ABSTRACT
This chapter provides some guidelines for building a small specialised corpus, and discusses, with concrete examples, what can be learnt from such a corpus. As with any corpus, the most important consideration in designing a small, specialised corpus is that it should be representative. Biber defines representativeness as 'the extent to which a sample includes the full range of variability in a population'. The most important consideration regarding corpus design is that the corpus should be set up in a way that is suitable for the purpose of the research. In the case of small specialised corpora, such contextual information is extremely valuable: it is often essential for interpreting the data, and it can be drawn on in qualitatively analysing the corpus results. For both types of contextual links between corpus and context, small specialised corpora have a clear advantage over large corpora.
