Nawatl Corpus

Recently I read an article about argument structure by Thompson and Hopper - based on a study of a conversational corpus, they argue that 1) conversational English mostly employs intransitive predicates and 2) a vast amounts of the total number of predicates don't fit into the theory of argument structure because they can't be said to take place on a scene where one of more participants may be present. Of course they give a number of examples to back up their claims, but the strength of corpus linguistics as opposed to traditional approaches, is that 1) they only cite actually occurring speech and 2) they can tell us the exact percentage, that is the probability of usage - I find this highly appealing in a linguistic argument - its like when I try to find out if some expression that I want to use is commonly used in English by Googling the exact phrase and comparing numbers of hits - only here we're talking about spoken speech which in my view, and in that of Thompson and Hopper and many others, is closer to the language humans have stored in their brains than written language.

This has given me further inspiration to go ahead with something that I have had in the planning for a while - namely the creation of a corpus of Sierra de Zonglica Nawatl - I have a number of recordings available - both my own and those of others, and I plan to transcribe, translate and interlinearize them using Toolbox so that they will become searchable and exportable to other media. This will become a powerful tool for doing corpus linguistics AND other kinds of linguistic investigations for that matter - and I belive that it will be the first spoken corpus of a Nawatl dialect.

The only downside, of course, is that transcribing is highly time consuming so it will take a while before the corpus grows large - but ones it does, it'll be worth it!