Perhaps in 30 years time, we shall take it for granted that software will outperform humans when it comes to translating, précising, editing or simplifying texts. If this is the case, then it will be a consequence of work in the field of corpus linguistics. Furthermore to the delight of spin-doctors, the specifics of language have been shown to affect our attitudes and thoughts. Hence influence pedlars also want to piggyback on work in corpus linguistics showing how language changes. These examples provide some reasons why we all should be aware of the field. They also show how this book could have been made more exciting.

Now, we are led to believe that there is a little friendly rivalry between those who study our own understanding of words, and those who look at the recognized meanings of our remarks. Corpus linguistics is the latter field. It ignores the psychology of thought and whether there is an association between a verbalized concept and the real world. Instead it restricts itself to objective analyses on large representative volumes of written and spoken text. These volumes are the corpora.

The book is an easy to read, if a little mundane. Meaning is discussed as something collectively agreed by a language community. Thus meaning is adaptable and variable amongst groups. As I have indicated, it would have been nice if the book was spiced up a little with a little sociology or psychology. Instead we are given the straight basics of corpus linguistics.

Building and annotating large corpora have proved a challenge in the past. Today we are more concerned with their processing. Probably most people would naively think that the basic units of meaning are words (or even parts of words). Yet most words are just too ambiguous for language analysis. This is where collocations (key words and their accompanying verbal cues) assume an importance. Much current work in this field goes into considering how one should treat these collocations. ( )