Everything else in this course moved because somebody wanted it to, and the exception killed more people than all the trade ever enriched.
The previous lesson described the century when the routes were safest and most integrated, and ended with the epidemic that arrived at the western end of that system between 1346 and 1353. This lesson is about how we know what it was, where it came from, and how much of the popular story, that Silk Road trade brought the Black Death to Europe, is actually established. The answer is: more than it was ten years ago, and still not all of it.
What the disease is
Yersinia pestis is a bacterium whose natural home is not humans at all. It lives in populations of wild burrowing rodents, marmots, gerbils, ground squirrels, in which it circulates via fleas, and in which it persists indefinitely in what epidemiologists call a reservoir. Central Asia has one of the world's major reservoirs, in the marmot populations of the Tian Shan and neighbouring ranges.
Humans are an accident. When a rodent population crashes, its fleas seek new hosts, and a flea that bites a human transmits the bacterium into the lymphatic system, producing the swollen nodes, the buboes, that give bubonic plague its name. Untreated, it kills something over half of those infected. If the infection reaches the lungs it becomes pneumonic plague, which spreads directly from person to person in droplets and is almost invariably fatal without antibiotics. If it reaches the blood it becomes septicaemic and kills faster still.
The transmission mechanics matter for the history. Bubonic plague travels the way rats and fleas travel, which means in cargo: grain, cloth bales, and the holds of ships. Pneumonic plague travels the way people travel. So a disease of Central Asian marmots has two quite different routes into a European city, and both of them run along the channels this course has been describing.
Three pandemics
Plague has broken out of its reservoirs three times on a scale that reorganised societies. The first, the Justinianic plague, is dated from 541 CE, when it appears at Pelusium in Egypt, and recurred in waves for two centuries. The second is the Black Death of 1346 to 1353 and the long series of returns that followed it into the eighteenth century. The third began in Yunnan in the nineteenth century, reached Hong Kong in 1894, where Alexandre Yersin identified the bacterium that now carries his name, and went on to kill something like twelve million people, mostly in India.
The third pandemic is the one that made the other two intelligible, because it was observed by people who could culture bacteria, and it established the rodent and flea cycle that nobody had previously suspected.
Why the old evidence was not enough
Until very recently the second pandemic was known entirely from written accounts, and they are less solid than their vividness suggests.
Gabriele de' Mussi provides the famous origin scene: the Mongol army besieging the Genoese trading post of Kaffa in the Crimea in 1346, stricken with plague, catapulting its own dead over the walls, and the infection passing to the Genoese, who carried it home by ship. It is a superb story, it is the source of every account of the plague's arrival in Europe, and de' Mussi was in Piacenza at the time, writing from report.
More seriously, the identification of the disease was contested for decades. Beginning in the 1980s, several historians and zoologists, Graham Twigg and later Samuel Cohn among them, argued that the Black Death could not have been bubonic plague, on grounds that look strong from the documents: it spread far faster than modern plague does, it spread in winter and in cold northern countries where the rat flea is inactive, and no chronicler mentions the mass rat deaths that precede modern outbreaks. Their candidates ranged from anthrax to an unknown viral haemorrhagic fever.
That debate could not be settled with texts, and it was not settled by argument. It was settled by digging up the dead.
The genomes
In 2011 a team led by Kirsten Bos recovered and sequenced Yersinia pestis DNA from skeletons in the East Smithfield emergency cemetery in London, dug in 1348 and 1349 specifically for plague victims. The pathogen was plague, and the case was closed. Subsequent work confirmed Y. pestis in a sixth century cemetery at Aschheim in Bavaria, doing the same job for the Justinianic pandemic.
Then, in June 2022, a team led by Maria Spyrou published in Nature a result that fixed the origin. Two cemeteries in the Chu valley of northern Kyrgyzstan, at Kara-Djigach and Burana, excavated in the 1880s, belong to a Syriac Christian community of the kind the ninth lesson described. Their tombstones are dated and inscribed, and among them is an unusual cluster from 1338 and 1339, several of which give the cause of death as pestilence. Teeth from burials in that cluster yielded Y. pestis genomes.
The genomes did something the tombstones could not. Placed on the family tree of the bacterium, the Kara-Djigach strain sits immediately ancestral to the polytomy, the sudden fourfold branching that biologists call the big bang, from which the lineages responsible for the Black Death and for most modern plague descend. And its closest living relatives circulate today in marmots in the Tian Shan, a few hundred kilometres away.
So the second pandemic's point of departure is now located to a region and dated to within a year or two, by three independent lines, a dated inscription, a genome and a modern animal reservoir, none of which is a chronicle.
Example. The Chu valley is about 3,170 km from Kaffa, and the plague appears there in 1338 to 1339 and at Kaffa in 1346. Kaffa to Messina is about 1,810 km, covered between the siege and October 1347. Compare the two speeds and interpret.
The overland leg is km a year. The sea leg is about 1,810 km in roughly a year, four times faster, and the pattern continues: from Messina in October 1347 the disease reaches London, about 1,920 km away, within about a year, having gone by ship along the Mediterranean and Atlantic coasts before spreading inland. So the plague behaved exactly like the freight of the third lesson. It crawled overland, where it depended on caravans, local rodent populations and short hops between oases, and it accelerated the moment it reached a maritime network, where a ship carries rats, fleas and infected crew a thousand kilometres in a fortnight with no intervening stops. This is a useful corrective to the phrase "the Silk Road brought the Black Death". The overland routes brought it out of its reservoir and moved it slowly westward over years; the sea lanes of the Black Sea and the Mediterranean did the killing work, in months.
Now you. The 1338 tombstones were excavated in the 1880s and the inscriptions were published then. Why did it take until 2022 to learn what they meant?
Answer
Because the inscriptions on their own were suggestive and unprovable. Scholars had noticed the excess mortality of 1338 and 1339 and the word for pestilence, and it had been proposed as a plague outbreak more than a century ago, but "pestilence" in a medieval inscription can mean any epidemic, and a cluster of deaths in a small cemetery can have many causes. There was no way to distinguish plague from typhus, smallpox or famine fever from a gravestone, and no way to connect a local outbreak to the pandemic eight years later and three thousand kilometres away. What 2022 added was not the archaeology but the biology: a genome identifies the organism exactly, and its position on a phylogenetic tree relative to later strains establishes ancestry rather than mere coincidence in time. This is worth generalising, because it is how the subject is now advancing. The limiting factor in Silk Road history has usually been not the absence of material but the absence of a method that can make the material answer a specific question, and the material sat in museum collections for a hundred and forty years waiting for the method.
Example. The sceptics' best points were that the Black Death spread too fast for rat-borne plague, spread through northern European winters when rat fleas are inactive, and produced no reports of mass rat deaths. The genome proves it was plague. How are those observations now explained?
By accepting the observations and revising the transmission model rather than the identification, which is the right order once a direct test exists. Three mechanisms do the work. Pneumonic plague passes directly between people in droplets, needs no rodent at all, and spreads as fast as people move, which covers the winter outbreaks and much of the speed. Human ectoparasites, the human flea and the body louse, can transmit the bacterium between people, and modelling published by Katharine Dean and colleagues in 2018 found that human parasite transmission fitted the observed spread of medieval European outbreaks better than a rat-based model did. And the absence of reported rat deaths is weak evidence in any case, since medieval chroniclers were writing about human catastrophe and had no reason to record rodent mortality. What the episode shows is not that the sceptics were foolish but that their argument was the strongest available until a method arrived that could test the premise directly, and the correct response to that method was to keep their observations and drop their conclusion.
Now you. Plague returned to Europe repeatedly until the eighteenth century and then stopped, without antibiotics, vaccines or any understanding of the cause. What kinds of explanation are on the table, and how would you choose between them?
Answer
Several, and the honest position is that the question is open. Quarantine and cordon systems, developed in Italian cities from the fourteenth century and enforced with increasing severity, are a favourite because they are documented and deliberate, and the Austrian military cordon along the Ottoman frontier is often credited with keeping later outbreaks out of central Europe. Changes in housing, with brick and tile replacing thatch and wattle, may have reduced human contact with rodents. A change in the rat population, or in the bacterium itself, is possible and would be testable with ancient DNA. And the last European reservoir may simply have been extinguished, since plague needs a persistent rodent host and the western European foci may never have been self-sustaining. Choosing between them requires evidence each would leave: genomes from late outbreaks to test pathogen change, archaeological work on housing and rodent remains, and the archives of the quarantine authorities, which are excellent. This is worth including here because it is the state of a real research question rather than a settled story, and a reader who has followed this course should be comfortable with the difference.
How much died, and the case that is not proven
Mortality in the second pandemic was catastrophic by any measure and is imprecise by all of them. The usual range for Europe is between a third and a half of the population; Ole Benedictow has argued for around sixty per cent. English manorial records, which are the best series available anywhere because they record tenant deaths for fiscal reasons, indicate something like forty to forty five per cent in the first wave. For the Islamic world and North Africa the documentation is thinner and the estimates correspondingly looser. Recovery to pre-plague population levels took roughly two centuries.
Now the honest complication, because the first pandemic is a live dispute and reasoning about it well is more instructive than the settled case.
Yersinia pestis in sixth century Bavaria is proven. What is not proven is that the Justinianic plague depopulated the Mediterranean world. In 2019 Lee Mordechai, Merle Eisenberg and colleagues published an argument in PNAS that it did not, and their method is the interesting part: rather than counting chroniclers' corpses, they examined proxies that would have to move if a third of the population had died. Pollen sequences showing whether cereal cultivation contracted. Coin production and mint output. The volume of papyri and of inscriptions. Legislation, which responds to labour shortage. Across these, they found continuity rather than collapse, and argued that the pandemic's demographic impact has been assumed from Procopius rather than demonstrated.
The argument is contested and the last word has not been said. What it models is the right procedure: a pathogen's presence and a pathogen's consequences are two separate claims, requiring two separate bodies of evidence, and confirming the first tells you nothing about the second. It is exactly the discipline this course has tried to apply to trade, where the presence of a Roman glass bowl in a Chinese tomb was likewise not evidence of volume.
Example. Someone argues that because the Black Death began in Central Asia and arrived in Europe by trade routes, the Silk Road caused the Black Death. Assess the causal claim.
Break it into steps and check each. That plague emerged from a Central Asian reservoir is now well evidenced. That it reached the Crimea and then Europe along established trade and military routes is highly likely and consistent with everything known, though the eight years between the Chu valley and Kaffa are documented by nothing at all. That the routes were necessary is much weaker: reservoirs also exist elsewhere, human populations have contacted marmots for millennia, and outbreaks had presumably occurred and burned out locally many times without a pandemic. What trade supplied was not the pathogen but the connectivity, the chain of dense, connected, rat-infested settlements down which a local outbreak could propagate instead of dying out. So the defensible statement is that the routes were the mechanism of spread rather than the cause of emergence, and that the maritime network did most of the spreading. As so often in this subject, the strong version of the claim is popular precisely because it is dramatic, and the defensible version is more interesting.
Now you. What would count as evidence that plague also travelled these routes before 1338, leaving no pandemic behind?
Answer
Ancient DNA from dated burials outside the reservoir zone, which is the only class of evidence that can identify the organism, and it is already accumulating: Y. pestis has been recovered from Bronze Age skeletons across Eurasia, thousands of years before any recorded pandemic, showing that the bacterium has been moving with humans for a very long time without necessarily producing recorded catastrophes. What would strengthen the case for the medieval routes specifically is dated genomes from Central Asian and Chinese burials between the two pandemics, and the phylogeny already implies their existence, since the lineages have to have been somewhere. The methodological point is worth keeping: absence of a pandemic in the written record is weak evidence of absence of the pathogen, because most outbreaks in thinly populated regions burn out unrecorded, and the written record is thickest exactly where population is dense and a chronicler is employed.
What it changed
The consequences at the European end are outside this course's territory but not outside its argument. A third or more of the labour force died while the land, the tools and the livestock remained, which raised the value of a working pair of hands and lowered the value of an acre. Real wages in England roughly doubled over the following century, serfdom in western Europe decayed because a landlord who insisted on services found his tenants had somewhere better to go, and the relative price of labour and land shifted in a way that some economic historians trace forward for centuries.
The point to carry away is about the subject rather than about Europe. The plague is the best evidenced thing that ever moved along these routes: better dated than the silk, better localised than the Buddhism, and better quantified than the trade. It is also the one nobody was trying to move, which is a fair summary of how the largest consequences of exchange usually arrive.
The last three lessons have all pointed at the same overlooked place. Freight went four times faster by sea, the plague went four times faster by sea, and the porcelain of the tenth lesson went entirely by sea. The next lesson goes there properly, and finds that the maritime route was carrying more than the overland one for most of the period this course covers.