Today we are pleased to continue with our three-part series from Tom Ewing, Sinclair Ewing-Nelson, and Veronica Kimmerly.
Part Two: “The Epidemic of Influenza” as an Editorial Intervention
On January 4, 1890, the Medical Record first noticed that the disease had reached the United States by publishing first-hand accounts by two New York City doctors of influenza cases under their care. Yet it was the editorial that appeared in the same issue that truly attracted national attention as an authoritative statement of the anticipated scope and severity of the disease outbreak (Illustration 4).
4: Medical Record Editorial, Title and First Paragraph, January 4, 1890
This short piece began with a statement that acknowledged widespread impact of this disease: “There is no doubt that influenza, or what is popularly known as ‘the grip,’ has become epidemic in this city and is scattered broadcast over the country.” The influenza epidemic was thus significant because of the large number of those affected, but this sentence also distinguishes between medical understanding and popular perceptions – as signified by the distinction between the medical term, “influenza,” and the “popular” term, “the grip.” In the rest of the article, Shrady only uses the term “influenza,” thus affirming the medical expertise that set his opinions apart from popular understanding.
The most important contribution of this editorial, however, was to downplay the perceived danger of this epidemic. The final sentence of the introductory paragraph makes this point, although in an indirect manner: “Fortunately, it is of such a mild type that, save for its invasion of large districts and for the large numbers of persons attacked, it would scarcely be noticed in the category of ailments connected with the usual ones which occur during winter months, associated with rapid and varied changes of temperature and humidity.” Positive words and phrases such as “fortunately,” “mild type,” and “usual” ailments, conveyed the message that the disease was not dangerous, which was further emphasized by statements that the “period of illness is comparatively brief,” the influenza has “comparatively low mortality” and the epidemic “is likely soon to pass over.” Even more frequently, the editorial negated indications of any real threat: reports of fatalities, particularly in Paris, were “very much exaggerated,” the dangers of associated diseases such as bronchitis and pneumonia were “greatly exaggerated,” the effects on individuals, while temporarily unpleasant, were “by no means dangerous,” and serious complications were “so few as scarcely to merit recognition alongside of those which are constantly occurring with the usual diseases of the season.”
Dr. Shrady asserted his expertise in the second paragraph by listing common symptoms such as increased temperatures, severe pain in the head, back, and extremities, and “coryzal symptoms” referring to inflammation of the mucous membranes. This section distinguished between types of influenza: “the present epidemic partakes more of the febrile and nervous variety than that of the purely catarrhal.” While “febrile” and “catarrhal” refer to physiological conditions (fevers and mucus discharge, respectively), the latter term, “nervous” had a more complex meaning, as it suggested that illness was primarily the result of mental or emotional conditions.
This editorial made recommendations for treating the influenza with antipyrin and bromides for headaches, calomel and salines to use as laxatives, and tonics such as quinine for stimulating expectorants. In making these recommendations, however, the editorial relied directly on popular experience, as certain remedies “have found most favor” or have “the confidence of the majority.” In other words, rather than legitimize treatment recommendations on chemical components or proven outcomes, the editorial drew upon observed preferences expressed either by patients or doctors.
Finally, the editorial drew a clear distinction between two categories: first, “the feeble and old,” who were “the first to succumb and the slowest to recover,” and second, those who were “robust,” had maintained their good health by avoiding fatigue, exposure, and insufficient food, and were likely to “quickly rally.” Yet even the latter population was not immune from influenza, “as no one living in an invaded district can have his safety guaranteed.” Even on this point of widespread susceptibility, however, the editorial provided a reassuring note, as the final sentence declared that those who fell ill to influenza would have increased immunity to future diseases, and thus “will not be somewhat the worse for an entirely unsought experience.”
This exploration of a single editorial, amounting to less than one page of printed copy, is an example of “close reading,” a common analytical exercise in humanities fields such as history and literary studies. In this approach, the scholar looks for meaning in the text, including the choice of words, the combination of phrases, the overall structure, and the discussion of examples. This editorial makes broad claims about the scope and severity of the disease, yet presents neither detailed numerical evidence or nor direct personal observation of patients to support the claims, relying instead on generalized reports and asserted expertise.
5: Cirrus Cloud, Medical Record, 1890, Complete Text, top words
Yet the Medical Heritage Library also allows for new methods for interpreting disease reporting in historical contexts. In the case of the journal, Medical Record, for example, the Internet Archive allows users to save the complete text of the two 1890 volumes, which amount to nearly two million total words. Using the digital humanities tool, voyant-tools.org, this entire text can be visualized in a number of ways, including the most common form of text visualization, a cirrus cloud (Illustration 5). Voyant also allows for several more insightful tools which begin to move beyond text frequency to interpretation. The context feature positions a key word in relation to the phrases that appear immediately before and after this term, thus allowing the user to situate a key term in its textual context. In the case of “influenza,” for example, this tool provides 264 phrases that have this term in context. The collocation tool in Voyant adds another layer of depth to the visualization process by identifying which words appear in proximity to each other (the user can set the parameters for the number of words within which two terms are collocated). For the purposes of this project, three terms were selected: influenza, epidemic, and pneumonia. The collocation tool in Voyant generated a list of nearly 500 collocations for these three terms, sorted by frequency. All of these terms were visualized using the Fusion Charts tool in a google spreadsheet. Illustration 6 shows the top fifty of these collocations, in a chart where the edges (links) connect nodes (three key terms and their associated words). These tools allow researchers to suggest interpretations on a larger scale than is possible through a close reading. In the case of the Medical Record for the year 1890, the collocation of influenza and epidemic suggest an awareness of the potential large-scale impact of this disease, while the collocation of influenza and pneumonia indicate how this disease was perceived as a secondary cause for the more deadly condition of pneumonia. Finally, the fact that epidemic was collocated with cholera and pox, but influenza was not, at least not with the same frequency, indicates that the Medical Record did not associate these diseases in the same category. Further investigation, including close reading of key texts, is needed to confirm these more suggestive interpretations based on large scale text analysis.
6: Mapping Collocations in Medical Record, 1890, Key Words Influenza, Epidemic, Pneumonia, Top 50 Nodes, from Fusion Charts (Click this image to view the live chart)