Lessons on Misreading Digital "Evidence" from the Past

The Digital Humanities have opened a very wide door into humanity’s medical past. They offer extensive, and often free, online access to digitized disease and medical treatment images, texts, and other materials fromcenturies, if not millennia, ago. They also offer recently digitized “big data” sets of historical disease outbreaks, patient records, and demographic trends, among many other topics. Both of these new points of entry into the past have, in turn, been a huge boon to education, research, and even public outreach in the history of health and medicine. DH has, in short, changed the way that we study health and medicine of the past by allowing more of us to enter the historical world through our computer screens. We can now readily see manuscript images of medieval doctors bleeding their patients or attending to the ubiquitous bent-over haemorrhoid sufferer, of people in the panicked throes of a plague outbreak, and of the Zodiac Man that show us in vivid colour what disease and medicine looked like in the past. We can now easily read first-hand accounts of people’s medical experiences, from both sides of the patient-practitioner divide. And we can now eagerly develop new insights into how historical epidemics appeared in one place after another through databases that can be re-sorted and re-analyzed in multiple different ways. But is it really that simple?

One of the key limitations of DH is that it has seemingly allowed the otherwise rigorous standards of evidence and interpretation normally applied by historians and scientists alike to be loosened. Online historical medical images typically circulate as cropped versions that have been removed from their original, most often textual, context. The original meanings behind these digitised, cropped, and decontextualised historical images are often lost and, worse, new meanings substituted in their place. Hence, we effortlessly find digital online images of the medieval Black Death that are, in fact, not images of the plague at all.[1] Similar problems arise with the uncritical use of historical data sets. Taking an original data set at face value overlooks potentially critical biases in source collection, selection, interpretation, recording, and even purpose. Re-sorting such databases may generate new insights, but any failure to critique the data sets’ origins and limitations renders such insights suspect.[2] By carelessly using and reusing these images and data sets in our scholarship, we too easily create a distorted view of the past and misinform the present. In other words, we run the risk of undermining our own attempts to produce solid scholarship.

The Digital Humanities do offer tremendous potential for expanding the nature, quality, and extent of our investigations into the history of health and medicine in new and different ways. But DH also comes with caveats that we have too often ignored. Being aware of these caveats, and carefully heeding the new set of challenges the DH poses to modern scholarship, will help to ensure that DH-based scholarship is not only more interesting and appealing, but also more rigorous and thus valid from historical and scientific standpoints.

[1]Lori Jones and Richard Nevell, “Plagued by doubt and viral misinformation: the need for evidence-based use of historical disease images,” The Lancet Infectious Diseases 16, no. 10 (October 2016): e235–40. http://dx.doi.org/10.1016/S1473-3099(16)30119-0. Michelle Ziegler, “Lazarus Does Not Have the Plague!” Contagions blog post. 3 January 2018. https://contagions.wordpress.com/2018/01/03/lazarus-does-not-have-the-plague/

[2]Joris Roosen and Daniel R. Curtis, "Dangers of noncritical use of historical plague data," Emerging Infectious Diseases 24, no. 1 (January 2018), online first 1 December 2017.  https://wwwnc.cdc.gov/eid/article/24/1/17-0477_article.