Every human being alive today shares a single common ancestor who lived as recently as 3,000 years ago — not a mythical figure, but an ordinary person who happened to be in the right place at the right time — meaning every emperor, every slave, every nameless laborer of the ancient world is either an ancestor to everyone now living, or to no one, and we will never know which

The most recent person from whom every human being now alive is descended lived approximately 3,000 years ago. The figure sounds wrong. It is not. It is the result of one of the cleaner findings in mathematical genealogy, published in Nature in 2004 by Douglas Rohde of MIT, the science writer Steve Olson, and Joseph Chang of Yale, and it has been repeatedly verified by more detailed subsequent modelling. The most recent common ancestor of every person reading these words, and every other person on Earth in 2026, was an ordinary individual — neither famous nor influential in any sense the historical record would have noticed — who happened to live in a location and at a time from which descendants could eventually reach every modern population through enough lines of inheritance. The estimate places this person somewhere between 1400 BCE and 300 BCE.

The argument behind the number is simple to state and counterintuitive in its consequences. Every person has two parents, four grandparents, eight great-grandparents, and so on, doubling with each generation back. After 33 generations, or roughly a thousand years assuming 30 years per generation, the number of “ancestor slots” in any individual’s family tree is more than 8 billion — already more than the current world population, let alone the much smaller population that existed a thousand years ago. The slots cannot all be filled by distinct people. They must collapse, with the same ancestors appearing in many different positions in the same person’s tree. When you extend the same logic to all 8 billion people alive today, the ancestor slots from generation to generation force shared ancestors with mathematical inevitability.

What the Rohde, Olson and Chang paper actually proved

According to the Yale press release accompanying the 2004 Nature paper, Chang had first explored the question in 1999, in a paper titled “Recent common ancestors of all present-day individuals” published in Advances in Applied Probability. The 1999 paper used a simplified mathematical model that ignored geography, migration patterns, and the social structures that determine who actually has children with whom, and placed the most recent common ancestor (MRCA) of all humans less than a thousand years ago. The objection was obvious: real human populations are not randomly mating clouds of mathematical points. They are clustered by geography, by language, by culture, by social structure, by religion, and by physical barriers. The simplified model gave the right shape of answer but the wrong number.

The 2004 paper added geography and migration. According to ScienceDaily’s coverage of the paper, Rohde, Olson and Chang built a computer simulation of the global human population over the past several thousand years, modelling the inhabited world as a network of populations separated by realistic barriers (oceans, mountain ranges, deserts) and connected by realistic levels of migration (occasional trade routes, occasional colonisation, occasional crossing of ocean gaps). They incorporated the historical record of major migration events — the settlement of the Americas, the Austronesian expansion across the Pacific, the colonisation of remote islands. The resulting estimate for the MRCA was 2,300 to 3,400 years ago, with a central value around 3,000 years ago. The increase from the simplified model’s <1,000 years was substantial, but not as substantial as intuition might suggest. Three thousand years is still a very recent date.

The reason it is not 30,000 or 300,000 years is that even small amounts of migration are sufficient to connect the human family tree given enough generations. According to NBC News’s account of how Chang and Olson came to collaborate, Olson had been working on a book about human relatedness when he contacted Chang to help him refine the simplified model. The two reasoned that a single fertile traveller from one population to another, every century or two, would be enough to ensure that the populations eventually share ancestors. Over thousands of years, even the most isolated communities — remote Polynesian islands, the indigenous peoples of the Andaman Islands — receive enough genetic input from outside to be connected to the global pedigree.

The identical ancestors point

The 2004 paper introduced a second, even more striking concept, called the “identical ancestors point” (IAP). The MRCA is the most recent individual from whom everyone alive today is descended. The IAP is something different and slightly more recent in scope: it is the most recent point in the past at which everyone alive then either is an ancestor of every person alive now, or is an ancestor of no person alive now. Below the MRCA, there are still many people in the past who are ancestors of some living humans but not others. Above the IAP — that is, further back in time — there is no such middle category. Every individual who lived then and who left any descendants at all left descendants who, by chains of marriage and reproduction, eventually populate every corner of the modern world.

According to the Wikipedia reference on the Identical Ancestors Point, the Rohde-Olson-Chang model places the human IAP somewhere between 5,000 and 7,400 years ago, in the period from roughly 3000 to 5400 BCE. This means that every individual from the ancient world before approximately 3000 BCE — every Sumerian scribe, every Neolithic farmer, every Egyptian who lived before the Old Kingdom — is either an ancestor of every person now living, or an ancestor of no person now living. There is no third option. Most of them, whose lineages survived in any form, are ancestors of everyone.

What this means for ancient lives

The implication for how we think about ancient people is more disconcerting than it sounds at first hearing. The Pharaoh Khufu, who oversaw the construction of the Great Pyramid around 2560 BCE, is either an ancestor of every human now alive, or an ancestor of none. Cleopatra VII, who died in 30 BCE, is more recent than the IAP but old enough that she is very likely an ancestor to everyone. The Buddha, who lived in the 5th century BCE, is too recent to fall squarely within the IAP, but is statistically likely to be a universal ancestor. Confucius, born around 551 BCE, is in the same statistical position — as Mark Liberman of the University of Pennsylvania has explored on Language Log, the question of whether everyone is descended from Confucius is a near-certain “yes” by Chang’s model, even though no living person can prove it through documented genealogy.

The Rohde paper noted this consequence in its closing summary, which the Yale press release quoted directly: “No matter the languages we speak or the color of our skin, we share ancestors who planted rice on the banks of the Yangtze, who first domesticated horses on the steppes of the Ukraine, who hunted giant sloths in the forests of North and South America, and who labored to build the Great Pyramid of Khufu.” The phrasing is not metaphorical. It is the literal mathematical implication of the model. Every reader of this article is descended from these people. So is every person who is not reading this article. The genealogical history of humanity is much more connected than the cultural history of humanity has been able to remember.

Why we can never identify the person

The reason we cannot know who the most recent common ancestor was is that the mathematical argument does not need to know. The model establishes that such a person must have existed, that they must have lived approximately 3,000 years ago, and that everyone alive today is descended from them. It cannot tell us their name, their birthplace, their language, or their occupation. The historical record is too sparse, the genetic signal from any single individual is too diluted after 100+ generations of intermarriage to be detectable, and the genealogical record is too incomplete for any culture more than a few centuries old. According to the paleoanthropologist John Hawks of the University of Wisconsin-Madison, in his 2025 overview of the field, the MRCA is best understood as a mathematical certainty rather than an identifiable historical figure: the family tree of humanity provably converges on such a person around 3,000 years ago, but the convergence happens through so many independent inheritance lines that any single ancestor’s genetic contribution is statistically indistinguishable from noise after roughly thirty generations.

The Chang paper closes with a forward-looking version of the same result. The MRCA of the human population in the year 4026 will be someone alive today, or someone who lived in the recent past. “Within two thousand years,” Chang noted, “it is likely that everyone on earth will be descended from most of us.” Whoever is reading this in 2026 has roughly even odds of being a universal ancestor by the year 4026, depending only on whether their descendant lines persist or die out. The category of “ancestor to all of humanity” is one that ordinary people drift into and out of without any awareness of the change. Some of you reading this will be in it. Some will not. The mathematics determines the eventual outcome; the historical record will never record it.