Language Log does it again. Here's the recipe: (1) take a silly article on a silly study and (2) reproduce the experiment using the text of the article to show that the findings (top 20 words count for 33% of all words used) of Professor McEnery hold true for any text.

Learned a new term, Vicky Pollard.

Thanks for that, zmjezhd. Language Log wins at life.

So often when we hear something on BBC (or NPR). we just believe it. Pullum always takes it one step further. He is so inspiring. I just pulled up his CV, and he has 222 publications, with a few more "in press." He is a true scholar.