If you could trace, plot, graph, and map how historical research in and beyond North America has changed over a century, what would you like to know?
In this project, I draw on tens of thousands of book reviews from the flagship journal of historical research in North America: the American Historical Review. I currently am analyzing 10,000 born-digital book reviews from the AHR for the past decade.
To this collection of recent book reviews, I also want to add the 112 years and approximately 50 to 100,000 AHR book reviews that preceded the born-digital reviews. These earlier reviews are available on JSTOR. I currently am beginning work with JSTOR's "Data for Research" text-mining service. With this service, I can download large numbers of these book reviews at once with one caveat - they have been disaggregated into frequency lists of 'n-grams,' which list individual words, two-word pairs (bigrams), and three-word groups (trigrams) by frequency. This form of dissemination is meant to prevent piracy. Unfortunately, it also makes some forms of digital text analysis impossible (i.e. identifying which adjectives are commonly found near particular peoples, places, or events). Nonetheless, the n-grams and frequency lists still allow for some basic analysis of trends in historical scholarship.
Questions labeled with a 'd' can only be answered with the born-digital reviews and with 'd?' are better answered by these reviews. I will be able to use the frequency lists for older book reviews to answer the remaining questions, although perhaps with less precision.