The Archive the Accountants Left

APlague is remembered as a rural catastrophe only by people who have not read the paperwork. The bacterium travelled with grain and cloth, and grain and cloth travelled to towns, so the documents that record its passage are municipal documents: burial fees, gate tolls, orders to whitewash a lane, payments to the men hired to carry bodies. Those papers have survived unusually well, and for an unromantic reason. A chronicle was copied because somebody judged it worth copying; a treasurer's roll was kept because it proved a debt, and a town that could not prove its debts could not borrow. Historians of epidemics therefore work in an archive assembled by accountants rather than by physicians, and that origin shapes what may sensibly be asked of it. The Black Death of 1347–1351 is the episode everyone names, but plague returned to European towns roughly every twelve years until the last western outbreak at Marseille in 1720, and it is the returns, not the first arrival, that left the thickest paper trail.

BUntil recently these records were read one town at a time. Digitisation has changed the scale but not, its critics say, the reliability. Three kinds of series are now machine-readable: fiscal accounts, which run continuously but say little about cause of death; parish burial registers, which count the dead but cannot supply the denominators that turn a body count into a rate; and quarantine correspondence, which is vivid but survives only where a magistracy kept letter-books. Between 2009 and 2022 a consortium working in 63 civic archives raised the number of transcribed plague-related entries from about 4,000 to more than 38,000. What did not move was the evidence needed to interpret them: the proportion of towns for which a defensible population figure exists for the year of an outbreak stayed near 22 per cent. Camille Ostrowski of Leiden, who has directed several of the campaigns, puts it flatly: 'we count the dead far better than we count the living'.

COstrowski's own argument rests on 214 towns with usable outbreak sequences. Sixty-one of them created a permanent health board before 1500. In that group the median interval between epidemics after 1500 was 19 years; in the remaining towns it was 11. Where no board existed, the same functions were performed, when they were performed at all, by committees dissolved as soon as the danger passed. Ostrowski reads the difference as a matter of administrative capacity rather than of medical knowledge: nobody in 1520 understood the disease any better in Milan than in a village of the Apennines, but Milan could close a gate, license a burial ground and pay a watchman for eleven months without inventing a new office each time. On this reading the retreat of plague from western Europe is a story about paperwork, and the intervals between outbreaks measure the durability of institutions more faithfully than any surviving treatise on contagion.

DTomás Rennick, of Durham, accepts the intervals and rejects the inference. In his account a health board is not a cause but a symptom: the towns that founded one were already the towns that ran a grain magistracy, and a place able to administer bread in a famine could administer isolation in an epidemic long before it wrote the statute. He also presses a harder objection. The archives that reach us are not a sample of the towns that existed but of the towns whose government never broke, and unbroken government is precisely what Ostrowski's boards are supposed to explain, so the correlation may be an artefact of what survives rather than a finding about what worked. Rennick concedes two points. His comparative material comes from a single region, the middle Rhine, and its towns are dated largely from chancery hand rather than from dated charters, a method he elsewhere calls the weakest link in the chain.

EThe exchange has narrowed on two points. Ostrowski accepts, in a paragraph that costs her something, that the interval effect is confined to towns of more than 8,000 inhabitants: among smaller places the difference disappears, and after 1600 it runs the other way. She also grants that street-widening ordinances can settle nothing between the two explanations, since towns issued them for carts as readily as for contagion. Rennick answers that the larger towns are the interesting ones precisely because they kept accounts, and that those accounts should now be searched for the wages the boards imply. Neither position has been abandoned. Since 2015 most funded digitisation has been directed at fiscal series, on the grounds that they are complete, which means the dispute will be settled, if it is settled at all, by documents that record what a town paid for and nothing about what it believed.

FOutside the seminar room the argument has a market. In 2003 a district of a southern Italian port was promoted to visitors as a quarter laid out to defeat plague, its straight streets and wide junctions presented as fourteenth-century public health. Archival work published in 2020 traced the grid to a road commission of 1758, convened after a fire. The finding fixes the date of the street plan and says nothing about the drains beneath it, which excavation still assigns to the sixteenth century, a distinction that reached the site's new signage but not most of the press coverage. City governments now commission historians for their pandemic plans as routinely as they once commissioned architects. Ostrowski regards this as the clearest gain of the field; Rennick, who has advised two such plans, warns that a method able to expose a false origin story is not thereby able to establish a true one, and that the second service is the one clients believe they are buying.