Quantitative Corpus Linguistics with R

5Some Statistics for Corpus Linguistics

[I]t seems to me that future research should deal with frequencies in a much more empirically sound and statistical professional way. […] In the long run, linguistic theorising without a firm grip on statistical methods will lead corpus linguistics into a dead-end street.

(Mukherjee 2007:140f.)

Let me first briefly say something about the role I think statistical analysis should play in corpus linguistics. I actually believe that statistical analysis is an indispensable ingredient of corpus-linguistic work even though, unfortunately, it has not figured as prominently in corpus linguistics as one would like it to: even now, there is still much work that contents itself with reporting raw (co-) occurrence ...