EFL through the Digital Glass of Corpus Linguistics

Introduction

In Through the Looking-Glass, Alice suggests a new way of perceiving the world. She tells the cat: “First, there’s the room you can see through the glass—that’s just the same as our drawing-room, only the things go the other way” (Carroll, 1972, p. 181). This chapter follows the example by going through the looking-glass of language to see what can be revealed. As Sinclair (1991) stated, “the most exciting aspect of long-text data-processing, however, is not the mirroring of intuitive categories of description. It is the possibility of new approaches, new kinds of evidence, and new kinds of description” (p. 36). Later, he added that “we must gratefully adjust to this new situation [the growing evidence of language in use] and rebuild a picture of language and meaning which is not only consistent with the evidence but also exploits it to the full” (Sinclair, 2004d, p. 10).

Having said that, for a long time, the input available to language students had been restricted to what could be contained in a single book. As Frankenberg-Garcia (2005) explains:

Along the same line of thought, Johns (1991) had already stated that most grammars used to be based on intuitions rather than on authentic data. In sum, in those days, students had to rely on their course books, reference books, or teachers as the only sources of learning and solving their doubts (Aston & Burnard, 1998, p. 20).

That was not an ideal world for at least two reasons. First, no matter how complete a textbook was, it could not possibly keep up with the constant changes of a given language. This means that most course books ran the risk of becoming anachronisms when made available to the market. There are also other issues these publications did not address explicitly, such as the kind of language they used as models (McEnery & Wilson, 1996, pp. 103-104), the choice of these models, the extent to which they represented the different registers students would have to deal with, the making up or editing of texts, and the extent of their authenticity. As a result, many of these textbooks did not reflect the real-world situations students would have to face when interacting with other people in English.

Second, research in the corpus tradition can now offer solid ground to claim that speakers’ intuitions are not reliable enough in language teaching. Sinclair and Renouf (1988) illustrated this point by showing that the most frequent use of ‘see’ is not the action performed by one’s eyes, as most speakers of English as a first language (henceforth EL1) would have it, but in the expressions ‘you see’ and ‘I see’. Over a decade ago, Sinclair (1991) argued that “the problem about all kinds of introspection is that it does not give evidence about usage…Actual usage plays a very minor role in one’s consciousness of language and one would be recording largely ideas about language rather than facts of it” (p. 39). The same view was later reinforced by Hunston (2002), who stated that “corpora can give information about how a language works that may not be accessible to native speaker intuition” (p. 13).

The problem with intuition and introspection is further explained by Tsui (2004) in that “they describe what people know about language, or what they perceive language to be, rather than how language is used” (p. 39). It is true that speakers can judge whether sentences are well-formed based on grammatical rules. However, this is not enough, especially if we understand form and meaning to be inseparable (Stubbs, 1993, p. 2). When it comes to the possible combination of words, grammatical rules may not explain what occurs. Besides, they do not tell us anything about word or pattern frequency in a given register (Granger, 2002, p. 4). In a word, as Francis (1993) puts it, “the evidence of the ways in which language is really used is available in plenty, and there is no longer any need to invent example sentences in the time-honoured way” (p. 138).