Learners and users – Who do we want corpus data from?

Learner corpora and lingua franca corpora differ in important ways in social and interactional aspects. Yet in the cognitive domain of language processing they have much in common, as reflected in lexicogrammatical and phraseological features. They can therefore be seen as complementary takes on second language research. We can expect advanced second/foreign language learners to show similar linguistic features to lingua franca speakers, and supporting evidence is accumulating. This paper suggests that although some features in an English as a lingua franca (ELF) corpus can be explained on cognitive principles similar to those likely to operate in learners, such as economy of effort, others cannot. For instance the common use of not quite native-like phraseological units requires a use-based rather than learning-based explanation. On the whole, the major differences between learner and ELF corpora make it necessary to keep them separate. At the same time, both can contribute results of considerable mutual interest.