One of the things that I continually find so compelling and exciting about digital humanities projects in general, is the epic sense of scale, combined with a refreshing (and, in my opinion, incredibly necessary) drive to build truly interdisciplinary coalitions. Traditional close reading and latter-day hyper-specialization will remain pivotal in the humanities for many years to come, but digital humanities projects continue to reveal the limitations of these methods and perspectives by demonstrating new ways of, quite literally, visualizing our relationship to language, and by extension, of querying the historical definition of literature itself.
One particularly ambitious and fascinating example is the Visualizing English Print, 1470-1800 project, based at the University of Wisconsin-Madison and co-investigated by Robin Valenza. As co-investigator Jonathan Hope’s (of Glasgow’s Strathclyde University) abstract puts it, “The narrative of the new humanities lies in the data: claims stand or fall on the size of the sample, the statistical significance of the results, and the care with which the procedures have been applied.” Obviously, with traditional methods of close-reading and textual analysis, sample sizes are going to be very limited, with the result that even the most well-researched comparison of literary texts and linguistic trends will be similarly limited. Such projects all-but-inevitably lead to far-reaching claims by of general literary trends and broad social movements. However, by combining digitized texts with computer modeling tools, project researchers are able to generate “something more like an epidemiology of literary populations, light sluicing across a map where the virus has passed,” generating maps and charts that are both academically and visually riveting.
For example, this representation of word popularity from 1660 to the 2000s. Each row highlights the most relatively popular words of that decade, with purple representing words popular from 1660 through the 2000s, and orange representing words that lost popularity along the way. The preview image shown here compresses the data into a high-impact photograph, but the full-size image really opens the story up. As noted here in moderate detail, one interesting example of the simultaneous epic scope, yet microscopic intensity of the findings, is seen in the prominence of the word “d.” The practice of printing a word such as “uncovered” as “uncover’d” became most prominent in the 1700s, dying out in the early 1800s. As a word, “d” reappeared in the 1990s, but in a surprising fashion; it became popular for mathematicians (scholars, textbook writers, engineers, and so on) to name mathematical points with letters of the alphabet, thereby showing a resurgence of the letter as a stand-alone word.
What’s the meaning of all this — and especially for Shakespeare? The answer to that seems deliciously vague at best, meaning there are many as-yet-un-articulated possibilities to be explored with data such as this. Using digitized texts of Shakespeare’s work to, say, determine which words appeared most frequently across the plays, or in the sonnets, is hardly new (and such research far pre-dates the age of digitized texts, to be sure). However, work such as this makes it possible to begin charting Shakespeare’s use of language within the context of all works written by his contemporaries, with all playwrights of the seventeenth and eighteenth centuries… and beyond. It becomes possible to begin not just imagining, but literally visualizing, Shakespeare’s literary and scholarly footprint in a whole new light. What is Hamlet’s reach across the span of English literature post-1600? Which plays / characters received more attention in scholarship / literature across decades/ centuries? Visual charts and maps that address questions like this will have the power to re-configure our understanding of Shakespeare’s popularity and influence throughout the centuries, and I’m looking forward to seeing what the results will be!
Other Shakespeare and digital humanities information and updates: