The problems with Google Books are well publicised, not least quality issues, such as the appearance of sinister condom-clad fingers. It remains, however, an indispensable source for those of us who work with books. Google’s Ngram Viewer allows researchers to generate graphs based on phrase-usage during a particular period. For example, I’ve been doing some work on novels featuring the ‘New Woman’, an often derisory term used to describe late nineteenth-century feminists. By running a search in Ngram, I can see exactly when it starts appearing in print. Here I can see that it emerges in 1890, reaches its peak around 1897, then trails off very rapidly after the turn of the century. Of course, the results are based only on Google’s collection, but that does now include over 5 million books from a period of 500 years. It’s possible to run case-sensitive searches, too, so I can differentiate between a feisty “New Woman” and someone who just feels like a “new woman”. You can also plot multiple trajectories to compare terms. Here you can see the relative popularity of three famous nineteenth-century literary creations. Perhaps surprisingly, Sherlock Holmes is discussed far less than Frankenstein or Dracula. Underneath the graph are links through to a Google Books search, so you can actually see the text (if available). If you want to get really geeky, I commend to you the wildcard feature. An asterisk can represent any word, so using one along with a time period will show you the top ten most frequently occurring words. Combining it with parts of speech makes it more effective, so “*_NOUN” will find only the nouns. No mention of women here – it’s all men and temporality. There are lots more examples and precise instructions on the Ngram help pages and you can find an excellent post over on The Atlantic. And a final warning: this is highly addictive.
Thank you to Nadine Muller for alerting me to the disembodied finger.