Saturday, June 04, 2011

Fun with words and history

kw: words, wordplay, history, graphics

Google has a useful and fun tool in its Ngram viewer! At the URL shown, a page opens up with a couple of sample words already loaded, and the default time limits set from years 1800 to 2000. For this first graph, I reset the first date to 1600 and put in a couple of words I know are historical and nearly out of use: thee and thou:

You'll need to click on these images to see all the detail. They are reduced to about 45% here. Those who are familiar with thee and thou know they are the archaic singular pronouns for "you". In this graph, we see primarily the references in literature that use Shakespearean language (the language of the King James Bible), up to the early or middle 1700s, and later on primarily those found in literature that quote the KJB or early versions of Shakespeare. Many modern printings of Shakespeare's works have replaced thee and thou with you, and removed numerous other archaisms.

While there is lots a person can do with the Ngram tool, I'll just introduce the possibilities with another example. First, I changed the collection from "English" to "English Fiction", set the smoothing factor from 3 to 1, and then put in the name of three major cities in which I've spent some time (eleven years, in the case of L.A.):

This is interesting. San Francisco is the most-mentioned city of the three prior to 1910, then New York City is neck-and-neck with it for thirty years; this is followed by NYC dropping out as L.A. rises, to run equal with S.F. after 1970. "So," I thought, "let's add the best-known Midwestern city, Chicago.":

What a difference! Since about 1870, Chicago has been by far the most-mentioned of these four, in English-language literature. To keep this short, I'll refrain from adding graphs for overseas cities, but as a hint for the US-centric: London is mentioned at least twice as often as Chicago, every year, and at that same level, steady, going back 250 years or more.

In a part of the web page that I didn't show, you are offered the chance to download portions of their word index to do your own larger-scale experiments. There's your chance to try this experiment with dozens of cities, or word groups of your choice.

No comments: