As I'm sure many of you have heard already, Google Labs has released a contraption that allows you to graph the frequency of words in books published in six languages (plus variations on English) since 1575. It also allows you to narrow the timeframe and compare multiple words/phrases. It's largely silly--but also good fun. Tom Scocca has demonstrated some of that fun, with amusing (though also sometimes depressing) results here, here, and here. For example, "television" surpassed "bible" around 1967, "feelings" exceeded "duty" in 1970, "fun" moved passed "honest" in 1978, and so on.
Some of the random ones I've made so far, in American English since 1800:
Of course , these data are idiosyncratic and drawing firm conclusions from them would be a premature (and require more mathematical acuity than I possess).
But messing around with it is certainly a whole lot of fun. I'm just happy they released it at the end of the semester, so I don't have to feel so guilty spending my time thinking up words to plug in.
Do any of you have favorites?