I recently discovered this interesting Google Labs project called Google Ngram Viewier that piggy backs on the data Google is collecting via Google books. You probably already know that Google books is is not only adding new texts, but scanning (and OCRing) all the books currently available in print at massive libraries all over the world. You can imagine the interesting types of things you could do with that massive data set. When you combine all the world’s digitized books with a google algorhythm and their massive amounts of processing power, you get Ngram Viewer.
It is pretty basic right now. You put in a list of terms separated by commas, and it gives you a line graph showing how popular those words were in print from 1500 on (or any subset of years). Turning our attention towards our graph of Heaven and hell from 1800-2000, we can see that, with a few exceptions, heaven was firmly in the lead until it took a turn around 1930 from which point hell has been on the rise and not looking back.
I look forward to seeing the interesting sets of terms you come up with!