A sheet of medieval parchment is an animal, flattened. Before it carried a gospel or a chronicle it was the skin of a sheep, limed and scraped and stretched on a frame, and whatever was circulating in that sheep’s blood and skin at the time of slaughter went into the page with it. A team led from University College Dublin has now read some of that residue: 21 ancient genomes of sheeppox virus, recovered from manuscripts and archaeological remains, spanning roughly 1700 BCE to early modern Europe — a record indicating the disease has moved through herds for at least 3,500 years.
The work appears in Science Advances (DOI 10.1126/sciadv.aeh3571) and was covered by Jennifer Ouellette at Ars Technica, whose headline called the manuscripts biological time capsules. The paper’s own framing is roughly 3,500 years of sheeppox evolution, assembled from two very different kinds of source: bones and teeth dug out of the ground, and books that have sat on shelves being read.
The flock inside the page
Sheeppox is a capripoxvirus, highly contagious among sheep and often fatal. It was largely pushed out of European livestock over the twentieth century, though outbreaks still occur — Ars Technica notes a Greek outbreak last year. For most of recorded European history, though, it was simply part of what could happen to a flock, and the flock was also the stationery supply.
The team sampled parchment held at Yale, Cambridge, Corpus Christi College, UCD, Lambeth Palace Library, and Trinity College. Among the named volumes in the Ars account are the York Gospels, about a thousand years old; a copy of the Speculum Historiale; manuscripts from Clairvaux; the Achadeus Psalter; the Trinity Gospels; and the Lambeth Bible, along with loose gospel fragments. These are not obscure offcuts. They are working books that were bound, copied, corrected, and handled for centuries, and they turned out to be carrying viral DNA the whole time.
The pattern within individual books is the part that changes the picture. Multiple pages from the same manuscript frequently tested positive — which points toward infected animals being used for parchment rather than a single stray contaminated skin. A codex is not one sheep; a large medieval bible could consume a small herd. If several leaves from one volume carry the virus, the batch of skins that went to the parchmenter likely came from a flock that was sick, and sick flocks turning up repeatedly across the sampled manuscripts suggests outbreaks were a regular feature of medieval animal husbandry rather than rare events.
Which means the physical page has been doing double duty. The text on it records what a scribe chose to write down. The substrate records something the scribe had no way to see and no reason to mention: the health of the animals whose skins were being turned into writing surface that season.
Bronze Age teeth and a 3,500-year line
Parchment gets the story back roughly a millennium. To go further, the researchers went to archaeology — Bronze Age sheep teeth from the Eurasian steppe, from sites in Russia and Central Kazakhstan, per Ars Technica. The oldest genome in the set dates to about 1700 BCE.
That anchor point, combined with the medieval and early modern parchment genomes, produces the 21-genome span and the at-least-3,500-year figure in the paper’s reconstruction. It also lets the authors put dates on the deeper branches of the capripoxvirus family tree. Sheeppox, goatpox, and lumpy skin disease — the three lineages that afflict sheep, goats, and cattle respectively — appear to have diverged from one another somewhere between about 11,500 and 3,700 years ago, according to the authors’ estimate as reported by Ars. That window overlaps with the domestication and spread of the animals themselves, and the authors read the two histories as connected: as people gathered herds into dense, managed populations and moved them across landscapes, the viruses had a new set of conditions to adapt to.
Lead author Louis L’Hôte, a PhD candidate at UCD, and co-author Kevin G. Daly worked from the premise that the manuscript collections of Europe constitute an underused archive of animal biology. The genomes recovered support the premise fairly directly.
What the genomes say that the texts do not
The written record for sheeppox is thin and late. Ars Technica points to first-century CE Roman descriptions of pustules on sheep as the earliest known textual evidence, with medieval references to pockes and variola that are hard to pin to a specific pathogen — pox terminology in premodern writing tends to cover a wide range of skin eruptions in a wide range of species. Genetics now reaches about seventeen centuries deeper than the earliest text, and does so with an identification no vocabulary problem can blur.
One finding the team flags as unexpected: sheeppox DNA also showed up on parchment made from calfskin and goatskin. The authors offer two possibilities without settling between them. Either the virus was crossing species boundaries into animals it is not usually associated with, or the DNA arrived during manufacture — skins of different species processed in the same workshop, in the same lime pits, over the same beams, picking up traces from one another. Both explanations are plausible, and distinguishing them will take more sampling.
What remains, either way, is 21 recovered genomes and a dated span running from Bronze Age steppe pastures to early modern Europe: a herd disease documented in the material of books for three and a half millennia, on pages whose scribes were recording something else entirely.
