Monday, September 05, 2011

Self control and Google Ngram Viewer

In the New York Times Sunday book review, Steven Pinker reviews Willpower: Rediscovering the Greatest Human Strength by Roy F. Baumeister and John Tierney, a book that was already on my to-read list after the recent summary. In doing so, Pinker writes,
"Nonetheless, the very idea of self-­control has acquired a musty Victorian odor. The Google Books Ngram Viewer shows that the phrase rose in popularity through the 19th century but began to free fall around 1920 and cratered in the 1960s, the era of doing your own thing, letting it all hang out and taking a walk on the wild side."
Being the anal fact-checking type I am, I went straight to the Google Books Ngram Viewer and searched for self-control. Nothing. Not a single hit, which is rather strange since the hyphenated version is not so uncommon. But after playing around a bit, I found that the Ngram viewer seems to have some problems with hyphens. So here's the graph of the frequency of self control sans hyphen.

From this graph, it seems Pinker is about a decade early in diagnosing free-fall, and getting on 20 years late in placing the crater. In fact, by the late sixties, self control had gained back a good deal of its losses, and Pinker doesn't mention that by 2000 we were back near historical highs. 

Maybe he's looking at a different graph. Perhaps he had more success with the hyphenated version, or he might be looking at one of the sub-corpora, say American English, or the English One Million. But none of the other graphs seem to match his description either. In fact, the British English graph tells a completely different story:

As I pointed out in my TEDx talk in June, even if we date the changes accurately, it's really not clear what fluctuations in the frequency of a particular phrase would mean. It would depend on many things including the change in popularity of synonyms (e.g., self restraint, willpower, etc.). It could indicate a shift in the frequency of the hyphenated and non-hyphenated spellings. And people can use self control both approvingly and dis-.  

Despite the trouble with interpreting changes in word frequency over time, though, I predict that there will be a rise in the frequency of this trope in the media.


Gheuf said...

The English used to spell "controul." I think it's an eighteenth century thing. What if you try that spelling?

Brett said...

Too long ago to be relevant.