New Google database search tool

[Epictetus]

Google has developed an interesting search tool for all the books it has scanned. Here is a Wall Street Journal article about it and here is the actual tool:
http://ngrams.googlelabs.com

Now a person can find the frequency of words in published texts. This could conceivably be used for all sorts of things, like linguistic, censorship or political studies. I've been playing with it to see how the popularity of Roman personalities have changed during time. For instance, Julius Caesar's relative popularity seems to have peaked circa 1810 but has been rather stable. Poor Marcus Aurelius seems to have hit his high-water mark about 1905 and has been sliding, more or less, ever since.

You can also compare terms, such as the relative popularity of "Eastern Roman Empire" versus "Byzantine Empire."

This could probably be used for all sorts of historical studies.

[Sean Manning]

That is a marvelous toy. Beware that it often gives false positives for early use of words when a 1999 reprint of a book published in 1923 gets filed under 1923, both the body and the introduction, notes, etc. Most of the pre-1970 uses of "internet" seem to be like that. At first the computer scientist in me was indignant (clearly, a book is most naturally modeled as an ordered collection of texts, each with properties such as an author or date) but then someone would have to go through each book marking it up, and I'd guess that would make the process too expensive.

Two of my favorite searches are the impact of changing modes of transportation on language and evidence for the date at which various sf/fa authors' careers peaked..

[D B Campbell]

Two of my favorite searches are ... and evidence for the date at which various sf/fa authors' careers peaked..

I guess they forgot to scan the R.E.H. data in?!

[Sean Manning]

Looking at their information, punctuation seems to be treated specially. So names with periods, like "Robert E. Howard" or "G. Tiberius Gracchus" sometimes fail to show up. I should read the Science paper but first I have some languages to practice.

Meanwhile Isaac Asimov looks more famous than any of the other authors, probably by virtue of writing more books than all of them put together!

[Robert Vermaat]

You can also compare terms, such as the relative popularity of "Eastern Roman Empire" versus "Byzantine Empire."

Or the use of the words reenactment vs. living history bewteen 1800 and 2008...