Counts all the different parts of speech in %[KWIC]%.