Centuries after their creation, works of medieval literature, including tales of King Arthur’s Camelot, Icelandic sagas filled with accounts of Viking victories and more, continue to fascinate and inspire. In a pop-culture world simultaneously obsessed with familiar characters and constant reinvention, the longevity of these stories speaks to their resonance. A new study published in the journal Science, however, suggests that these beloved legends represent just two-thirds of the total fiction produced during the medieval era. As Michael Price reports for Science magazine, a team using techniques more commonly used to track wildlife estimates that 68 percent of chivalric and heroic works produced in medieval Europe survive today. For individual manuscripts, or handwritten copies of literary works, that figure drops to 9 percent.
To conduct the analysis, the researchers selected 3,648 manuscripts written in Dutch, French, Icelandic, Irish, English and German between 600 and 1450 C.E. They plugged the data into the “unseen species” model, which was developed by co-author Anne Chao, a statistician at National Tsing Hua University, to survey biological diversity in a given area. Based on the number of species recorded in an area, the model estimates how many other “unseen” ones make up the local population. (More generally, explains Bruce Bower for Science News, the technique can be used across disciplines to gauge how many “unobserved events of any type … accompany relatively frequent observed events of the same type.”)
“These models use the pattern of the observed evidence to estimate what we’re not seeing,” study co-author Daniel Sawyer, a medieval English literature expert at the University of Oxford, tells New Scientist’s Chris Stokel-Walker.
Under the statistical model, the team’s analysis theorized that the 3,648 manuscripts represented a larger universe of 40,614 medieval manuscripts, around 90 percent of which no longer survive. Crucially, writes Sophie Bushwick for Scientific American, these figures refer to “the loss of physical documents, not the stories preserved in them.” For a work of medieval literature to be considered completely lost, all existing copies of it must be destroyed—much like how animals are only declared extinct after all living members of their species die. Looking at literary works rather than individual copies, the researchers found that 68 percent (799 out of around 1,170 stories) likely survived to the present.
“It’s basically a distribution of manuscripts over works,” co-author Katarzyna Anna Kapitan, an Old Norse philologist at the University of Oxford, tells Jennifer Ouellette of Ars Technica. “If you have a literary tradition that is even, it means that all literary works have more or less equal numbers of copies of manuscripts that preserve it. If it’s uneven, then you will have work that is preserved in a lot of manuscripts and then some works that are preserved in just one or two,” making them less likely to survive.
Icelandic and Irish literature boasted relatively high survival rates for both individual manuscripts (17 percent and 19 percent, respectively) and overall works (77 percent and 81 percent, respectively), as well as similar evenness profiles—perhaps because they were produced on relatively isolated islands.
“This is something that is very interesting for our study because it seems like island ecosystems ... are better in preserving their biological biodiversity, right?” Kapitan says to Scientific American. “It’s an interesting hypothesis to entertain whether the same patterns could facilitate better survival of cultural heritage in those island societies.”
English literature, meanwhile, had a low survival rate, with 7 percent of medieval manuscripts and 38 percent of works surviving to the present. Though England is also an island, it had stronger ties to the European continent during the medieval period than Iceland and Ireland. At the time, adds Sawyer in the statement, English “had little international prestige, and was rarely learned beyond England.”
Kathleen Kennedy, a medieval English literature expert at the University of Bristol who was not involved in the study, tells New Scientist that the study’s findings “generally support existing scholarly assessment.”
“Applying statistical models from ecology offers a tantalizing workaround,” she says, “[but] in the end, we cannot ever prove or disprove either traditional or statistical estimates of lost literary works, or the manuscripts containing them.”
Exactly what happened to the thousands of lost medieval manuscripts remains unclear. But the researchers offer a range of possible explanations, including library fires, recycling of older manuscripts, insect damage, and normal wear and tear. In one instance, a 13th-century collection of Old Norse tales was used to reinforce a bishop’s miter, or headdress. The team’s next project will be diving into the question of why medieval literature disappeared by examining whether certain genres or languages were more likely to survive.
“I do think this research touches on the preservation of culture,” co-author Mike Kestemont, a literary scholar at the University of Antwerp, tells Ars Technica. “Now we have estimates of the size of the loss. What we don’t have is a statistical explanation of why books were lost. We don’t know what drives cultural loss. We’re all thinking very hard now about how we can build a statistical model that also explains why certain things survived and others did not.”