Analog Intelligence

In lieu of an abstract, here is a brief excerpt of the content:

• Analog Intelligence
• Andrew M. Stauffer (bio)

Twenty summers, with the length of twenty long winters: it’s a little epoch, the two decades that have passed since the “Whither Victorian Poetry?” issue appeared. Back then, I stressed the importance of physical books and their place within our institutions and our field, with a weather eye on the rapidly evolving digital technologies that were poised to transform our apprehension of the historical materials we study as scholars of Victorian poetry. Have they ever. As we look ahead through the thicket of social media towards the AI-generated text-and-mediascape that is just around the corner, we may find the nineteenth-century material record more urgently necessary than before. In 2023, scholars, teachers, and most of all students of Victorian literature and culture need physical books and libraries that manage collections of those artifacts of analog intelligence.

In 2003, digital technologies had made relatively narrow inroads: no Facebook, no Twitter, no YouTube, no WordPress, no Gmail. Incredibly, no smartphones. Digitization had just begun to change the ways we accessed scholarship and primary texts. Project MUSE and JSTOR had each concluded their first decade of existence and still had plenty of room to grow. In 2003, JSTOR contained only about three hundred academic journal titles across all fields. It is now ten times that size.¹ ProQuest was moving forward in its digitization of nineteenth-century materials, but it wasn’t at all clear how extensive the collections they were building were going to be. Perhaps most significantly, there was no Google Books (or HathiTrust), meaning that online coverage of nineteenth-century printed books was still very small in scale. Scholars of Victorian poetry in 2003 had to spend a lot of time in brick-and-mortar libraries, looking things up in books, hunting down allusions, locating sources, and gathering information by hand. Several years later, John Walsh could report as news the fact that “when a web user searches the internet, he or she searches the nineteenth century.”² In a now-almost-touching use of the word remote, Walsh also celebrated the fact that Victorian material could increasingly be accessed “free from the traditional confines of often remote archives and libraries,” and observed quite rightly that “scholars [End Page 553] of nineteenth-century British and American literature are awash with an ever-growing number of high-quality digital resources.” Awash and sailing forth: the early 2000s were a utopian moment for Victorian digital literary scholarship. Thanks to Jerome McGann and many others, the Rossetti Archive was up and running, and NINES was about to emerge to federate a large number of first-generation digital archives (e.g., Blake, Whitman, Dickinson) under an umbrella of peer review.

My experiences with NINES and digital nineteenth-century editions have revealed just how complex books are—that their many layers can only be partially captured through processes of abstraction and modeling. Again and again, we dreamed of ways to “put books online,” and each time we found that the books themselves put up a fight: their myriad overlapping structures resisted assimilation, leaving us with only a partial capture, a lossy rendering, despite the gains. Using computers, we could amplify or reveal or make newly operational certain aspects of books (and manuscripts et al.). But the books remained, some part of them always occluded, some aspects outside of the logic of the digitization schema. In the end, this led me back to the information-rich scene of individual copies: the irreducible uniqueness of each bibliographic object made extraordinarily and often movingly visible in readers’ annotations, inscriptions, and other marks left in their personal copies.³ My stint as a digital humanist taught me most about the powerful recalcitrance of the material printed record—a recalcitrance that we may turn to with a renewed sense of its value now.

News broke recently of pirates: namely, tech companies that have been drawing on vast libraries of copyrighted literature in their development of Large Language Model generative artificial intelligence software. As Alex Reisner put it in The Atlantic, “Pirated books are being used as inputs for computer programs that are changing how we read, learn, and communicate...