A Small Scottish Grant Is Reopening the Mystery of Where the Book of Kells Was Made
In a field where arguments can hinge on a pinprick, a modest grant is poised to test a big claim about one of the most celebrated objects of early medieval art. The Society of Antiquaries of Scotland has awarded £2,779 ($3,700) to master craftsman Thomas Keyes to reconstruct an eighth-century washing tank at the monastery site of Portmahomack, on Scotland’s northeastern coast, and to make vellum using the same methods the monks may have used 1,200 years ago.
The goal is not simply to revive a lost craft. It is to generate new, practical evidence in a debate that has occupied scholars for roughly two centuries: where the Book of Kells was produced.
The illuminated manuscript, a lavishly ornamented presentation of the four Gospels, is named for the Irish abbey that protected it, not for the place it was made. Most researchers agree it arrived at the Abbey of Kells at the beginning of the ninth century, but its origin remains uncertain. The book’s material demands alone are staggering: it required nearly 200 calfskins and is estimated to have taken 75 years to complete.
For much of modern scholarship, the leading candidate has been St. Columba’s monastery on Iona, an island off Scotland’s west coast. The prevailing narrative holds that monks fled Iona at the end of the eighth century, carrying manuscripts and relics to Kells as Viking raids intensified. Yet despite Iona’s central place in the story, the site has not produced the physical archaeological evidence of a workshop capable of manufacturing vellum at the scale and sophistication the Book of Kells would have required.
Portmahomack has increasingly filled that evidentiary gap. Excavations conducted there between 1994 and 2007 uncovered what researchers described as a highly developed vellum-making operation. Among the finds were bone pegs used to stretch hides, pumice stones for scraping, fire pits that could have produced soda ash for dehairing, and a large stone tank where skins were soaked. The discovery was notable not only for its completeness but also for its rarity: Portmahomack remains the only early medieval vellum workshop identified in Northern Europe.
Keyes’s new project is designed to move from archaeological inference to experimental proof. He will build a replica of the washing tank and undertake the unglamorous, labor-intensive sequence that turns animal skin into a writing surface: digging and engineering the water flow into the tank, then scrubbing, scraping, and stretching hides with historically accurate tools.
One variable, however, could be decisive. Lye was used to clean and prepare skins, and at Portmahomack the usual ingredient, lime, is not available locally. One theory is that monks produced an alkaline solution from seaweed instead. Keyes has already tested seaweed lye; the next challenge, he has said, is controlling temperature shifts in the water to a degree that would match the skill of early medieval makers.
The experiment also intersects with a telling detail on the manuscript itself. The Book of Kells, held by Trinity College Dublin since the 17th century, shows small pock marks attributed to bacteria that ate into the hide during soaking. That kind of damage, Keyes has noted, does not occur when lye is made from lime, the method associated with monasteries in Ireland, Iona, and northern England.
“The theory that the Book of Kells was made at Portmahomack is already well supported by the circumstantial evidence,” Keyes said in an email. “For me, it’s now about getting into the granular detail and working out firstly specifically how each step of the process was carried out and secondly how each process could be used as evidence, either for or against the theory.”
Beyond the question of geography, the work may also sharpen estimates of time and labor. By tracking the pace and scale of production under plausible eighth-century conditions, Keyes’s reconstruction could offer a more grounded sense of how long a project like the Book of Kells might have taken in practice.
Results from the experiment will be published online by the Tarbat Discovery Centre and presented in a public lecture in late 2026, adding a new layer of material evidence to a debate that has long relied on fragments, probabilities, and the stubborn silence of the historical record.
