Buboes in Thirteenth-Century China: Evidence from Chinese Medical Writings

TL;DR

Hymes argues that Chinese medical writings show a new symptom—large purulent lumps/sores, called *geda*—added to older epidemic categories after the Mongol invasions, supporting plague in thirteenth-century North China. Li Gao’s 1247 account of the 1232 Kaifeng siege epidemic and his students’ later use of *geda* are central evidence.

Key findings

• Li Gao’s 1247 account reports that after the 1232 Mongol siege of Daliang/Kaifeng was lifted, “all but one or two out of every ten thousand people” became sick; at the capital’s twelve gates, 1,000–2,000 dead were sent out daily for almost two months, and doctors mistook the disease for Cold Damage. Hymes reads Li Gao’s treatment clues—croton seed “pressed”/rubbed to expel pus—as implying large purulent sores.
• The term *geda* (“lumps/sores”) is almost new in medical discourse: before the Jin-Yuan works discussed, it had a medical meaning just three times in surviving Tang and Song records; in Yuan times about half of all medical *geda* mentions record it as a symptom of epidemic disease, falling to less than a fifth in the Ming and rising to nearly a third in the Qing to 1800. The earliest definitely datable epidemic *geda* is Wang Haogu in 1264 (possibly 1253 in *Yilei yuanrong*), followed by Xu Guozhen 1267, Luo Tianyi 1281, Shiyuan ca. 1300, and Zhu Su 1406.
• Four preexisting epidemic categories—Thunder Head, Big Head, Seasonal Toxin, and Seasonal Epidemic—acquired purulent lumps/sores as a newly specified symptom after the Mongol advent; Hymes argues this reflects encounters with bubonic plague. Two of Li Gao’s students, Wang Haogu and Luo Tianyi, were leaders in bringing *geda* into epidemic discourse.
• Chart 2.6 shows epidemic *geda* mentions peak in 1275–1324 and again in 1625–1674, the latter higher than the Yuan peak; Hymes links this to possible late Ming/Ming-Qing plague, while cautioning about source copying and survival.

Methods

• Historical and chronological analysis of Chinese medical texts from the eleventh through eighteenth centuries, using the Airusheng *Zhongguo jiben gujiku* and Wenyuange *Siku quanshu dianziban* databases, plus some nonmedical sources such as the *Jinshi* and Yuan Haowen’s epitaphs.
• Constructs a detailed timeline/table of events, disease descriptions, and sources; traces the first addition of lumps/sores/*geda* to each preexisting epidemic category; compares *geda* distribution with the older sore term *chuang* to control for source availability.
• Not simple retrospective symptom-matching: Hymes asks whether disease descriptions change in ways suggesting an encounter with a new phenomenon, while treating Chinese medical categories historically.

Limitations

• The paper states that most sources are general catalogues or prescription guides, not dated epidemic records; dates are usually authorship/publication dates, not outbreak dates; time and place often cannot be derived; copying and echoing can inflate peaks; Shiyuan’s identity and date are uncertain; there is no perfect fit with the Qara Khitai timing; past distribution cannot be inferred directly from present strains; ancient-DNA evidence is needed.
• Hymes acknowledges Li Gao gives little physical description; *geda* descriptions are rare and incomplete; there is no direct proof of *Yersinia pestis*; the horse *geda* source is not plague and shows the term could be non-epidemic; Zhang Congzheng’s case may be a red herring.
• Evident from the text: retrospective diagnosis remains a bridging assumption, and source loss—such as Yuan Haowen’s Kaifeng siege journal not surviving past the eighteenth century—severely limits the evidence.