This week featured an in-depth look at using corpora to inform discourse analysis, in this case analyses of British newspaper discourse regarding Muslims, immigration, and other related terms/ideas. The lecture and readings were full of interesting points and information. It’s a good demonstration of how keywords, collocations, word sketches and such can illuminate otherwise hidden and hard-to-see features in texts.
Some basic info about corpus building was also introduced, and I assume that corpus building will become a bigger topic in the coming weeks.
As for more technical matters, the Antconc tutorial continued with how to search for clusters of words and ngrams which are, IMO, extremely useful as a concept (at least) for helping learners to produce natural sounding expressions. Using tagged corpora in Antconc was introduced as well. I’m assuming, also, that tagging a corpus will be a topic in the next week or two.
I think the best things about this week were the provision of annotated LOB and Brown corpora and the free (limited duration) account for the Sketch Engine. I am particularly excited to use the Sketch Engine as I have only used it under a trial account before. Particularly for those already familiar or semi-familiar with corpus linguistics, the provision of these kinds of resources really increases the practical value of the course.