He followed that up with a question about the odd bump around 1900 (turns out, I gather, that it's the result of some publications being tagged with their date of founding rather than the date of publication of the actual work in which the word appears. Whew; wouldn't want to think that some careless time-traveler had gone and published a critique of Neuromancer back in 1902: or, as someone Tweeted to Gibson: "@GreatDismal Apparently you were quoted by the Manchester Literary and Philosophical Society in 1888. http://j.mp/gKZJac")
Then he started having fun. Next, he graphed out "flying saucer:"
Then he posted a few more searches - "Cthulhu," "opium," "buggery" - and announced he was done, and heading off to bed with the latest Fortean Times. But what was interesting was the graphs other people created following that, which he reposted. I was particularly interested in the "graph of fears," comparing how often the words "negro," "terrorist" and "communist" appeared:
Sure, just the idea of software that can graph out the frequency of a word across that many books is amazing enough. The idea that Google has that many books digitally stored is also pretty mind-boggling. But it's funny how we can take that for granted and go straight to playing with it, taking the psychological pulse of the last 300 years by way of killing time before bed. We're awash in this kind of massive sea of information, and we have these little toys with which we can all dabble around in it. Here you go, kids: everything Google's got in the literary world from between 1700 and now. Make pretty pictures with it, and maybe learn something.