Many Mothers

Copyright: Sanjay Basu

 How a 44 person genome study from southern Africa quietly retired the lonely Eve

There is a story most of us absorbed somewhere between high school biology and a National Geographic cover. A single woman, somewhere in East Africa, gives birth to the modern human lineage. We were taught to call her Mitochondrial Eve, although the textbooks were always careful to remind us, in slightly embarrassed footnotes, that she was not the only woman alive at the time. She was just the most recent common matrilineal ancestor of every person who happens to be reading this. A statistical bottleneck wearing a halo.

It was a beautiful idea. Tidy, monotheistic in shape, easy to draw on a slide. And it was, as a new wave of African genomic research has been politely pointing out for several years now, almost certainly wrong in the way that it mattered most.

The latest reminder came across my feed this morning, courtesy of UC Davis and the work of Brenna Henn, Simon Gravel, Aaron Ragsdale and colleagues. The paper itself, A weakly structured stem for human origins in Africa, dropped in Nature back in 2023, but the conversation has only gotten richer since, and the press release that prompted this piece is the kind of thing that deserves to be read past the headline. Especially if, like me, you suspect that the way we tell the human origin story tells us at least as much about us as it does about the past.

Here is the short version. They sequenced 44 whole genomes from Nama individuals in southern Africa, the Khoe-San population that has long sat awkwardly inside any tree-shaped model of human descent because their genetic diversity simply refused to fit. They added eastern and western African genomes to the mix. They borrowed a clever trick from population genetics, looking at linkage disequilibrium and diversity statistics together, which lets you discriminate between competing demographic histories without needing fossils to settle every argument. Then they ran the models. Lots of them. The one that fit best is what they call a weakly structured stem.

Translation, before the jargon eats us alive. There was no single ancestral population. There was no clean trunk that one day branched neatly into peoples. What there was, somewhere between roughly half a million and a hundred and thirty five thousand years ago, was a loose federation of human groups, scattered across the continent, occasionally drifting apart, frequently drifting back together, swapping genes the way you swap recipes at a long family reunion that nobody quite remembers having started. The earliest split that still leaves a fingerprint in living people is dated to about 120,000 to 135,000 years ago, and even then the splitting was leaky.

So not Eve. Many mothers. Plural, distributed, in conversation with each other across deep time.

I have been thinking about why this finding lands harder than it should. The technical content is, frankly, not that surprising. Anyone who has been watching African genomics for the last decade has seen this coming. Hammer’s work on archaic admixture in Africa hinted at it back in 2011. Scerri and colleagues argued for a structured African origin in 2018. The Henn group has been chipping away at the single origin model for years. What changed in 2023 is that the model finally became quantitative enough, and parsimonious enough, that the older “ghost lineage” hypotheses started looking like extra epicycles.

That is the part worth pausing on. For a while, geneticists were comfortable explaining anomalies in modern African DNA by invoking contributions from some unknown archaic hominin that had interbred with our ancestors. A mystery cousin, never sampled, never named, but mathematically required to balance the books. The new model says you do not need that cousin. Internal structure within the ancestral Homo stem itself, with continuous low level gene flow between weakly differentiated populations, is enough to explain the patterns. Occam, somewhere, raises a small glass.

And the numbers are humbling. Only about one to four percent of the genetic differentiation among living human populations traces back to variation between those ancestral stem populations. We are, by a wide margin, more alike than the deep time geography suggests. Which is also the answer, by the way, to the perennial bad faith question of what a “race” is in biology. The honest reply is that the differences are real, very recent, very shallow, and considerably less interesting than the long centuries of mixing that produced them.

Here is what I find genuinely beautiful in this. The model has consequences for the fossil record that almost nobody is talking about. If the early human stem populations were continuously exchanging genes, they would have looked broadly similar to one another. Which means fossils that look strikingly different from anatomically modern humans, the most famous example being Homo naledi down in the Rising Star cave system, are unlikely to be lineages that contributed to us. They were doing their own thing. Cousins, not parents. The story of our species has been simplified by being made messier in the right way.

I want to put this next to something that has been rattling around my head for a while. In QBism, the version of quantum mechanics I keep returning to, there is a useful slogan that the universe does not arrive pre-labeled. An agent has to do the work of carving it up, placing bets, and updating priors when the world talks back. The reason single origin stories tend to capture our imagination so completely is that they pre-label the world for us. There was an Eve. There was a Garden. There was a moment when the lineage began. The narrative is closed, the categories are stable, the metaphysics gets to take the rest of the afternoon off.

The weakly structured stem refuses to do us this favor. It says, instead, that the boundary between species and population is itself a rough human convenience, and that the actual history is one of contact, hybridization, near separation, reconnection, and a great deal of getting on with daily life across a continent that was warmer and wetter and more crisscrossed than we usually picture it. There is no first mother. There were many. They knew each other, occasionally. Their grandchildren met. Some of those grandchildren walked north, eventually, but the stem they came from was already a braid.

A small aside on methodology, because I think it matters and because most coverage glides past it. The reason this paper could be written at all is that the Nama participants in the study, between 2012 and 2015, allowed researchers to collect saliva samples in their villages while they went about their lives. The genetic diversity of the Khoe-San is the single most informative dataset in human population genetics. Not because they are somehow more “ancient” than the rest of us, that is a confusion you still see in the popular press, but because their effective population size has been large for longer, which means they preserve a deeper coalescent history. If you want to read the early chapters of the human story, you read them in southern African DNA. You do not read them in mine, or yours, or in a Neanderthal toe bone from a Siberian cave, although those are interesting books too.

Two follow up studies have made this picture sharper since 2023. A 2024 paper in Nature Ecology and Evolution documented nine thousand years of genetic continuity in southernmost Africa, which is the kind of thing that should make any historian nervous about claims of recent population turnover. And a more recent Nature analysis of 28 ancient southern African genomes, ranging from 10,200 to 150 years before present, found Homo sapiens specific genetic variants that are no longer present in living people, along with patterns of variation that simply cannot be reconstructed from any modern population. Some of the variation we used to have, we lost. Some of what we are missing might be exactly what an evolutionary biologist would most want to read.

If you take all of this seriously, and I do, the cleanest way to phrase the finding is that human evolution looks less like a tree and more like a river delta. Channels braid. Some silt up. Others reopen after a wet season. The water that finally reaches the sea is the same water that has been moving back and forth between channels for a hundred millennia. Anyone who has watched the Sundarbans from the air knows what this looks like, and anyone who has tried to draw it on a slide also knows that the slide always lies a little.

There is a temptation, when a result like this lands, to treat it as another small humiliation of the scientific intuitions of fifty years ago. I would resist that. The Eve model was never wrong about the fact that we share extremely recent ancestry. It was wrong about the geometry. The replacement model is harder to draw, harder to teach in a single class, and harder to fit on a magazine cover. It is also, almost certainly, closer to what actually happened.

We did not descend from a woman. We descended from a conversation.

And the conversation, as far as anyone can tell, never really stopped. It just took a long walk out of Africa and forgot, for a while, who its mothers were.

If you found this interesting, the underlying paper is Ragsdale et al., “A weakly structured stem for human origins in Africa,” Nature 617, 755 to 763 (2023). The follow up work on ancient southern African genomes is well worth your weekend, particularly if you have ever wondered how much of the human story is missing from your own chromosomes.

Human Evolution

Homo Sapiens