intended as resources for linguistic research, teachers in higher education soon realised that attested examples of language use, both written and spoken by native speakers, also had potential in the context of language learning and teaching. Consulting a corpus and producing concordances could supplement or replace the invented examples which had hitherto been the norm in language-learning coursebooks, and could thus create a much richer learning environment by providing multiple genuine examples. From the 1980s onwards publications on this new development in language teaching began to appear. Tim Johns is an early example of a teacher whose research publications emerged directly from his experimentation with the use of corpus data with his students (Johns 1986), hereafter referred to as a teacher-researcher. It soon became evident that these teacher-researchers did not always rely on the existing reference corpora, but tended to create smaller corpora to meet their learners' needs (Stevens 1991; Cobb 1997). Terms used for these corpora include quick-and-dirty, specialist micro-corpus (Tribble 1997), home-made (Gavioli 2005: 53), learner-centred (Rodgers et al. 2011) and pedagogic (Willis 1998: 46; Hunston 2002: 16; Braun 2005; Meunier and Gouverneur 2009: 286). Considerable experimentation was undertaken in the use of corpus data with learners, and a substantial body of research articles was published. Boulton has identified more than 116 relevant publications, and has published overviews of different aspects of teachers’ use of corpus data with learners (Boulton 2010, 2012; Boulton and Tyne 2014). A notable characteristic of the vast majority of publications on corpora and language learning is that they focus on native speaker (NS) language use; Gilquin and Granger (2010) are a notable exception in that their chapter on the use of corpora in language learning and teaching includes both NS and non-native (NNS) corpora. The use of learner data by learners themselves has not been the subject of empirical studies until relatively recently (see, for example, Lee and Swales 2006; Belz and Vyatkina 2008; Wen-Ming and Hsien-Chin 2008; Bloch 2009, 2010).