Reflection 11

Investigating the viability of a collocation list

for students of English for academic purposes

By Philip Durrant

This class session was based on the discussion of collocations. We spoke about BNC (British National Corpus) which contains 100 million samples of written English. Another was the Word Smith which represents a corpus of spoken English, and still another is COCA (Corpus of Contemporary American English) with 450 million text samples. I think corpus linguistics is an impressive achievement in that it reveals patterns of frequency and co-occurrence, which are not intuitively apparent, either to use or analyst. The computer analysis of text provides detailed information about the co-textual relations of words in use: how they relate lexically in collocation, and how they relate to particular grammatical features in colligation. In short, they provide a vast amount of basic fact which could dictate what the content of language courses should be. However, I used could because I acknowledge the limitations of these descriptions.

Although corpus descriptions capture a linguistic reality that others do not, they represent partial reality, like all linguistic descriptions being necessarily limited. What they account for is one aspect of language behavior, namely the texts that language users produce, and these productions do not exhaust the possible. Moreover, they provide basic fact about text, but not about discourse. This means they do not reveal facts about the pragmatic aspect of behavior, how people use language appropriate to context to achieve a particular force and effect. We, as language teachers, need to understand what corpus findings represent. But we should also recognize that these findings are not absolute truths, and that description does not equal prescription; it just provides insights and direction for prescription.

 For academic writing, the list of such collocations is more specific, though, as stated in Durrant’s study, cross-disciplinary collocations differ in type. The main finding of this study was that the type of these collocations does not tend to be lexical, which has traditionally been focused on, but rather grammatical. Our class also captured the discussion of lexical and grammatical collocations, the former being described as the combination of two lexical words, and the latter – as that off one lexical and one grammatical word. I think, the list of grammatical collocations is as important in academic writing as lexical collocations have been valued by researchers. So, both researchers and educators should not downplay the role of either type, and if these collocations are not sufficiently frequent in the language to be learnt implicitly, at least they can be learnt explicitly.