Comparing Real and Ideal Language Learner Input: The Use of an EFL Textbook Corpus in Corpus Linguistics and Language Teaching

It is widely accepted that both native speaker corpora and non-native speaker or learner corpora can be of extremely high utility in foreign language learning and teaching. But beside analysing learner output and ideal learner input, i.e. the language actually produced by competent speakers of English, it might also be helpful to look at the input pupils really get in their English lessons. In the present paper I would like to discuss how an electronic corpus of EFL textbook texts can be used to compare so-called "school English " (Mindt 1996: 232) with real English as used by native speakers, and in which way such a corpus may help us answer two crucial questions: "Do we teach our pupils authentic English, i.e. do we confront them with the same type of English they are likely to be confronted with in natural communicative situations?" and "What can we do to improve EFL teaching materials? " It will be shown that a textbook corpus (TC) can prove a useful resource if one wants to find out more about the status of authenticity in ELT. The importance of using authentic language in the EFL classroom has recently been stressed by Beaugrande (2001) and Granger who states that " … textbooks are more useful when they are based on authentic native English. " (1998: 7