Corpus methods in syntax

This chapter concerns the use of corpora in the study of syntax, or rather, how corpus methodology is and can be used for understanding, describing, and perfecting existing theorizations of syntax. Although this perspective takes for granted the role of corpus linguistics as a methodology rather than a linguistic theory, I will try to show how the availability of large principled collections of language in the form of electronic corpora has prompted a novel view of syntax as inherently probabilistic. This view challenges a categorical modelling of language and replaces grammaticality judgments with the recognition of variability, gradience, and fuzzy phenomena and also enables observations on current change.