Google Ngram Viewer is a useful tool that unfortunately was used as a whipping boy for a rather strong attack on the digital humanities written by Adam Kirsch in The New Republic back in May 2014. The DH community has since responded in force in too many responses to link here.
So what is Google Ngram? From the website:
When you enter phrases into the Google Books Ngram Viewer, it displays a graph showing how those phrases have occurred in a corpus of books (e.g., “British English”, “English Fiction”, “French”) over the selected years.
This “corpus of books” refers to the just over five million books digitized by Google as part of its famous Google Books site up to 2012. Certainly the tech giant has a long way to go before it has digitized every book out there, and users of Google Books know that the OCR running behind it allowing users to quickly search through the collection is not 100% accurate. Still, this is a remarkably large dataset and most academics, in history at least, acknowledge that they could only look at so many sources before publishing and tenure deadlines force them to be selective in their research.
Having worked on Founders Online full-time for the last three years, I thought I would see how the six Founding Fathers on the site stack up against each other from the start or the American Revolution (1775) to the last year included in Ngram Viewer (2008). Here’s the result:
Yes, Virginia, George Washington and Thomas Jefferson are clearly the top two Founding Fathers mentioned by first and last name in Google Books, that is in the books they’ve scanned that were published during the 21st century. But let us not overlook John Adams’s impressive spike in the 1840-1880 range. I believe that this can be explained by his devoted grandson, Charles Francis Adams, who published large volumes of correspondence from his famous grandparents during this time. As the original, or perhaps most prolific, documentary editor of the nineteenth century, Adams did his best to ensure his family’s reputation was well-preserved in national memory.
While Kirsch complained that digital humanists merely used Ngram to try to confirm something they already know, the graph above, like many other visualizations in the humanities, could also be useful for non-experts. For example, the above Ngram graph could help undergraduates formulate research questions for their own class work. Is Washington really the most popular Founding Father today? Why does Hamilton appear to be the least popular? Why is there a spike in mentions for all of the founders in the 1940s and is there any relation to the American war effort in WWII? Beyond the example above, students could use Ngram to begin to evaluate the relative historical prominence of many other historical events, persons, or ideas over time. While Ngram certainly has its limitations, as long as scholars and students understand them they can use the tool successfully and as a convenient starting point for future research or class assignments.
To learn more about this Ngram Viewer and some of its more advanced features, read its about page here.
Updated on December 23, 2014.