ABSTRACT
Corpus linguistics has just cracked the surface of explaining how grammatical structures work in discourse contexts. Certainly, corpus linguistics is an excellent methodology for determining the frequencies of words and grammatical structures–– – frequencies that are often broken down by genre, style, and register–– – and those findings help inform the ways in which such structures should be taught. Just a few corpus studies, though, have been able to pinpoint where grammatical structures appear in a particular stream of discourse and what their pragmatic functions are. For example, the present progressive form is often used to interpret prior utterances in spoken discourse (in addition to indicating tense). Also, a present perfect form is thought to often mark the initiation of a narrative. One can infer these patterns based on corpora of written and spoken discourse that reflect authentic language. Yet computers are still not able to adequately tag and analyze discourse-level corpora; human eyes remain necessary, which means the corpora used cannot be too large. The ultimate goal of this sub-branch of corpus linguistics – part of what is known as corpus pragmatics – is to overcome that limitation to produce results from much larger discourse-level corpora and arrive at a better accounting of entire complexes of grammar structures. Chief among these is the tense-modality system, and a firm understanding thereof will eventually greatly benefit language learners and teachers.
