The Great Dying: A History of Pandemics

The earliest pandemic in this record struck Athens in 430 BCE, and it was described by the historian Thucydides. It killed an estimated 25 percent of the Athenian population, including the statesman Pericles. What it was remains genuinely unresolved. Typhoid fever, smallpox and viral hemorrhagic fever have all been proposed, among others. Without preserved biological material to test, the identification cannot be confirmed microbiologically, so this stays an open question in historical epidemiology rather than a settled one, and any confident retro-diagnosis you meet is reading further than the evidence goes.

The Antonine Plague, from 165 to 180 CE, is the second unresolved case, and possibly smallpox. Estimates of its toll run from about 5 million people across the Roman Empire to as many as 10 million, and sources genuinely vary on that upper bound. Whether it was a decisive factor in the Roman Empire's later decline, or one stressor among economic, military and political ones, is argued among historians: Kyle Harper has made the case for disease as a primary driver, while others weight the non-disease factors more heavily. Neither reading has closed the other out, and this article does not pretend to settle it.

One thing from Athens is not uncertain at all, and it turned out to matter more than any diagnosis. Thucydides recorded that survivors of the plague were never attacked a second time, and could nurse the sick without becoming reinfected. That is the earliest known observation of acquired immunity, set down more than two thousand years before anyone could explain the mechanism, and it is precisely the phenomenon that variolation, and much later vaccination, would put deliberately to work.

Yersinia pestis, the bacillus that causes plague, was identified by Alexandre Yersin in 1894, in Hong Kong, during an outbreak that was itself the third of its own pandemics. Four years later Paul-Louis Simond demonstrated the transmission vector: the rat flea, Xenopsylla cheopis. The cycle he described, infected rat to flea to human, is why plague has traveled with grain, with cargo and with ships, and why a port city is the natural place to catch it.

That one bacterium produced three documented pandemics, separated by centuries and by entire worlds of technology. The Plague of Justinian, from 541 to 750 CE, began in Egypt and spread along the grain trade; it is estimated to have killed 25 million to 50 million people. The Black Death, from 1347 to 1353, reached Europe on Genoese trading ships out of the Crimean port of Caffa and killed an estimated 30 to 60 percent of Europe's population, on the order of 25 million to 50 million people. The Third Pandemic, from 1855 to 1960, began in Yunnan in China and traveled the world by steamship, killing an estimated 12 million, primarily in India and China. It was that third wave, still running in living memory, that finally gave up the organism to a microscope.

The first of the three arrived at the Egyptian port of Pelusium in 541 CE and reached Constantinople by the spring of 542. The Byzantine historian Procopius recorded death rates as high as 10,000 a day at the peak, a figure most historians consider likely exaggerated, though even a skeptical reading of it leaves a genuine catastrophe behind. And it did not end there: the plague returned in roughly 18 waves over two centuries before disappearing around 750 CE.

What that did to the world is actively disputed, and both sides deserve their strongest form. The maximalist reading holds that the plague significantly weakened the Byzantine Empire and helped clear the ground for the early Islamic conquests between 632 and 750 CE. A revisionist reading, associated with the historians Lee Mordechai and Merle Eisenberg, argues that the literary sources overstate the mortality and that the archaeological evidence for demographic collapse is uneven across the empire. Both camps are working from the same thin and awkward evidence base: chroniclers with reasons to exaggerate on one side, and an archaeological record that is patchy by nature on the other. We are not crowning a winner here, because the field has not.

One element of it is not contested at all: the pathogen. Ancient DNA recovered from First Pandemic burials has confirmed Yersinia pestis directly, in finds reaching from Bavaria to a mass grave at Jerash in Jordan. Whatever historians eventually settle about consequences, the organism responsible has been identified in the ground.

In 1347 the plague came back, and over six years it killed an estimated 30 to 60 percent of the population of Europe, roughly 25 million to 50 million people. That range is wide because the evidence is parish records, manorial records and chronicles rather than censuses, and the range is not decorative: later manorial-record research, associated with Ole Benedictow, argues for the higher end against an older consensus closer to 30 percent. Anyone who hands you a single confident figure for the Black Death is rounding away a live scholarly argument.

How it got onto the ships carries a famous detail that deserves its own label. The contemporary chronicler Gabriele de' Mussi reported that the Mongol army besieging Caffa catapulted plague-infected corpses over the city walls, which would make it possibly the first recorded instance of biological warfare. The transmission event itself is well established: plague reached the Genoese ships at Caffa and traveled to Europe with them. The corpse-catapult detail rests on a single period source, and historians treat its literal accuracy as less certain than the route it is attached to. It is a real report from a real chronicler, and that is a different claim from a verified event.

The pathogen question was closed by genetics, not by argument. Bos and colleagues, publishing in Nature in 2011, reconstructed a Yersinia pestis genome at 30-fold average coverage from Black Death victims in London securely dated to 1348 to 1350, and work by Spyrou and colleagues in 2019 confirmed the same organism across multiple European sites. Whatever remains contested about the Black Death, what it was is not.

A death rate on that scale rearranged the society that survived it. The labor shortage that followed empowered surviving peasants and contributed to the end of serfdom, with the Peasants' Revolt of 1381 in England as a documented consequence. The authority of the Catholic Church was disrupted. The danse macabre entered European art, skeletal figures leading people of every social rank in one procession. And some historians argue that the whole upheaval helped create the conditions that led to the Renaissance, which is a reading of the evidence rather than a fact recorded in it.

Nor did the plague leave when the first wave burned out. Bubonic plague returned to Europe in periodic outbreaks for centuries afterward, the Great Plague of London of 1665 to 1666 among the worst of them, before the recurring pandemic pattern eventually subsided in Western Europe. For centuries, European cities lived with the standing expectation that it would come again.

Where the second pandemic actually began is a question ancient DNA has recently moved. In 2022, Spyrou and colleagues recovered two ancient Yersinia pestis genomes from fourteenth-century individuals buried at the Kara-Djigach and Burana cemeteries near Lake Issyk-Kul, in the Tian Shan region of present-day Kyrgyzstan. The tombstones at those sites are directly dated to 1338 and 1339 CE and record pestilence as the cause of death. Those genomes are the closest known ancestor to the major diversification of Yersinia pestis lineages associated with the Black Death's emergence, which supports a Central Asian origin for the second pandemic rather than a direct Crimean or Black Sea one. It is strong evidence, and it is not the last word: an ancestral strain found in one place is not by itself proof that the place is the source.

That diversification is sometimes described as a Big Bang in the bacterium's evolution. Most modern Yersinia pestis lineages descend from the fourteenth-century pandemic strain, and the diversity now catalogued, more than 30 known lineages, is thought to have been generated by that single radiation. Plague also never went away. The World Health Organization records roughly 3,000 cases a year worldwide; Madagascar, the Democratic Republic of Congo and Peru are the current hotspots; and antibiotic-resistant strains have been identified, in Madagascar in 1995. The organism behind the Black Death is a treatable disease today and an actively monitored one, which is not the same thing as a defeated one.

In 1377 the city of Ragusa, modern Dubrovnik, enacted the first documented quarantine law. The Great Council's decree of 27 July that year required ships and travelers arriving from plague-affected areas to spend 30 days in isolation before entering the city: the trentino, from the Italian trenta, thirty. The period was later extended to 40 days, quaranta giorni, and that second number is the direct origin of the word quarantine. The measure was taken with no theory of infection behind it, on the observation alone that arrivals from a stricken port were dangerous and that time seemed to make them safe.

Venice built the permanent version. In 1423 it established the first permanent lazaretto, a dedicated quarantine station on the island of Lazzaretto Vecchio, for isolating and treating plague victims, and it became the model for quarantine facilities across Mediterranean ports.

Then the Italian city states made the response institutional. They established permanent boards of health, the magistrati della sanita, the first standing public health institutions in the world, and gave them real authority: to impose quarantine, to draw a cordon sanitaire around an affected area, to enforce burial regulations. That model spread across Europe and eventually worldwide. The International Sanitary Conferences, beginning in 1851, sought to standardize quarantine practice between nations, and that long process culminated in the World Health Organization's International Health Regulations of 2005. The line from a decree in Ragusa in 1377 to the rules governing outbreak reporting today is a real institutional one, not a rhetorical flourish.

For most of that history the explanation was wrong. Miasma theory held that disease was caused by bad air rising from swamps, corpses and filth, and it was the dominant account of disease before germ theory. It is worth being precise about how wrong it was, because the shape of the error is instructive: acting on miasma theory drove genuinely beneficial sanitary reform, the cleaning and draining and clearing of filth, while simultaneously obscuring the person-to-person and waterborne routes by which many of these diseases actually moved. A false model can produce useful action for the wrong reason, which is exactly why 'it worked' has never been sufficient evidence that an explanation is true.

The most recognizable image in all of pre-modern medicine comes straight out of that theory. The plague doctor's costume, a beaked mask filled with aromatic herbs, a long wax-coated robe, and a stick for examining patients without touching them, is credited to the French physician Charles de Lorme, who devised it in 1619 during an epidemic in Paris. Physicians were hired by cities during outbreaks and were identifiable at a distance by it. The beak was not theater: it was meant to filter miasma. The costume spread through Europe, became firmly established in medical practice during the severe plague outbreaks in Rome and Naples in 1656, and was worn during the Great Plague of London in 1665 and the Great Plague of Marseille in 1720.

Not everyone accepted the bad-air account, and one dissent stands out. Ibn al-Khatib, an Andalusian physician who lived from 1313 to 1374, argued during the Black Death itself that plague spread by contagion rather than by miasma, and his work Muqana'at al-Sa'il set out empirical evidence for person-to-person transmission centuries before germ theory could explain the mechanism. He was killed for holding views considered heretical at the time. The correct answer was available, argued from evidence, and it did not win.

There was also a working technology of immunity long before there was a theory of it. Variolation, the deliberate inoculation of material from a smallpox sore to induce a mild infection and lasting immunity, was practiced in China from at least the fifteenth century, by nasal insufflation, and in the Ottoman Empire and West Africa. Lady Mary Wortley Montagu introduced the Ottoman practice to Britain in 1721. It predates Edward Jenner's cowpox-based vaccination by centuries, which means the deliberate use of Thucydides' observation is older than the science that finally explained it.

After 1492, European contact carried smallpox, measles, typhus and influenza into populations that had never encountered them and had no prior immunity. What followed is the largest demographic catastrophe in this record, and the figures attached to it are estimates with real spread rather than counts. Estimates of Indigenous American population decline range from 50 percent to more than 90 percent. The pre-contact population against which that is measured is itself debated: Denevan's 1992 estimate is roughly 54 million for the Americas as a whole, other scholars suggest a range of 40 million to 80 million, and some estimates trace a pre-contact population of 50 million to 60 million falling to roughly 5 million to 6 million by 1600. The relative role of disease against violence, displacement and forced labor is also debated, though most historians hold that introduced disease was the primary driver. The precision of any single mortality figure is uncertain. The catastrophic scale is not in dispute.

On 31 August 1854 a cholera outbreak began in the Soho district of London. In the first ten days about 500 people died. The physician John Snow plotted the addresses of the dead on a map of the neighborhood and looked at where they fell: in a tight cluster around a single water pump on Broad Street. He persuaded the local authorities to remove the pump's handle, and new cases subsided. By the end, the outbreak had killed 616 people. The pump handle is the part everyone remembers. The map is the part that mattered.

Snow's larger work, published as On the Mode of Communication of Cholera in 1855, established that cholera is transmitted by water. He did it decades before Robert Koch isolated the responsible bacterium, Vibrio cholerae, in 1883, and without any means of seeing the organism at all. What he had instead was systematic case mapping combined with hypothesis testing, and that combination is a founding method of modern epidemiology, which is why he is widely called its father. Note what this means: the method did not require the microbe. You can find a cause you cannot see, act on it, and be right, provided you are disciplined about where the deaths actually fall.

The microbes arrived anyway. Between the 1860s and the 1880s the germ theory revolution, driven by Louis Pasteur, Robert Koch and Joseph Lister, established that specific microorganisms cause specific diseases, transforming epidemic response from empirical sanitary measures into targeted identification, prevention and treatment. Koch formalized the standard of proof in 1884 with his postulates, four criteria for showing that a given microorganism causes a given disease: the organism must be found in every case of the disease; it must be isolated and grown in pure culture; the cultured organism must cause the disease when introduced into a healthy host; and it must be recoverable again from that experimentally infected host.

What the method could do at full strength has been demonstrated once, completely. In 1980 the World Health Organization declared smallpox eradicated, under the leadership of D.A. Henderson, and it remains the only human disease ever deliberately eradicated. The campaign did not work by vaccinating everybody. It worked by ring vaccination: find a case, vaccinate all its known contacts, and starve the chain of transmission. A targeted epidemiological strategy, aimed precisely, finished off one of the deadliest diseases in human history.

The 1918 influenza pandemic is the most lethal pandemic in modern history. An estimated third of the world's population, roughly 500 million people, were infected. It struck in three waves: spring 1918, the autumn of 1918 which was by far the deadliest, and the winter of 1918 to 1919. The death toll is genuinely uncertain and is best held as a range. Commonly cited estimates run from 50 million to 100 million worldwide; the CDC's own baseline figure is at least 50 million; some historians cite a lowest credible modern estimate near roughly 35 million; and a small number of studies have proposed figures as low as roughly 17 million using different methodology. Every one of those numbers is a reconstruction, and the honest way to carry them is together.

One feature of it was genuinely strange. Influenza normally kills hardest at the two ends of life, the very young and the very old, producing a U-shaped age-mortality curve. In 1918 it also killed young, healthy adults in large numbers, producing an unusual W-shaped curve instead. That extra peak, in the age group least expected to die, is the pandemic's most distinctive epidemiological signature.

Where it began is still debated. Haskell County in Kansas has been proposed, following Barry in 2004, as have Camp Funston in the same state, France and China. Genomic studies identified the pathogen as an H1N1 virus with avian characteristics, but the specific animal reservoir and the precise geographic origin remain uncertain. The name it carries is a wartime artifact rather than a clue. Spain was neutral in World War I and its press reported the outbreak freely, while the warring nations suppressed news of it to protect morale. The name therefore attached itself to a country that was reporting the disease rather than to any country where it began.

HIV and AIDS, from 1981 onward, is not a closed historical episode. Per the UNAIDS fact sheet for 2023, an estimated 40.4 million people have died of AIDS-related illness since the epidemic began, and roughly 39 million people were living with HIV in the most recent reporting year. Antiretroviral therapy has transformed HIV from a near-certain death sentence into a manageable chronic condition in countries with treatment access, and that last phrase is carrying an enormous amount of weight: global treatment access remains sharply unequal between wealthy and lower-income regions.

COVID-19, caused by SARS-CoV-2 and emerging late in 2019, is the most recent entry in this record rather than its subject, and its numbers are worth stating precisely because they are so often stated loosely. The World Health Organization's official estimate of excess mortality directly and indirectly associated with the pandemic is 14.9 million deaths for 2020 and 2021 alone, within an estimated range of 13.3 million to 16.6 million. Confirmed COVID-19 deaths reported worldwide passed 7 million by 2024. That WHO figure covers two years, not the pandemic to date, and any wider total is a reasonable estimate rather than a WHO-attributed count. mRNA vaccines were developed and authorized within roughly a year of the pandemic's start, a record pace for vaccine development, and the pandemic exposed and worsened social and economic inequalities that were already there.

Two claims recur constantly in the history of pandemics and neither survives contact with the record. The first is the oldest: that plague is divine punishment for sin. As an explanation it has no explanatory or predictive value, because disease transmission follows biological and ecological patterns rather than moral ones. As a historical force it was enormous, and it was lethal. During the Black Death the interpretation produced pogroms and massacres of Jewish communities, falsely blamed for poisoning wells, along with flagellant movements and apocalyptic preaching. The pattern of scapegoating marginalized populations during pandemics did not stop in the fourteenth century: it recurred against Chinese communities during the Third Pandemic in San Francisco in 1900, and against gay men during the early AIDS epidemic in the 1980s. In every case the blame substituted for an effective disease response instead of supporting one. This belongs in the record as something people genuinely believed and genuinely acted on. It does not belong in it as an account of why epidemics happen.

The second is recent and academic. A minority of scholars, notably Samuel Cohn in 2002 and Susan Scott and Christopher Duncan in Biology of Plagues (Cambridge University Press, 2001), argued that the Black Death was not bubonic plague at all but a viral hemorrhagic fever or some other pathogen, citing its rapid spread, its mortality pattern, and a supposed scarcity of rat remains at burial sites. It was a serious argument when it was made, and ancient DNA has decisively refuted it. Multiple independent research teams have recovered Yersinia pestis DNA directly from Black Death era mass graves across Europe, including London, Barcelona, Marseille and Bergen op Zoom. The pathogen identification is no longer in scientific dispute. We carry this one as refuted rather than as contested, which is a different thing from saying it was never worth asking.

Preparedness measured on paper is not preparedness in practice. In October 2019 the Global Health Security Index, produced by the Johns Hopkins Center for Health Security, the Nuclear Threat Initiative and the Economist Intelligence Unit, assessed 195 countries and ranked the United States and the United Kingdom as the two most pandemic-prepared countries in the world, with scores of 83.5 and 77.9. Both went on to experience among the highest per capita COVID-19 mortality rates of any country in the pandemic's first years. That is not an argument against measuring preparedness. It is an argument against mistaking the measurement for the thing.

One of the hardest problems in this history cannot be solved by epidemiology at all. Public health measures during a pandemic, lockdowns, quarantine and surveillance, run directly against civil liberties: freedom of movement, privacy, bodily autonomy. The tension is genuine, and it cannot be resolved by epidemiological science alone, because it involves a value judgment about how much individual liberty a society is willing to trade for how much reduction in collective harm. Science can tell you what a measure is likely to do. It cannot tell you what a society should be willing to pay for it, and anyone presenting that as a purely technical question is quietly answering a different one.

Then there is forgetting, which may be the most consistent pattern in the whole file. Historical memory of pandemics fades rapidly even after catastrophic mortality. The 1918 influenza killed an estimated 50 million to 100 million people and was then comparatively little discussed in mainstream public history until a scholarly and popular revival beginning in the 2000s. Whether the lessons of COVID-19 will be retained in public memory and public health policy, or fade in the same way, is an open and currently unresolved question. The one precedent we have is not encouraging.

The last item is not history but a research consensus about what comes next. The World Health Organization maintains a priority research designation called Disease X, which is an explicit acknowledgement that public health authorities consider a future pandemic caused by a currently unknown pathogen a real and likely eventual risk, while its timing, its specific pathogen and its severity cannot be predicted. The factors identified as increasing the risk of zoonotic spillover are deforestation, wildlife trade, factory farming, climate change and global travel. A related modeled estimate, from Morse and colleagues in The Lancet in 2012, suggests roughly 1.7 million currently undiscovered viruses exist in mammalian and avian host species, of which an estimated 631,000 to 827,000 could have zoonotic potential. Those are order-of-magnitude planning figures produced by a model, not an inventory of anything anyone has counted, and they should be read that way.