Revealing Data: Close Reading and Textual Analysis as Historical Methods

Historical medical journals provide unique perspectives on the development of expert understanding of transmission, morbidity, and impact during an epidemic. Examining the ways that medical journals contributed to the spread of information, evaluation of interpretations, and creation of new knowledge in a specific historical process can contribute to current discussions about the relationship between expert perspectives and public understanding. Information on the Russian Influenza (1889–1890) in the British Medical Journal offers an excellent case study for evaluating historical significance and contemporary relevance. Circulating Now welcomes guest bloggers E. Thomas Ewing, Ian Hargreaves, Jessica King, Andrew Pregnall, and Tyler Talnagi who examine different dimensions of the role of medical journals in collecting, presenting, and interpreting knowledge of a disease outbreak in this last of three posts for our ongoing Revealing Data series.

During the 1889–1890 influenza epidemic, the question of how influenza spread became the focus of public discussion among medical experts and can now be explored through the articles published by medical journals. This post combines two techniques, close analytical reading and large scale text interpretation, to explore how contemporaries understood the etiology of influenza and how their understanding shaped expert and public perceptions of the disease.

The February 8, 1890 issue of the British Medical Journal published a letter under the headline, “The Etiology of Influenza,” responding to a study of epidemic influenza at a psychiatric hospital in Edinburgh published in the previous issue, dated February 1, 1890. The original study and the letter were both written by physicians, who drew upon their own experience treating patients and documenting the course of the disease to explain the spread and impact of the influenza. The studies, one from the Royal Asylum in Edinburgh by Drs. George M. Robertson and Frank A. Elkins and one from Dr. William Bruce on his observation of his own patients in Dingwall, were similar in ways of collecting, classifying, and analyzing information about the influenza epidemic.

The question of causation became a point of disagreement. Robertson and Elkins cited examples of seemingly random case outbreaks among patients and hospital officials to support this guarded interpretation: “We can express no very definite opinion as regards the contagiousness of the disease; if however contagiousness exists, it only does so to a very slight degree.” By contrast, Dr. Bruce directly challenged this statement as an example of the “anti-contagionist” argument by asserting that uneven patterns of illness, simultaneous outbreaks in different locations, and greatest susceptibility of those spending time outdoors are all consistent with an infectious disease. While praising the Edinburgh study as “a splendid storehouse of facts, and facts are what are still needed to complete the life history of this peculiar epidemic,” Bruce nevertheless argued that it did not “throw sufficient light” on the question of “incubation,” and he concluded with a broad statement of the need for more research into this topic: “It is a pity that we have had so much speculation with so few attempts at really detecting the germ which, according to the analogy of similar diseases, must be present. I am no bacteriologist, but surely something might be done by examining the sputum or filtering the expired air of infected patients. I may be excused adding to the numerous hypotheses, and saying that I believe it will be found to be closely allied to the germ peculiar to contagious pneumonia, where in fresh sputum I have seen the bacillus active and moving.”

The “etiology of influenza,” which Bruce connected to the broader argument for and against “contagion,” is a topic well-suited to large-scale textual analysis of the British Medical Journal, thus demonstrating the value of historical scholarship that combines traditional close-reading techniques with the possibilities of examining extensive quantities of digitized text using analytical tools. Large-scale text analysis provides an alternative method to explore the themes identified by a close reading of these two articles.

A frequency chart, Figure 1, shows which medical terms were collocated with infect* and contag* (terms were truncated to recognize variations, such as infection, infectious, infected, infectivity, and infective) in the corpus of the British Medical Journal for one year before and after the Russian influenza, from July 1889 to June 1890. Not surprisingly, common medical terms appear most frequently. This list is suggestive, however, because it identifies a glossary of of specific diseases, in addition to influenza, which appear in proximity to these two terms, including diphtheria, leprosy, measles, pneumonia, as well as tuberculosis and phthisis referring to the same disease), pox (presumably small-pox), scarlet (presumably scarlet fever).

Figure 2 shows how disease terms were collocated with infect* and contag*, as calculated as a percentage of medical terms. with some interesting patterns: influenza and phthisis/tuberculosis appeared in close proximity to both infect* and contag*, whereas small-pox and measles only appeared close to infect* while pneumonia and leprosy were only associated with contag*.

A network diagram provides further examples of the collocation of terms, as shown in Figure 3. As suggested above, this network diagram shows a clustering of terms connected to all three words (influenza, contag*, and infect*) that were general medical terms (epidemic, patient, hospital, medical, case, and disease). Some of the more interesting terms were located on the periphery, with connections to only one of the terms. Authorities, pox, and isolation were connected only to infect*, while death and attack were connected only to influenza. As with the above chart, this kind of network visualization suggests possible connections that require further analysis in order to make definitive statements about expert understanding of the infectious or contagious nature of influenza.

These suggestions can be explored with a hybrid approach that uses large-scale text analysis to identify a smaller set of relevant phrases to be analyzed using a close reading technique. The Voyant tool makes it possible to list more than 250 phrases from one year of British Medical Journal, July 1889 to June 1890, that position the word influenza in context, with fifteen words on either side. Approximately 10% of these phrases also contain either infect* or contag*, thus creating the subset of phrases shown in Table 1 above. An analysis of more than twenty phrases produces the following distribution of answers to the question: during the Russian influenza epidemic, did medical experts consider influenza to be an infectious and/or contagious disease? One-third (7) were interpreted as indicating an affirmative answer, just over one-third indicated that the answer was complicated or contested (8), and a similar share that cannot be classified based on this selection of text. None of these phrases, however, can be classified as denying that influenza is infectious or contagious. Given the complexities evident in the first-hand commentary of the physicians published in the British Medical Journal, the interpretations suggested by large scale text analysis confirm that medical experts were moving toward a consensus belief that influenza was contagious or infectious–but such an understanding was complicated and contested.

We hear about data every day. In historical medical collections, data abounds, both quantitative and qualitative. In its format, scope, and biases, data inherently contains more information than its face value. This series, Revealing Data, explores how, by preserving the research data of the past and making it publicly available, the National Library of Medicine (NLM) helps to ensure that generations of researchers can reexamine it, reveal new stories, and make new discoveries. As the NLM becomes the new home of data science at the National Institutes of Health (NIH), Circulating Now explores what researchers from a variety of disciplines are learning from centuries of preserved data, and how their work can help us think about the future preservation and uses of the data we collect today.

Thomas Ewing is a professor of history at Virginia Tech and director of the Tracking the Russian flu project. Funding for this project, including support for undergraduate student researchers, was provided by the National Endowment for the Humanities and a 4VA research grant.

Ian Hargreaves graduated from Virginia Tech with degrees in Foreign Languages (German) and International Studies, and he is enrolled at the The George Washington University Law School.

Jessica King graduated from Virginia Tech with majors in Communication and International Studies with a minor in German, and she is enrolled in the Virginia Tech graduate program in Communication.

Andrew Pregnall is a senior at Virginia Tech pursuing degrees in Microbiology and History.

Tyler Talnagi graduated from Virginia Tech with degrees in German and International Studies.