The basic message which comes across from the three chapters in Part A is that many L2 learners have been disadvantaged because, until very recently, textbooks have been typically based on idealised data about the language they are teaching. Some have taught a prescriptive model of how their authors think the learners should use the target language, many have been based on the authors’ intuitions about how the target language is used, most have been informed by a model of the target language based on information from reference books rather than from actual data, and nearly all have taught learners to speak written grammar. None of this is too surprising, given that until very recently textbook writers had no access to comprehensive and representative data of authentic language use. They had to make use of reference books based on rules and constructed examples rather than on instances of language use. Or they based their books on their own abstract awareness of how they, as typical educated users of the language, expressed themselves in the target language. Such awareness was inevitably biased towards the norms of planned discourse (e.g. essays, lectures) as it is difficult to be aware of how we use language in unplanned discourse (e.g. spontaneous informal conversation) in which by definition we do not plan what to say and are not usually aware of exactly what we have said. So we had, for example, the ridiculous situation of writers insisting that learners use complete sentences in their conversations when the writers rarely did so themselves. Now we have no excuse. We have access to data which tells us how the target language is typically written and spoken and we know for a fact that language use is variable and depends on the context in which it is being used. We know that the grammar of the spoken language is distinctively different from that of the written language, that the degree of intimacy and of shared experience between the participants are crucial determinants of the lexis and the structures used in discourse, that all language use is subjective, attitudinal, purposeful and strategic, and that the purposes of a communication will exert a strong influence on the language which is actually used.