Another role of corpora in semantics has been in establishing more firmly the notions of fuzzy categories and gradience. In theoretical linguistics, categories are usually seen as being hard and fast - either an item belongs to a category or it does not. However, psychological work on categorisation suggests that cognitive categories are not usually "hard and fast" but instead have fuzzy boundaries, so it is not so much a question of whether an item belongs to one category or the other, but how often it falls into one category as opposed to the other one. In looking empirically at natural language in corpora it is clear that this "fuzzy" model accounts better for the data: clear-cut boundaries do not exist; instead there are gradients of membership which are connected with frequency of inclusion.
For examples of the above read Corpus Linguistics, Chapter 4, pages 96-97.