I've been told (thanks to an academic contact on Twitter) about Google's Ngrams: they use their database of pretty much every book written to extract raw data. For instance, I'm wondering about writing a piece on the use of 'banditti' in literature and the media, if - as I suspect - it's used in English with an anti-Catholic subtext. With Ngrams, I get a really good chart showing me how much and when it was used (click to enlarge):
Endless fun. What a great tool. Who knew, for instance, that the word 'git' reached its peak in 1940? Annoyingly, it doesn't distinguish between 'git on up' etc and 'git' as in idiot. I'm certainly looking forward to reading The Magic Git-Flip. (Neal's favourite word, 'gitwizard', appears not to have been taken up in literary circles as yet). 'Plashing' peaked in 1860, used in a very poor poem, in Dickens' All The Year Round magazine. Interesting, my names bump along as a choice in fiction until about 1980, since when usage has increased massively. Which makes me cool. Doesn't it?