I’ve spent the morning playing with Google’s Ngram tool, a fascinating (and addictive) way of visualizing cultural trends. It can search through up to 500 years of published material (Google claims to have over 10% of the books ever published) and “count” the use of words. It yields some pretty interesting results (click on the images below for the Google Ngram pages):
Some would be useful for teaching the birth of certain “concepts”:
And some are just fun, like this one I saw posted on Facebook:
Admittedly, I’m always a bit wary of using computers to analyse culture. (Am I the only one who cringed at the term “culturenomics” here?) But in this case, the reasons to be cautious are not just epistemological and methodological. There are technical issues with OCR; Google Books does not always classify the publication dates of books accurately; and then there are problems with spelling, including national and regional differences, differences over time, and — of course — the ever-problematic medial “s” (case in point: you might think the eighteenth-century was a bit potty-mouthed if you do a search for the history of the term “fuck,” but if you look at some of those results you’ll find frequent references like “infants desire to fuck the blood…” etc.) Here’s a good example: a chart of the term “person” would make you think there was barely such a concept before 1800. But try adding “perfon” to the search, and you find this:
Despite these cautions about reliability, some of these results are rather fascinating. I’m doing some research into the relationship between the “technologies” of sensation and detective fiction, and if you plug these terms into Google Ngrams you get the following results:
The hype about sensation fiction that lasted throughout the 1860s and 70s is clearly represented on this graph. But although the detective novel was a popular genre throughout the second half of the nineteenth-century, as a concept it had little impact in comparison with the “sensation novel” until the twentieth century.
No Comments