Patterns and Meanings: Using Corpora for English Language Research and Teaching (review)

Such has been the influence of a theory which relieved linguists of the need to look for data beyond their own and their friends’ intuitions that nearly 40 years after the first large corpus was made available for everyone to consult (the Brown corpus of American English, released in 1964), corpus linguists are still busying themselves explaining not just that intuitions need to be checked against ‘real’ data, but that this checking process often generates new intuitions that were completely unsuspected. Clearly, though corpus research may now have ‘been accepted into the mainstream of linguistic practice’ (148), books like this one, illustrating ‘the variety of different language areas which may be looked at through corpus analysis’ (148), would not be written if the use of corpora had become self-evident. However, the recently released second edition of the British National Corpus—the so-called ‘World Edition’ because unlike the first it is available worldwide—and the arrival of the American National Corpus, currently being developed, may truly bring about the paradigm shift some are already talking about. This book, ‘essentially a set of suggestions for future research’ (149), may not only serve as a source of inspiration to the many more that will soon turn to corpora but can also inform them about the considerable expertise already built up in Europe in this area.

Basically, what corpus linguists do is look for recurring ‘patterns’, so-called ‘collocations’, in a ‘concordance’, a list of unconnected lines of text at the center of which is the item being studied. Ch. 1 offers a brief history of thought about collocation, outlines its role in discourse building and comprehension and its importance in language acquisition, and ends with a brief overview of different types of collocation. The next seven chapters each contain a case study exemplifying how the examination of collocations can shed light on various aspects of meaning on different structural levels. Chs. 2–4 focus on words and phrases, showing how corpus analysis can be employed to differentiate between near synonyms in the same language, to identify false friends in different languages, and to reveal the connotations associated with certain items. Queries for words like if allow you to move beyond the phrase level to the sentence, and Ch. 5 demonstrates how a corpus analysis of if constructions can uncover that the rules of didactic grammars are merely a model for, rather than a constraint on, natural language production: The majority of such constructions do not conform to any of the three types of conditional sentence usually presented in student grammars. Ch. 6 moves up yet another level to the text, investigating the cohesive role of general nouns like move, labelling nouns like allegation and claim, and general verbs like happen and occur. The seventh case study looks into metaphor, more particularly as used in business and financial journalism, confirming that language is heavily metaphorical, but questioning that thought is as well since most metaphor is of the dead or dying kind, i.e. has become literal. In the last illustration, ‘unusuality’ is explored, the way ‘authors exploit the framework of habit that collocation imposes on language’ for literary and humorous effects, the psychological implication of which is that the mental lexicon contains not just discrete lexical items, but also patterns.

A final chapter very succinctly addresses some criticisms and limitations of corpus study. This is a thin book spanning a wide range of phenomena, but undoubtedly the discipline is still in need of pamphlets, and as such it is inspiring for the researcher and teacher alike.