When I was doing research on rickets for my masters thesis, I used a lot of primary texts from the 16th to 19th century. These were primarily medical or government texts, and were written for a number of purposes. The goal of my dissertation was to compare osteological evidence of rickets against the historical record during the early modern period in London (late 16th to early 18th centuries). From contemporary texts there is a reported rise in a new disease, and the Bills of Mortality report that rickets is increasingly a cause of death during this period. My hypothesis was that there was a much higher prevalence of rickets in this period than we were finding based on skeletal remains alone, and therefore should seek new methods of diagnosis. In order to argue this, the historic text becomes a critical line of evidence.
A new article from the International Journal of Paleopathology discusses the role of historic texts in understanding disease in the past. Mitchell (2011) argues that there are a number of hazards that new to be recognized when studying using texts. While it is often thought that descriptions of disease come primarily from medical experts, there are a variety of sources that would record outbreaks and pathologies including histories of the era, epics and tales, biographies, personal letters, diaries, legal documents, religious proclamations, in additions to hospital and medical records. Bias and information can enter the text a number of ways. Different medical beliefs change the perception of the function of the body and what is considered normal. Information may be mistakenly borrowed from earlier documents when a disease or pathology is new to an individual. Also, the text may have been written with a slant in order to create a change, such as political documents arguing a disease was prevalent for a reason that personally benefitted them rather than the actual cause. Finally, just as today, misdiagnosis and misperception was fairly common. Often common sense, word of mouth and rumor were considered to be viable evidence in diagnosis and determination of prevalence. Often the social diagnosis of a disease is vastly different than the biological reality.What further complicates understanding disease in the past through texts is that researchers are attempting to apply their own medical knowledge to give a retrospective diagnosis. Mitchell (2011:82) argues that “it is clear that retrospective diagnosis cannot be attempted for a large proportion of past disease descriptions found in written sources where the evidence is simply not specific or sufficiently detailed”. We cannot assume that the diseases present today are representative of the diseases of the past, nor can we assume that medical terminology has stayed the same. Researchers also need to determine whether medical texts are representative of the original body of texts, and of the social and medical perceptions of the time period. Often a number of different diagnoses were available, complicating modern determination of what the actual pathology was. To make matters more complicated, a lack of textual evidence does not mean that a disease wasn’t present.
In order to create more probably retrospective diagnoses, Mitchell (2011) offers a number of ways to avoid the pitfalls he discussed above. First, researchers must gather as many first hand testimonies of the disease as possible. Second, more reliable texts will have clear descriptions of symptoms, signs of the disease and pictures. When first hand accounts have recurring descriptions, it is more likely that the text is representative of the pathology. Third, the researcher needs to pay careful attention to locations and nature of the lesions and symptoms. Fourth, the cultural and social context of the author needs to be determined, as well as any factors that may bias their interpretation such as government or religious positions. Finally, in order to make a retrospective diagnosis there needs to be a clear and diagnostic connection between the modern and historic symptoms and signs.
When conducting my own research on texts from the early modern period, I ran into a number of these pitfalls that needed to be taken into consideration. A number of texts from medical experts and noted citizens in the 17th century recorded similar, but not directly copied, accounts of a disease which caused the bones in the arms and legs of children to soften. Unlike some more personal accounts, none of these attributed the disease to political, social or religious causes which may have shown a bias. An account from 1706 was suspect for stating the claim that every child was effected by the disease, a claim that no previous treatise had noted. Further inspection found that the individual was writing specifically for the king, and writing in favor of decreasing pollution which he argued would also decrease rickets. The only quantitative record of the prevalence of rickets comes from the London Bills of Mortality where rickets was recorded as a cause of death from the early 17th to mid 18th centuries. These numbers are problematic because one doesn’t die from rickets, and the individuals recording the cause of death were untrained laywomen who also recorded worms as a cause of death. By taking into account modern diagnosis and comparing this against the reliable historical texts, I was able to track the rise and fall of rickets in the early modern period. While texts are a valuable piece of information, they must be used carefully.
For more images of medical texts, check out the Vaulted Treasures Historical Collections at University of Virginia.