Playbook
Tasteful Skills
“Tasteful Skills” argues that the best agent skills are not documentation or best-practice lists.
Back to Podcast Digest
Did Modern Humans Help Make Neanderthals? – David Reich
TL;DR
David Reich thinks the standard human-family tree may be missing the real story — instead of Neanderthals and Denisovans simply splitting from modern humans 700,000–800,000 years ago, he’s interested in a 200,000–300,000-year-old episode where a modern-human-related population may have helped reshape Neanderthals.
That “small” admixture event could explain why Neanderthals look culturally closer to us than to Denisovans — Reich points to the spread of Middle Stone Age/Middle Paleolithic or Levallois-style stone tools, fire use, and long-distance stone transport across Africa and Europe, but not East Asia.
The genetics are weird in exactly the way his model tries to solve — Neanderthals and Denisovans cluster together in the whole genome, but Neanderthal mitochondrial DNA and Y chromosomes look much closer to modern humans, with shared ancestors only about 300,000–450,000 years ago.
Reich says the current consensus asks us to accept an unlikely coincidence — if modern-human-related ancestry in Neanderthals is only about 5%, it’s surprising that both mitochondrial DNA and Y chromosomes would have risen to 100% frequency just by chance.
A fossil population from Sima de los Huesos in Spain gives this idea some shape — these 300,000–400,000-year-old individuals look Neanderthal-like in their nuclear genome but Denisovan-like in mitochondrial DNA and Y chromosome, suggesting a turnover event really may have happened.
His bigger claim is that the same expanding population may have transformed both Europe and Africa — in Europe it may have mixed into local archaics and become ~95% local genetically while preserving culture, while in Africa it may have contributed to the ancestry of all living humans through a roughly 80/20 mixture with a deeply diverged archaic African lineage.
The Breakdown
Reich opens with a heretical tweak to the family tree
He starts from the standard post-2010 picture: Neanderthals and Denisovans split from each other around 500,000–600,000 years ago, and both split earlier from modern humans around 700,000–800,000 years ago. Then he immediately complicates it with evidence for a later interbreeding event, around 200,000–300,000 years ago, that may have contributed about 5% modern-human-related DNA into Neanderthal ancestors.
Why Neanderthals feel oddly close to us
Reich says the archaeology has always made Neanderthals look more similar to modern humans than Denisovans do, even though Denisovans are genetically closer to Neanderthals on average. That mismatch is what pulls him toward a different framing: maybe Neanderthals were genetically mostly “Denisovan-like” or archaic, but culturally much closer to modern humans.
The Middle Stone Age as the real turning point
He centers the story on the Middle Stone Age/Middle Paleolithic, also called Levallois or “Mode 1” in his telling: carefully prepared stone cores, high-quality flint moved long distances, and broader fire use appearing around 300,000–400,000 years ago in places like East Africa and the Caucasus. His hunch is that this wasn’t just parallel invention — it may reflect a shared population spreading a toolkit and way of life across Africa and Europe, while East Asia stayed outside that cultural zone.
The genetics that make the standard model feel “crazy”
Here’s the puzzle: whole genomes put Neanderthals with Denisovans, but Neanderthal mitochondrial DNA and Y chromosomes cluster with modern humans much more recently, only about 300,000–450,000 years ago. Reich says the current view effectively asks us to believe that a mere 5% admixture event replaced both lineages to 100% frequency, which he calls a surprising, low-probability outcome on first principles.
Sima de los Huesos is the eerie clue
He points to the famous Spanish site, Sima de los Huesos, dated to roughly 300,000–400,000 years ago. Those individuals look Neanderthal-like in their nuclear genome, but Denisovan-like in mitochondrial DNA and Y chromosome — exactly the kind of partial turnover pattern you’d expect if an incoming population displaced some lineages while leaving most local ancestry in place.
The wave-front model: modern culture, local genes
Reich’s speculative model is almost ecological: a modern-human-related population enters Europe from somewhere like the Middle East, spreads deme by deme, and interbreeds lightly at the expansion front. Simulations from other species suggest that even limited interbreeding during an expansion can lead to massive introgression of local genes, so by the time the population reaches the far side of Europe it’s about 95% local genetically but may still carry the incoming culture — maybe even through matrilineal transmission of toolmaking, preserving mitochondrial DNA.
The same expansion may also help explain us
He then widens the lens: modern humans, too, may be admixed, with roughly 80% ancestry from an early modern lineage and about 20% from an archaic African lineage diverged around 1.5 million years ago, mixing 200,000–300,000 years ago. In this version, one key population spreads Levallois technology in all directions, helps form Neanderthals in Europe, helps form the ancestors of everyone alive today in Africa, and turns an earlier 200,000–300,000-year-old revolution into a formative event on par with the later behavioral “revolution” 50,000–100,000 years ago.
The punchline: maybe Neanderthals were closer kin than we say
By the end, Reich is careful — “probably is wrong, who knows” — but you can feel the excitement. If this is even partly true, Neanderthals are not just our distant sister group: they share our Y chromosome, our mitochondrial line, a major technological package, and maybe a common formative population, making them in an important sense much closer cousins than the simple genome tree suggests.
Share
Keep Reading
Two Ways to Follow
The Weekly Echo. The inbox-shaped summary of what mattered.
New editorials announced here.
