For decades, historians have argued over what sparked the 1789 smallpox outbreak that devastated Indigenous Australians. Now a new modeling study published July 9 in the journal Nature Human Behaviour points to a specific, and disturbing, mechanism: vials of dried smallpox scabs brought by the British “First Fleet” in 1788.

In the study, co-author Alan Williams, an archaeologist at EMM Consulting, told Live Science that the likely source were vials of dried smallpox scabs that British medics kept for inoculations, a precursor of smallpox vaccines. That matters because it rejects the competing idea that the outbreak originated through contact with Indonesian islanders fishing in the far north of Australia.

The core of the dispute is timing and geography, and the new work attacks both. Some historians had argued the live smallpox virus could not have survived the eight-month ship journey from Portsmouth, England, to Botany Bay, where the British ships first landed. The modeling study counters that with a practical detail: the vials of dried smallpox scabs may have been replenished during the voyage when the fleet stopped at either Cape Town, South Africa, or Rio de Janeiro, Brazil, only a few months earlier, Williams said. In other words, the investigators are not just asking whether smallpox could survive a crossing. They are modeling whether the most plausible route to an infection event is consistent with how and when outbreaks typically unfold.

From there, the study moves to the question of where the outbreak began. The “First Fleet” to Australia consisted of 11 ships: two naval vessels, six convict transports, and three ships for provisions. The British initially anchored at Botany Bay, then moved to Sydney Cove in Port Jackson, which is now part of modern Sydney, after they judged the land and water at Botany Bay were poor. They raised the British flag there, claiming the whole continent. Crucially for the outbreak debate, there were no signs of a smallpox outbreak among nearby Indigenous communities when the First Fleet arrived, but disease took a devastating toll in the decades that followed.

The earlier north-coast theory had a simple storyline: because there were no recorded smallpox cases on the fleet, the disease must have arrived via Australia’s tropical north coast through contact with islanders from the Makassar region of Sulawesi, now part of Indonesia. The new computer modeling, however, indicates the disease could not have spread from that region to Australia’s southeast. Williams pointed to a major issue: thousands of miles of uninhabited desert separated the proposed north coast outbreak zone from Port Jackson. Meanwhile, modeling focused on Indigenous communities near Port Jackson indicates the outbreak had originated there, the researchers found.

Co-author Matthew Cody Nitschke, an expert in computer modeling at Flinders University in Adelaide, told Live Science that it is now clear the outbreak was very unlikely to have originated in the North, and that, by contrast, there is “very strong evidence” it was the First Fleet. Williams also contextualized why the spread might have happened even without anyone intending it. At the time, “disease theory” was not known, so the spread of smallpox may have happened by accident. That matters because it reframes causality from a deliberate act into a chain of assumptions, risk, and poor visibility into contagion.

The consequences go beyond historical blame. The study estimates approximately 220,000 people died from smallpox, and it says early-20th-century Australian authorities did not comprehend the true extent of the Indigenous population when they attempted to document historical accounts. That connects to a second wave of research tied to the same team. The Nature Human Behaviour paper refers to a second study by the same researchers that is now being prepared for publication by the same journal. It proposes that there were up to 3 million Indigenous Australians at the time the First Fleet arrived, many more than the 300,000 or so proposed in some earlier studies. That later estimate, based in part on archaeological investigations of early Indigenous Australian communities, suggests the theoretical population capacity of Australia before the British arrival was up to 5 million people, especially in resource-rich areas of the North and Southeast.

Why should executives and boards care about a 1789 outbreak? Because the story is also about how complex systems get modeled, what assumptions survive scrutiny, and how the “unknown unknowns” of population data can compound over time. When the numbers were revised, historians attributed the earlier roughly 300,000 Indigenous Australians to a study in the 1930s based only on depleted numbers, Williams said. “By the time people actually sat down and said how many people were here, we had 100 years of disease, frontier violence, and everything else going on,” he added. That is a pattern organizations recognize: after repeated shocks, measurement becomes distorted, and later conclusions inherit the bias.

External historians interviewed by Live Science also framed the update as a correction of old logic. University of Tasmania historian Henry Reynolds, who was not involved in the studies, told Live Science that the research debunks the old idea that the smallpox outbreak started on the north coast, saying the idea that it came from the North and arrived at exactly the time as the First Fleet was always improbable. Ann McGrath, a historian at the Australian National University in Canberra who was not involved in the work, agreed, calling the interdisciplinary approach convincingly refuting the north-coast explanation and providing an expanded estimate of the Indigenous population prior to 1788.

For decision-makers today, the second-order takeaway is not “history repeats.” It is that modeling and evidence can overturn comfortable narratives when the underlying assumptions get stress-tested. In sectors where risk, logistics, and contagion-like spread matter, this study is a reminder that your best explanation is only as good as your model inputs, your mechanism, and your willingness to revise when the data says your original story does not add up.