Twenty-six human species in seven million years
Drag the date: the page shows who was alive, and with which tools
01 · Timeline
Who was alive, and when
Each bar is the known fossil range of a species: the distance between the oldest and the most recent find attributed to it with reasonable consensus. It is not the real duration of the species, which is almost certainly wider in both directions. Dashed bars mark taxonomic attributions that are still under discussion.
1.80 Ma Timeline
5 species living at the same time in the fossil record at this date.
Technology already documented: LomekwianOldowanAcheuleanPrepared core (Levallois)Blades and laminar flakesMicroliths
No stone tool industry documented at this date.
Drag, click, or use the left and right arrow keys on the band. Open a bar for the species card. Logarithmic time scale: the last 100,000 years take up about a third of the width.
- basal hominins
- Australopithecus / Kenyanthropus
- Paranthropus
- early Homo
- Middle Pleistocene Homo
- late populations
- Homo sapiens
- genetically defined population
- dashed: attribution or dating disputed
02 · Encephalisation
The brain did not grow in a straight line
Volumes are given in cubic centimetres, measured as endocranial volume. The circles are scaled by volume, not by area, so the visual difference matches the real one.
- Sahelanthropus 365 cm³
- H. floresiensis 426 cm³
- A. afarensis 445 cm³
- A. africanus 465 cm³
- P. boisei 510 cm³
- H. naledi 540 cm³
- H. habilis 610 cm³
- H. erectus 950 cm³
- Middle Pleistocene 1,230 cm³
- H. sapiens 1,350 cm³
- H. neanderthalensis 1,410 cm³
Careful with the numbers
The Neanderthal mean (about 1,410 cm³) is above that of present-day Homo sapiens (about 1,350 cm³), and Upper Palaeolithic sapiens had brains larger on average than ours. Absolute size says little: what counts is the reference body mass, the internal organisation and individual variability, which is wide in every species. The means for H. erectus and H. neanderthalensis cover ranges of over 600 cm³.
03 · Stone tool industry
Six ways to break a stone
The technological sequence is older than the genus Homo and does not belong to any one species. The dates mark the first documented appearance, not the spread: every technique survives for hundreds of thousands of years alongside the ones that follow.
-
Mode 0
Lomekwian
first appearance 3.3 Ma
Large cores and flakes from Lomekwi 3, in Kenya, made by percussion on an anvil. They come half a million years before the oldest known Homo. Who made them remains undetermined, and some authors dispute that the context is intentional.
-
Mode 1
Oldowan
first appearance 2.9-2.6 Ma
Choppers, cores and sharp flakes with minimal retouch. At Nyayanga, in Kenya, the industry dated to 2.9 million years is found with Paranthropus teeth: stone tool production is not the preserve of the genus Homo.
-
Mode 2
Acheulean
first appearance 1.76 Ma
Bifaces and handaxes, that is, tools with a shape decided in advance, symmetrical on both planes. They appear at Kokiselei 4, in Kenya. They call for an operational sequence planned over dozens of removals. They last for over 1.5 million years.
-
Mode 3
Prepared core (Levallois)
first appearance about 400-300 ka
The core is shaped in advance so that a flake of predetermined form comes off with a single final blow. Adopted independently by Neanderthals in Eurasia and by Homo sapiens in Africa, where it defines the Middle Stone Age.
-
Mode 4
Blades and laminar flakes
first appearance about 50-45 ka
Serial production of blades from prismatic cores, with standardised blanks and systematic working of bone, antler and ivory. Thrown weapons, needles and equipment for extreme cold appear.
-
Mode 5
Microliths
first appearance about 65 ka onwards
Tiny stone elements hafted in series onto wooden or bone supports. The backed points of Howiesons Poort, in South Africa, are about 65,000 years old. The logic is modular: interchangeable components, repair instead of replacement.
Fire
Traces of burning in a human context appear at Wonderwerk (South Africa, about 1 million years) and at Gesher Benot Ya'aqov (Israel, about 790,000 years), but they stay episodic. Structured, repeated hearths, that is, genuinely habitual use, are documented only after 400,000 years. Cooking has demonstrable effects on digestibility and on the energy budget, but the idea that it drove the increase in brain size in H. erectus remains a hypothesis, not an established fact.
04 · Geography
The exits from Africa, in the plural
There is no single “Out of Africa”. There are at least two large dispersals of the genus Homo and several smaller expansions of Homo sapiens that left no living descendants. The outlines of the land change: during glacial maxima sea level falls by as much as 120 metres and the Sunda, Sahul and Beringia shelves emerge.
2.1-1.5 million years
The first exit: Homo erectus in Eurasia
By 1.8 million years ago hominins of the erectus group are at Dmanisi, in Georgia. The tools from Shangchen, in China, dated to 2.12 million years, push the dispersal back further still, although they are not found with human remains. In Java the Sangiran levels are around 1.5 million years. No sea crossing is needed: the route is entirely overland.
Europe peopled and emptied several times
The Happisburgh footprints, in England, are between 950,000 and 850,000 years old and document a human presence at cold latitudes. At Atapuerca the sequence spans almost a million years. This is not a continuous colonisation: European populations die out and are replaced several times, following the glacial cycles. In February 2026 the remains from Thomas Quarry, in Morocco, were dated to about 773,000 years.
Four humanities at the same moment
In Africa the first Homo sapiens: Jebel Irhoud in Morocco at 315,000 years, Florisbad in South Africa at 259,000, Omo I in Ethiopia at no less than 233,000. In Europe and western Asia the Neanderthals. In East Asia the Denisovan population, documented from the Altai to Tibet and as far as Harbin and Taiwan. In South Africa Homo naledi, with a brain of half a litre.
The exits with no descendants
Homo sapiens leaves Africa long before the main dispersal. The Misliya jaw, in Israel, is between 194,000 and 177,000 years old. At Al Wusta, in Saudi Arabia, a finger bone 88,000 years old. At Tam Pa Ling, in Laos, remains between 86,000 and 68,000 years old. None of these populations left a detectable genetic trace in people living today.
The dispersal that took hold
Every non-African human being alive today descends from this expansion. With sea level 120 metres lower, Sunda is a continental peninsula and Sahul joins Australia and New Guinea: even so, at least 70 km of open sea in Wallacea still have to be crossed. Madjedbebe, in Australia, is dated to between 65,000 and 60,000 years, with an open debate about how reliable the stratigraphy is. In Europe sapiens are documented at Ranis, Bacho Kiro and Grotta del Cavallo around 45,000 years.
Extreme cold, then the Americas
At Yana, in the Siberian Arctic, permanent settlements at 32,000 years. The White Sands footprints, in New Mexico, are dated to between 23,000 and 21,000 years: in 2025 a third independent line of evidence, radiocarbon on the mud of the ancient lake, confirmed the range, but part of the archaeological community is still sceptical, because no stone industry has been found with them. Monte Verde, in Chile, is 14,500 years old. Clovis, for decades taken to be the oldest culture on the continent, is 13,000.
Domestication, and the last islands
Agriculture arises independently in at least eleven regions. The last lands to be reached are the most remote: the Austronesian expansion sets out from Taiwan around 5,000 years ago, the Lapita culture reaches island Melanesia at 3,300, eastern Polynesia around AD 1000, New Zealand around 1280. Madagascar is settled by Austronesian navigators in the first millennium.
- archaic hominins
- Homo sapiens
- centres of domestication
- shelves exposed at glacial maxima
- meridians every 30°, parallels every 18° · dashed: tropics · ochre line: equator
The continental shelf shown follows the 200 metre isobath, used as an approximation of the coastline during glacial maxima. Actual sea level fall at the Last Glacial Maximum was about 120 metres, so the shelf drawn here is slightly larger than the area that was truly exposed.
05 · Molecular evidence
The archaic DNA we all carry
Neanderthals and Denisovans are populations with which Homo sapiens had fertile children, and their genomes are still inside ours. The percentages refer to the fraction of the nuclear genome of archaic origin in populations living today.
- Europe and western Asia 1.8%
- East Asia 2.0% + 0.2%
- Papua and Indigenous Australia 1.9% + 4.5%
- Sub-Saharan Africa 0.3%
- Neanderthal
- Denisova
- Sub-Saharan Africa: the Neanderthal fraction comes from migrations back out of Eurasia
-
50.5-43 ka
The window of interbreeding
Two independent studies published in December 2024 place the main Neanderthal gene flow in a window of about 7,000 years, peaking around 47,000 years ago. The genomes from Ranis and Zlatý kůň show a single event shared by all non-Africans.
-
430 ka
The oldest nuclear DNA
Recovered from the remains at the Sima de los Huesos, at Atapuerca. Beyond that threshold DNA degrades past recovery, and the only molecular access left is palaeoproteomics, which reads enamel proteins that have survived for millions of years.
-
146 ka
The face of the Denisovans
The Harbin cranium, in China, identified in 2025 through 95 endogenous proteins and mitochondrial DNA extracted from dental calculus. For the first time an archaic genome has an almost complete skull to go with it.
-
EPAS1
An allele that is still useful
The Tibetan variant that regulates the response to hypoxia at high altitude is of Denisovan origin. Other archaic alleles act on immunity, blood clotting and lipid metabolism; some affect the severity of respiratory infections.
-
Sediments
DNA without fossils
Hominin DNA can be taken straight from cave sediments, with no need for bones. It has revealed a Neanderthal and Denisovan presence at sites where not one skeletal fragment has ever been found.
06 · Symbolic behaviour
Before the cave art of Europe
The dates are minimum ages: they show when that behaviour was already present, not when it began. Open diamonds mark attributions or datings that are disputed in the literature.
- 176 ka
Structures built from stalagmites
Two rings of broken stalagmite fragments, deliberately arranged, 336 metres inside the cave of Bruniquel, with traces of fire. At that date and in that place the only humans present were Neanderthals.
Bruniquel-sur-Aveyron, France - 150-142 ka
Shell beads
Thirty-three Tritia shells, perforated and worn by hanging. They are the oldest personal ornaments securely dated: an object worn to say something to somebody else.
Bizmoune, Morocco - 130 ka
Worked eagle talons
Eight sea eagle talons with notches and polish from mounting, probably part of a composite Neanderthal ornament.
Krapina, Croatia - 115 ka
Pigments and perforated shells
Pigment residues in shells used as containers, in a Neanderthal context.
Cueva de los Aviones, Spain - 100 ka
An ochre workshop
Two abalone shells used as containers, holding a mixture of ochre, marrow fat and charcoal, together with the tools to prepare it. A recipe, that is, a procedure handed down.
Blombos, South Africa - 78 ka
A child burial
A child of about three laid in a dug pit, with the head supported and the body wrapped. The oldest deliberate human burial documented in Africa.
Panga ya Saidi, Kenya - 73 ka
A cross-hatched drawing
Nine crossing lines drawn with an ochre crayon on a flake of silcrete. The flake is not a decorated tool, but a surface used only to make a mark.
Blombos, South Africa - 64.8 ka
Neanderthal cave art
Calcite crusts over paintings in three Iberian caves, dated by uranium series to a time when there were no sapiens in Iberia. The dating has been challenged by several groups and the question is still open.
La Pasiega, Ardales, Maltravieso, Spain - 51.2 ka
The oldest narrative scene
Three human-like figures interacting with a warty pig, dated by laser ablation uranium series. It is the oldest figurative art and the oldest known story told in pictures.
Leang Karampuang, Sulawesi, Indonesia - 42 ka
Musical instruments
Flutes made from vulture bone and from mammoth ivory, with holes spaced to produce a scale.
Geissenklösterle and Hohle Fels, Germany - 40.8 ka
Painted discs
The oldest dated cave paintings in Europe, made by blowing pigment around a hand and onto discs.
El Castillo, Spain - 36-33.5 ka
Chauvet Cave
Over a thousand figures using shading, perspective and overlap to suggest movement. The repertoire is dominated by predators, not by prey: lions, bears, rhinoceroses.
Ardèche, France - 34 ka
Burials with grave goods
Two children laid head to head with over ten thousand ivory beads, spears of straightened tusk and ornaments. An enormous investment of labour, and the first clear evidence of social differentiation in the treatment of the dead.
Sungir, Russia - 17 ka
Lascaux
The best known cycle of the European Palaeolithic, close to the end of a painting tradition that lasted over twenty millennia.
Dordogne, France
07 · Domestication
Agriculture was invented at least eleven times
The Fertile Crescent is the oldest and best documented centre, not the single origin. In at least eleven regions, with no contact between them, different populations domesticated local plants and animals within the space of a few millennia. The dates are in years before present and mark the earliest evidence of domestic species.
-
11,500 years
Fertile Crescent
Einkorn and emmer wheat, barley, lentils, chickpeas, peas, flax. Sheep, goat, pig and cattle over the following two millennia.
-
about 10,000 years
Mesoamerica
Squash first of all, then maize from teosinte in the Balsas basin around 9,000 years, and beans. No large animal available that could be domesticated.
-
about 9,000 years
Yangtze basin
Short-grain rice, domesticated in wetland. In the Yellow River basin, in parallel, foxtail millet and broomcorn millet.
-
about 8,000 years
Central Andes
Potato, oca, quinoa. Llama and alpaca are the only large pack mammals domesticated in the Americas.
-
about 8,000 years
South-western Amazonia
Cassava, peanut, chilli, within a practice of forest management that has left traces in the composition of the forest today.
-
about 7,000 years
Highlands of New Guinea
Taro and banana at Kuk Swamp, with artificial drainage systems dug into the peat.
-
about 5,000 years
Eastern North America
Sunflower, goosefoot, little barley. An autonomous crop complex, later largely displaced by Mesoamerican maize.
-
about 4,500 years
Sahel and West Africa
Pearl millet and sorghum, suited to short and irregular rainy seasons. African rice in the inner Niger delta.
-
about 4,000 years
Ethiopian highlands
Teff, enset, noug. Crops that remain almost entirely regional to this day.
-
about 4,000 years
Western New Guinea and Wallacea
Coconut, sugar cane, greater yam, the basis of the Austronesian expansion into the Pacific that followed.
-
20,000 years
Southern China, out of sequence
The pottery from Xianrendong comes ten millennia before agriculture. Pottery, grinding stones and permanent villages do not arrive as a block: the idea of a single “Neolithic package” does not hold.
The cost of the Neolithic
The transition brings a food surplus and population growth, and with them a less varied diet, falling average height, more caries and dental disease, zoonotic illness from living alongside livestock, and the first material inequalities readable in grave goods. Göbekli Tepe (Turkey, about 11,500 years) shows collective monumental architecture before agriculture, which reverses the causal order of the classic account.
08 · Ancestry
How the populations of today were formed
No population living today descends from a single ancient group: every living genome is the result of successive mergers between populations that had stayed apart for millennia. Select a region and see which ancient populations it derives from, when each admixture happened and where it came from.
Before reading the numbers
The components are not races. They are the coefficients of a statistical model that breaks a present-day sample down into proportions attributable to ancient genomes used as references. Change the reference populations and the numbers change. They are descriptive tools, not biological entities.
About 90% of human genetic variation lies within single local populations. The fraction attributable to differences between large continental groupings stays at around 10%. Two people from the same village can differ from each other more than two continents differ on average. The result, obtained by Lewontin in 1972 on classical markers, has been confirmed by genome-wide data.
Variation is clinal. It changes gradually with geographic distance, with no discrete boundaries. There is no point on the map where one genetic group ends and another begins. Ethnic groups and nations are historical, linguistic and political categories, and they do not coincide with ancestry components.
The percentages below are approximate ranges, rounded and dependent on the model. In the publications they come with standard errors that cannot be shown here. They are to be read as orders of magnitude.
North-western Europe
- ancient reference population
- ghost population, no fossils
- selected region
The thickness of the arrows on the map is proportional to the contribution. The dashes mark ghost populations, inferred from the calculation and never found as fossils.
Three populations layered in five thousand years
The post-glacial European hunter-gatherers are largely replaced by the Anatolian farmers between 8,500 and 6,000 years ago. In some regions the replacement exceeds 90% of the gene pool, with a partial recovery of the indigenous component over the following millennia.
Around 5,000 years ago the third wave arrives: the steppe pastoralists, associated with the Corded Ware culture and then with the Bell Beaker phenomenon. In Britain the turnover is almost total, and the earlier Y chromosome all but disappears. The movement is strongly skewed towards the male side.
Lactase persistence into adulthood, that is, lactose tolerance, spreads with this expansion and after it, but only reaches high frequencies in the Iron Age, long after herding began.
Three-way model WHG + ANF + Steppe (Haak 2015, Olalde 2018, Allentoft 2024)
- Pontic-Caspian steppe pastoralists about 47% The Yamnaya population, itself a merger between eastern European hunters and a southern component tied to the Caucasus. They bring the domestic horse, the wagon and, on the most widely accepted model, the Indo-European languages.
- Anatolian Neolithic farmers about 35% The origin of European agriculture. From Barcın and the Anatolian sites their descent spreads across Europe between 8,500 and 6,000 years ago, by the Danubian and the Mediterranean route.
- Western European hunter-gatherers about 15% A post-glacial European population, represented by the Loschbour genome, in Luxembourg, and by the Villabruna cluster. Dark skin and light eyes in the combination reconstructed from known alleles.
- Neanderthal about 2% Present in every non-African genome at a fraction between 1.5% and 2.5%, and in traces in Africa through migrations back from Eurasia.
- Basal Eurasians (ghost) about 1% A ghost population inferred in the Near East: it splits from the other Eurasians before the interbreeding with Neanderthals.
The thickness of the arrows on the map is proportional to the contribution. The dashes mark ghost populations, inferred from the calculation and never found as fossils.
The island that took in the least steppe ancestry
Sardinia keeps the highest fraction of Anatolian Neolithic ancestry in the whole of Europe. Sardinian populations are the standard reference in models that reconstruct the early European farmers.
The steppe expansion arrives late and on a reduced scale. It is a case of geographic attenuation, not of absolute isolation: Phoenician, Punic and Roman contributions are documented, but they do not alter the underlying structure.
The Iceman of the Similaun, found on the border between Italy and Austria and dated to 5,300 years, is genetically closer to Sardinians today than to the populations who now live where he died.
Three-way model, ANF reference (Fernandes 2020, Marcus 2020)
- Anatolian Neolithic farmers about 78% The origin of European agriculture. From Barcın and the Anatolian sites their descent spreads across Europe between 8,500 and 6,000 years ago, by the Danubian and the Mediterranean route.
- Western European hunter-gatherers about 10% A post-glacial European population, represented by the Loschbour genome, in Luxembourg, and by the Villabruna cluster. Dark skin and light eyes in the combination reconstructed from known alleles.
- Pontic-Caspian steppe pastoralists about 10% The Yamnaya population, itself a merger between eastern European hunters and a southern component tied to the Caucasus. They bring the domestic horse, the wagon and, on the most widely accepted model, the Indo-European languages.
- Neanderthal about 2% Present in every non-African genome at a fraction between 1.5% and 2.5%, and in traces in Africa through migrations back from Eurasia.
The thickness of the arrows on the map is proportional to the contribution. The dashes mark ghost populations, inferred from the calculation and never found as fossils.
Where agriculture arrives last
In the eastern Baltic the hunter-gatherer component stays dominant for millennia after agriculture is introduced in the rest of Europe. The Anatolian farmers arrive late and in small numbers.
The steppe expansion, by contrast, is early and intense, and on that base the local indigenous component is not erased but absorbed. The result is the highest fraction of hunter-gatherer ancestry in Europe.
The mitochondrial lineages of the region show continuity from the Mesolithic, while the Y chromosome records the turnover: the same asymmetry between the sexes seen elsewhere in the steppe expansion.
WHG/EHG + ANF + Steppe model (Mittnik 2018, Allentoft 2024)
- Pontic-Caspian steppe pastoralists about 44% The Yamnaya population, itself a merger between eastern European hunters and a southern component tied to the Caucasus. They bring the domestic horse, the wagon and, on the most widely accepted model, the Indo-European languages.
- Western European hunter-gatherers about 22% A post-glacial European population, represented by the Loschbour genome, in Luxembourg, and by the Villabruna cluster. Dark skin and light eyes in the combination reconstructed from known alleles.
- Eastern European hunter-gatherers about 16% A population of Karelia and the Urals, carrying an Ancient North Eurasian component absent further west.
- Anatolian Neolithic farmers about 16% The origin of European agriculture. From Barcın and the Anatolian sites their descent spreads across Europe between 8,500 and 6,000 years ago, by the Danubian and the Mediterranean route.
- Neanderthal about 2% Present in every non-African genome at a fraction between 1.5% and 2.5%, and in traces in Africa through migrations back from Eurasia.
The thickness of the arrows on the map is proportional to the contribution. The dashes mark ghost populations, inferred from the calculation and never found as fossils.
Four divergent populations that come back together
Between 12,000 and 8,000 years ago the Near East holds populations that are genetically very distinct from one another: Natufians in the Levant, Anatolian farmers, groups of the Iranian Zagros, hunters of the Caucasus. The genetic distances between them are comparable to those that today separate Europeans and East Asians.
Over the course of the Holocene these populations merge step by step, and the homogeneity of the region today is the recent product of that fusion. The ancient structure is gone.
The basal Eurasian component has never been found as a fossil: it is deduced from the Neanderthal fraction being lower than expected in the populations of the region.
Four-way model with a basal component (Lazaridis 2016, Feldman 2019)
- Anatolian Neolithic farmers about 42% The origin of European agriculture. From Barcın and the Anatolian sites their descent spreads across Europe between 8,500 and 6,000 years ago, by the Danubian and the Mediterranean route.
- Iranian and Zagros Neolithic about 22% Farmers of the Iranian plateau, represented by Ganj Dareh. They contribute to the Near East and, to a substantial degree, to South Asia.
- Natufians and the Levant about 18% Settled hunter-gatherers of the Levant, the direct forerunners of agriculture in the Fertile Crescent.
- Basal Eurasians (ghost) about 10% A ghost population inferred in the Near East: it splits from the other Eurasians before the interbreeding with Neanderthals.
- Pontic-Caspian steppe pastoralists about 6% The Yamnaya population, itself a merger between eastern European hunters and a southern component tied to the Caucasus. They bring the domestic horse, the wagon and, on the most widely accepted model, the Indo-European languages.
- Neanderthal about 2% Present in every non-African genome at a fraction between 1.5% and 2.5%, and in traces in Africa through migrations back from Eurasia.
The thickness of the arrows on the map is proportional to the contribution. The dashes mark ghost populations, inferred from the calculation and never found as fossils.
A gradient, not two groups
The populations of the subcontinent are spread along a continuous axis between two theoretical extremes, called in the literature Ancestral North Indians and Ancestral South Indians. No population living today sits at either extreme: they are all intermediate points.
The deepest indigenous component, the Ancient Ancestral South Indians, has no ancient genome attached to it. It is inferred in full, because in the climate of the subcontinent DNA does not survive.
Steppe ancestry enters between 4,000 and 3,500 years ago, after the decline of the Indus civilisation, and its distribution correlates with geography and with historical social categories. It is the point on which public debate is fiercest, and it has to be kept distinct from claims about identity: a genomic proportion does not establish who belongs to a place.
ANI-ASI cline, three-way model (Narasimhan 2019)
- Ancient Ancestral South Indians (ghost) about 42% A ghost population: no ancient genome recovered. Inferred as a deep lineage that split early from the main dispersal and survived in the Indian subcontinent.
- Iranian and Zagros Neolithic about 34% Farmers of the Iranian plateau, represented by Ganj Dareh. They contribute to the Near East and, to a substantial degree, to South Asia.
- Pontic-Caspian steppe pastoralists about 18% The Yamnaya population, itself a merger between eastern European hunters and a southern component tied to the Caucasus. They bring the domestic horse, the wagon and, on the most widely accepted model, the Indo-European languages.
- Neanderthal about 2% Present in every non-African genome at a fraction between 1.5% and 2.5%, and in traces in Africa through migrations back from Eurasia.
- Denisova about 1% The fraction peaks in the populations of Papua and Indigenous Australia, at around 4-5%. About 0.2% in East Asia, absent elsewhere.
- East Asian farmers about 3% Populations of the Yellow River and Yangtze basins, tied to millet and to rice, which expand across the whole of East and South East Asia during the Holocene.
The thickness of the arrows on the map is proportional to the contribution. The dashes mark ghost populations, inferred from the calculation and never found as fossils.
Two farming basins that expand
East Asia today derives in large part from the Holocene expansion of farming populations from the Yellow River basin, tied to millet, and the Yangtze basin, tied to rice. These expansions overwrite the structure that came before.
The Palaeolithic populations of the region, such as the Tianyuan individual at 40,000 years near Beijing, are collateral relatives and not direct ancestors of the populations living there today.
The Denisovan component is present but minimal, around 0.2%. The bulk of Denisovan introgression is found instead in island South East Asia and in Oceania.
Two-way model, Yellow River and Amur references (Wang 2021, Yang 2020)
- East Asian farmers about 84% Populations of the Yellow River and Yangtze basins, tied to millet and to rice, which expand across the whole of East and South East Asia during the Holocene.
- North East Asia about 13% A population of the Amur basin and Manchuria, documented at Devil's Gate. It brings rice farming to Japan in the Yayoi period.
- Neanderthal about 2% Present in every non-African genome at a fraction between 1.5% and 2.5%, and in traces in Africa through migrations back from Eurasia.
- Denisova about 1% The fraction peaks in the populations of Papua and Indigenous Australia, at around 4-5%. About 0.2% in East Asia, absent elsewhere.
The thickness of the arrows on the map is proportional to the contribution. The dashes mark ghost populations, inferred from the calculation and never found as fossils.
Three ancestries, with a gradient from north to south
The classic model assumed two origins, Jomon and Yayoi. Palaeogenomic work since 2021 documents three: to the indigenous Jomon and the North Asian component that arrived with rice farming in the Yayoi period is added an influx from East Asia in the Kofun period, tied to the formation of the state.
The Jomon fraction varies a great deal from region to region. A study published in May 2026 on over 3,200 genomes estimates it at around 28.5% in Okinawa and 13.4% in western Japan, where the affinity with Han populations is strongest. The national average hides this internal structure.
The same study identifies a further component in north-eastern Japan, perhaps tied to the Emishi populations of the historical sources, and concludes that the archipelago is genetically more varied than earlier models assumed.
Tripartite model Jomon + Yayoi + Kofun (Cooke 2021, RIKEN 2026)
- East Asian farmers about 64% Populations of the Yellow River and Yangtze basins, tied to millet and to rice, which expand across the whole of East and South East Asia during the Holocene.
- North East Asia about 17% A population of the Amur basin and Manchuria, documented at Devil's Gate. It brings rice farming to Japan in the Yayoi period.
- Jomon about 17% Hunter-gatherer-fishers of the Japanese archipelago, separated from the mainland between 20,000 and 15,000 years ago with an effective population of about a thousand individuals, then isolated for millennia.
- Neanderthal about 2% Present in every non-African genome at a fraction between 1.5% and 2.5%, and in traces in Africa through migrations back from Eurasia.
The thickness of the arrows on the map is proportional to the contribution. The dashes mark ghost populations, inferred from the calculation and never found as fossils.
The reservoir of the Ancient North Eurasians
The Ancient North Eurasian component, represented by the Mal'ta genome of 24,000 years ago, survives in Siberia at fractions that are far more diluted elsewhere.
It is the same component that, by way of the steppe, enters Europe in the Bronze Age, and that by way of Beringia enters every Indigenous American population.
Some Siberian groups carry ancestry that came back from the Americas, through the Bering Strait in the opposite direction, around 5,000 years ago.
Multi-way model with Mal'ta as reference (Sikora 2019, Zhang 2023)
- North East Asia about 52% A population of the Amur basin and Manchuria, documented at Devil's Gate. It brings rice farming to Japan in the Yayoi period.
- Ancient North Eurasians about 30% A Siberian population represented by the Mal'ta genome, 24,000 years old. It contributes both to Europeans today, by way of the steppe, and to every Indigenous American population.
- East Asian farmers about 16% Populations of the Yellow River and Yangtze basins, tied to millet and to rice, which expand across the whole of East and South East Asia during the Holocene.
- Neanderthal about 2% Present in every non-African genome at a fraction between 1.5% and 2.5%, and in traces in Africa through migrations back from Eurasia.
The thickness of the arrows on the map is proportional to the contribution. The dashes mark ghost populations, inferred from the calculation and never found as fossils.
One founding population, then three thousand years of isolation
The ancestral lineage separates from the populations of East Asia around 36,000 years ago, with gene flow continuing until about 25,000. Between 25,000 and 20,000 years ago it takes on the contribution of the Ancient North Eurasians, and becomes isolated in Beringia.
After the entry south of the ice sheets the lineage splits into two main branches, northern and southern, between 17,500 and 14,600 years ago. What follows is extremely rapid diversification: Monte Verde, at the far south of Chile, is 14,500 years old.
The Ancient Beringians, identified from the genome of a girl of 11,500 years ago in Alaska, are an extinct branch: they left no descendants in the populations of today. Since 1492 the genetic structure of the continent has been transformed by European and African input on an enormous scale.
Two-way model (Moreno-Mayar 2018, Posth 2018)
- East Asian farmers about 63% Populations of the Yellow River and Yangtze basins, tied to millet and to rice, which expand across the whole of East and South East Asia during the Holocene.
- Ancient North Eurasians about 34% A Siberian population represented by the Mal'ta genome, 24,000 years old. It contributes both to Europeans today, by way of the steppe, and to every Indigenous American population.
- Neanderthal about 2% Present in every non-African genome at a fraction between 1.5% and 2.5%, and in traces in Africa through migrations back from Eurasia.
- Denisova about 1% The fraction peaks in the populations of Papua and Indigenous Australia, at around 4-5%. About 0.2% in East Asia, absent elsewhere.
The thickness of the arrows on the map is proportional to the contribution. The dashes mark ghost populations, inferred from the calculation and never found as fossils.
The most direct descent from the first dispersal
The Indigenous populations of New Guinea and Australia descend from the first expansion towards Sahul, and have been apart from every other non-African population for over 45,000 years. The continuity in Australia is among the longest documented anywhere in the world.
They carry the highest Denisovan fraction of any living population, between 4% and 5%, inherited from at least two separate interbreeding events with Denisovan populations that were genetically different from each other.
The Austronesian expansion of the last 4,000 years touches the coasts of New Guinea and the surrounding islands, with limited contributions inland.
Model with two Denisovan events (Jacobs 2019, Malaspinas 2016)
- Deep Sahul lineage about 92% Direct descendants of the first dispersal towards Sahul, apart from every other non-African population for over 45,000 years.
- Denisova about 5% The fraction peaks in the populations of Papua and Indigenous Australia, at around 4-5%. About 0.2% in East Asia, absent elsewhere.
- Neanderthal about 2% Present in every non-African genome at a fraction between 1.5% and 2.5%, and in traces in Africa through migrations back from Eurasia.
- Austronesians about 1% An expansion that set out from Taiwan around 5,000 years ago and reached the Philippines, Indonesia, island Melanesia, Polynesia and finally Madagascar.
The thickness of the arrows on the map is proportional to the contribution. The dashes mark ghost populations, inferred from the calculation and never found as fossils.
The Bantu expansion, and an archaic lineage with no fossil
From 6,000-4,000 years ago onwards, out of the plateau between Nigeria and Cameroon, Bantu-speaking populations expand across almost the whole of sub-Saharan Africa. Today about 350 million people speak one of the roughly five hundred Bantu languages.
The genomic analysis published in 2024 on over 1,500 Bantu speakers documents a serial founder model: genetic diversity falls with distance from Cameroon, and Zambia and the Democratic Republic of the Congo emerge as points of recombination. Along the way the populations absorb local groups.
In the genomes of West Africa a contribution of between 2% and 19% is detected from an archaic lineage with no known fossil at all. The African climate destroys ancient DNA, so what is known about the deep structure of the continent stays far poorer than what is known about Eurasia, even though it is there that most human diversity is found.
Models with an unidentified archaic lineage (Durvasula 2020, Fortes-Lima 2024)
- West Africa about 74% A population of the plateau between Nigeria and Cameroon, the origin of the expansion of the Bantu-speaking peoples.
- Hunters of the central African rainforest about 10% Populations of the equatorial forest, separated from the other African lines between 90,000 and 60,000 years ago.
- Unidentified African archaic lineage (ghost) about 10% In the genomes of West Africa a contribution of between 2% and 19% is detected from an archaic population for which no fossil exists. The estimate is wide and disputed, because it depends on the demographic model assumed.
- Ancient East Africa about 6% Represented by the Mota genome, in Ethiopia, 4,500 years old: the only ancient African genome with no Eurasian ancestry flowing back into it.
The thickness of the arrows on the map is proportional to the contribution. The dashes mark ghost populations, inferred from the calculation and never found as fossils.
The gene flow that came back
The Mota genome, in Ethiopia, dated to 4,500 years, is the only ancient African genome with no Eurasian ancestry. Comparison with the populations of the region today has made it possible to quantify a flow back from the Near East into East Africa over the last few millennia.
It is by this route that traces of Neanderthal ancestry appear in African populations: not through direct interbreeding, but through the return of Eurasian populations that already carried it.
The Horn of Africa is also the point of departure of the main dispersal 65,000 years ago. The populations who live there today are not, however, a preserved sample of those: fifty thousand years of demographic history separate them.
Mota as reference for quantifying the return flow (Llorente 2015, Skoglund 2017)
- Ancient East Africa about 56% Represented by the Mota genome, in Ethiopia, 4,500 years old: the only ancient African genome with no Eurasian ancestry flowing back into it.
- Natufians and the Levant about 22% Settled hunter-gatherers of the Levant, the direct forerunners of agriculture in the Fertile Crescent.
- West Africa about 14% A population of the plateau between Nigeria and Cameroon, the origin of the expansion of the Bantu-speaking peoples.
- Hunters of the central African rainforest about 4% Populations of the equatorial forest, separated from the other African lines between 90,000 and 60,000 years ago.
- Neanderthal about 2% Present in every non-African genome at a fraction between 1.5% and 2.5%, and in traces in Africa through migrations back from Eurasia.
- Anatolian Neolithic farmers about 2% The origin of European agriculture. From Barcın and the Anatolian sites their descent spreads across Europe between 8,500 and 6,000 years ago, by the Danubian and the Mediterranean route.
The thickness of the arrows on the map is proportional to the contribution. The dashes mark ghost populations, inferred from the calculation and never found as fossils.
The deepest divergence between living populations
The Khoe-San populations separate from the other human lines between 300,000 and 250,000 years ago, before the appearance of the oldest known sapiens fossils.
Their internal genetic diversity is the highest measured in any human population.
The arrival of pastoralists from East Africa and then of the Bantu expansion brings in outside contributions over the last two thousand years, in proportions that vary widely between groups. The term Khoe-San covers populations distinct in language, economy and history, and does not name a single unit.
Ancient South African references (Schlebusch 2017, Skoglund 2017)
- Khoe-San about 76% The deepest divergence between living human populations, estimated at between 300,000 and 250,000 years ago, and the highest internal genetic diversity of any group.
- West Africa about 14% A population of the plateau between Nigeria and Cameroon, the origin of the expansion of the Bantu-speaking peoples.
- Ancient East Africa about 8% Represented by the Mota genome, in Ethiopia, 4,500 years old: the only ancient African genome with no Eurasian ancestry flowing back into it.
- Neanderthal about 1% Present in every non-African genome at a fraction between 1.5% and 2.5%, and in traces in Africa through migrations back from Eurasia.
- Unidentified African archaic lineage (ghost) about 1% In the genomes of West Africa a contribution of between 2% and 19% is detected from an archaic population for which no fossil exists. The estimate is wide and disputed, because it depends on the demographic model assumed.
The thickness of the arrows on the map is proportional to the contribution. The dashes mark ghost populations, inferred from the calculation and never found as fossils.
The last colonisation of the planet
The Austronesian expansion sets out from Taiwan around 5,000 years ago, reaches the Philippines and Indonesia, and in island Melanesia mixes with Papuan populations. The Lapita culture arrives in Vanuatu and Tonga around 3,300-2,900 years ago.
The first Lapita individuals analysed genetically have almost entirely Asian ancestry: the Papuan component enters afterwards, through later contact. The sequence is therefore arrival and then mixing, not the arrival of populations already mixed.
Eastern Polynesia is reached around AD 1000, New Zealand around 1280, with double-hulled craft and navigation by the stars over routes thousands of kilometres long.
Lapita and Austronesian references (Skoglund 2016, Lipson 2018)
- Austronesians about 72% An expansion that set out from Taiwan around 5,000 years ago and reached the Philippines, Indonesia, island Melanesia, Polynesia and finally Madagascar.
- Deep Sahul lineage about 24% Direct descendants of the first dispersal towards Sahul, apart from every other non-African population for over 45,000 years.
- Denisova about 2% The fraction peaks in the populations of Papua and Indigenous Australia, at around 4-5%. About 0.2% in East Asia, absent elsewhere.
- Neanderthal about 2% Present in every non-African genome at a fraction between 1.5% and 2.5%, and in traces in Africa through migrations back from Eurasia.
The admixtures that produced the present
Every row is a demographic event documented by ancient DNA. Some replaced the population that came before almost entirely, others merged with it.
- 45,000 years
The first Europeans, with no descendants
The genomes of Zlatý kůň, in Bohemia, and of Ranis, in Germany, belong to the deepest split known from the line that left Africa. Neither leaves descendants today.
- 25,000-20,000 years
The forming of the ancestral Americans
In Beringia an East Asian lineage takes on about 35% ancestry from the Ancient North Eurasians. The group that forms stays isolated for a few millennia before entering the American continent.
- 14,000 years
The turnover of the European hunters
With post-glacial warming the Magdalenian populations of Europe are largely replaced by a group tied to the Villabruna cluster, which becomes the basis of the hunter-gatherers that follow. The replacement comes six thousand years before the arrival of agriculture.
- 8,500-6,000 years
The Anatolian farmers in Europe
Two routes, Danubian and Mediterranean. In some regions the indigenous component falls below 10%, and the replacement is almost complete. In others the integration is more gradual, with a partial recovery of the local component over the following millennia.
- 6,000-4,000 years
The Bantu expansion
From the plateau between Nigeria and Cameroon towards the Congo basin, East Africa and finally southern Africa, reached around 1,500 years ago. A serial founder model, absorbing local populations along the way.
- 5,000 years
The steppe into Europe
Yamnaya ancestry reaches most of Europe within about a thousand years. In Britain the turnover of the gene pool exceeds 90%. Strongly skewed towards the male side: the earlier Y lineages all but disappear.
- 5,000 years
The Austronesians from Taiwan
The widest maritime expansion of prehistory: over eight thousand kilometres from Taiwan to Madagascar.
- 4,000-3,500 years
The steppe into South Asia
Steppe ancestry enters the subcontinent after the decline of the Indus civilisation, by way of central Asia. In the same period the gradient that marks South Asian populations today takes shape.
- 3,300-2,900 years
Lapita in remote Oceania
Populations of almost entirely Asian ancestry reach Vanuatu, New Caledonia and Tonga. The Papuan component enters over the generations that follow, through contact with an already inhabited Melanesia.
- 3,000 years
Yayoi in Japan
Populations from North East Asia bring irrigated rice farming to the archipelago and mix with the Jomon. Unlike what happens in Europe with the Anatolian farmers, the indigenous component is not erased.
- 1,700 years
Kofun in Japan
A second continental influx, genetically closer to East Asian populations today, during the phase in which the state takes shape. It is the majority component in Japanese people now.
- 1,000 years
The last islands
Eastern Polynesia around AD 1000, New Zealand around 1280. With this the peopling of the habitable lands of the planet comes to an end.
- 500 years
The colonial admixture
Since 1492 the genetic structure of the Americas, of the Caribbean and of part of Africa has been transformed on a continental scale by European migration, the Atlantic trade in enslaved people and the demographic collapse of Indigenous populations. Most of the mixed populations of today come from it.
Ghost populations and severed branches
The genetic calculation detects contributions from populations for which no fossil exists, and identifies real, well documented groups that left no descendants.
- Ghost population
Basal Eurasians
No fossil. Inferred from an arithmetical anomaly: in the populations of the Near East the Neanderthal fraction is lower than expected, so part of their ancestry must come from a group that split before the interbreeding with Neanderthals.
- Ghost population
Ancient Ancestral South Indians
The deepest indigenous component of the Indian subcontinent, present in every population of the region today, has no ancient genome attached to it. The climate does not preserve DNA. It exists only as a parameter in a model.
- Ghost population
Unidentified African archaic
Between 2% and 19% of the genomes of West Africa derives from an archaic lineage that matches neither the Neanderthals nor the Denisovans, and for which no fossil exists. The range is wide because the estimate depends entirely on the demographic model assumed.
- Ghost population
Super-archaic in the Denisovan genome
The Denisova genome in its turn holds a contribution from a population more divergent still, perhaps tied to Homo erectus.
- Severed branch
Zlatý kůň and Ranis
The oldest modern human genomes in Europe, about 45,000 years old, tied to one another by distant family relationships. They represent the deepest split known from the line that left Africa. No descendants today.
- Severed branch
Ust'-Išim
A Siberian individual of 45,000 years ago, carrying Neanderthal segments that were still long, which made it possible to date the interbreeding. His lineage contributes to no living population.
- Severed branch
Ancient Beringians
Identified from the genome of a girl of 11,500 years ago in Alaska. Sister group to all other Native Americans, it splits off between 22,000 and 18,000 years ago. Extinct or absorbed, with no contribution detectable today.
- Severed branches
The failed exits from Africa
Misliya, Apidima, Al Wusta, Tam Pa Ling, Skhul and Qafzeh: populations of Homo sapiens outside Africa between 210,000 and 68,000 years ago, documented archaeologically and invisible genetically in people living today.
09 · Fact check
Nine things you often read that are wrong
Many school and popular summaries repeat claims that decades of literature have superseded. Each one here comes with the reference that refutes it.
“Human evolution is a ladder of progress”
It is a bush with many dead branches. Between 3 and 2.5 million years ago at least four hominin lines lived side by side in East Africa: early Homo, Paranthropus, Australopithecus garhi and the unnamed species from Ledi-Geraru described in 2025. In the same period A. africanus occupied South Africa.
The picture grew more complicated still in January 2026, with the first Paranthropus mandible from the Afar, more than 900 km further north than the known limit for the genus. For almost the whole history of the genus Homo the fossil record documents more than one human species at a time.
Villmoare et al., Nature 2025 · Alemseged et al., Nature 2026
“The first stone tools are the work of Homo habilis”
The tools from Lomekwi 3, in Kenya, are 3.3 million years old: half a million years before the oldest find that can be attributed to the genus Homo, which is the Ledi-Geraru mandible at 2.78 million years.
The Oldowan industry at Nyayanga, dated to 2.9 million years, is found with Paranthropus teeth. Stone knapping is not an invention of our genus and does not define it. The name “habilis”, coined in 1964 on that assumption, has stayed in use.
Harmand et al., Nature 2015 · Plummer et al., Science 2023
“The missing link is still missing”
The idea presupposes a linear chain with a hole in it. Evolution proceeds in mosaic fashion: every fossil combines derived and primitive traits in its own way, and none is the intermediate form between two points.
Ardipithecus ramidus has the pelvis of a biped and the opposable big toe of a climber. Homo naledi has an almost modern hand and an australopithecine brain.
White et al., Science 2009 · Berger et al., eLife 2015
“We are descended from chimpanzees”
We share with them a last common ancestor that lived between 6.5 and 9 million years ago, on molecular clock estimates, which vary with the mutation rate assumed. From that point on both lines have had the same amount of time to evolve.
The chimpanzee is not a backward version of us: knuckle-walking is a recent specialisation, probably not present in the common ancestor, as the anatomy of Ardi suggests.
Moorjani et al., PNAS 2016
“Homo sapiens appears 200,000 years ago at one point in Africa”
The Jebel Irhoud fossils, in Morocco, are dated to 315,000 ± 34,000 years. Florisbad, in South Africa, to 259,000. Omo I, in Ethiopia, was redated in 2022 to at least 233,000 years. They are three opposite corners of the continent.
The current model does not call for a single cradle: structured African populations, partly isolated and in intermittent exchange, in which modern traits assemble at different times and in different places. A genomic analysis published in April 2026 supports the contribution of several distinct ancestral populations.
Hublin et al., Nature 2017 · Vidal et al., Nature 2022
“Neanderthals were cognitively inferior and that is why they died out”
A mean cranial capacity of about 1,410 cm³, above that of sapiens today. They made birch tar for composite tools, which calls for controlled burning in a low-oxygen environment. They used pigments, ornaments and marine resources, and buried their dead.
Their disappearance, dated to between 41,000 and 39,000 years cal BP, fits a demographic model: small, fragmented populations, the climatic instability of marine isotope stage 3, competition for the same carrying capacity, and absorption through interbreeding. There is no evidence of any “physiological decline”.
Higham et al., Nature 2014 · Iasi et al., Science 2024
“The palaeo diet is the diet of our ancestors”
There is no single ancestral diet. Isotopic reconstructions, dental microwear and starch grains in calculus show enormous variation by latitude, by season and by group: from the Iberian Neanderthals who ate molluscs and plants to the historical Inuit who were almost entirely carnivorous.
The precise percentages in circulation, of the “80% of calories came from gathering” kind, cannot be generalised: they come from single recent ethnographic settings projected onto the past. And cereals, left out of commercial palaeo diets, were already being processed with grinding stones 30,000 years ago.
Henry et al., PNAS 2011 · Revedin et al., PNAS 2010
“Human races are biological groupings”
About 90% of human genetic variation lies within single local populations. The fraction attributable to differences between large continental groupings stays at around 10%. Two people born in the same village can differ from each other more than two continents differ on average.
Variation is clinal: it changes gradually with geographic distance, with no discrete boundaries. Any line drawn on a genetic map is a choice made by the analyst, not a fact of the data. And the deepest diversity between living populations does not separate the continents from one another: it separates two groups that are both African, with the Khoe-San populations diverging from all the rest 300,000-250,000 years ago.
The ancestry components read off genomic models are not races. They are coefficients that break a sample down against ancient reference genomes, and they change if the references change. Ethnic groups and nations are historical, linguistic and political categories, and they do not coincide with any of these components.
Lewontin, Evol Biol 1972 · Rosenberg et al., Science 2002 · Schlebusch et al., Science 2017
“Prehistory ends in 3200 BC”
That is a local convention, valid for Mesopotamia. Prehistory ends when written documentation appears, and that happens at wildly different moments: in Egypt around 3200 BC, in China around 1200 BC, in Italy between the eighth and the seventh century BC with the alphabet of Greek origin, in Australia only with European contact.
It is not a stage of development, it is a documentary condition. Technologically complex populations stay “prehistoric” for the sole reason that they produced no texts that survived, and their history is read through archaeology and genetics.
A convention of historiography, not an empirical fact
What these numbers do not say
Almost every date on this page carries a range, and several of those ranges are wider than the rounded figure suggests. The bars on the dating line show the known fossil record, that is the gap between the oldest and the youngest find attributed to a species. It is not how long the species actually lasted, which was almost certainly longer at both ends. Many taxonomic attributions are still argued over among specialists, and some of the species listed here may turn out to be variants of others. Dispersal routes are reconstructions: a few dozen dated sites scattered across whole continents carry lines that look like travelled itineraries once they are drawn on a map. Ancestry proportions are not races and not identities. They are coefficients of a statistical model that decomposes a present-day sample against ancient genomes chosen as references, and they shift when the references change; roughly ninety per cent of human genetic variation sits within single local populations, and geographic variation is gradual, with no boundary that can be drawn. The fossil sample itself is not random either. Remains survive where the sediment allows it and are found where people dig, so eastern Africa and western Europe are over-represented against tropical forests and much of Asia.
Methods and applicable ranges
- Radiocarbon (14C) Range: From roughly 300 to 55,000 years, and only on material holding organic carbon of biological origin: bone with surviving collagen, charcoal, wood, seeds. Past 50,000 years the residual 14C falls below the level at which the measurement separates from background, so the oldest dates should be read as minimum ages. Raw dates need calibrating against curves such as IntCal20, which is why they are written as cal BP. A small fraction of modern carbon entering the sample after burial makes it look younger, and the effect grows with age: near the limit of the method, very little contamination shifts a date by thousands of years.
- Uranium series (U-Th) Range: From 1,000 to 500,000 years, on carbonates that behaved as a closed system: stalagmites, calcite crusts, enamel and dentine. Laser ablation, introduced in 2024, allows very thin crusts to be sampled, including those covering wall paintings. A crust over a painting dates the crust, not the painting: it gives a minimum age for whatever lies beneath. If uranium migrated into or out of the carbonate after it formed, the system was not closed and the date does not hold. That is the point on which the challenge to the Neanderthal dates for Iberian cave art turns.
- Argon-argon (40Ar/39Ar) Range: From about 10,000 years to billions, on minerals from volcanic layers. It dates the eruption, not the fossil: a find sits between the tephra below it and the tephra above, and the age reported is that of the bracket enclosing it. This is the backbone of the East African chronology. Where there are no volcanic layers the method is simply unavailable, which is the main reason South African caves carry looser chronologies than the Rift Valley. An older crystal reworked into the tephra can also make a date too old, and has to be spotted grain by grain.
- Luminescence (OSL and TL) Range: From 100 to 500,000 years. It measures the energy stored in quartz and feldspar since the last exposure to light or heat, so it dates the sediment containing a find, and flint heated in hearths. It holds on the condition that find and sediment were laid down together. Animal burrows, washing and later reworking break that condition and mix grains of different ages. Single-grain analysis exists to spot those mixed populations, and it is the reason some chronologies have been revised substantially, as happened at Liang Bua.
- Electron spin resonance Range: From 10,000 to 2 million years, on tooth enamel. Useful where there is no datable organic material and no volcanic layer, which covers a good share of Asian and South African contexts. The age depends on the model assumed for how the tooth took up uranium during burial. Different models, early uptake or linear uptake, return ages that can differ appreciably, and papers report more than one estimate for exactly that reason.
- Palaeomagnetism Range: Relative dating, not absolute. It reads reversals of the Earth's magnetic field recorded in sediments and matches them to the global polarity timescale, so it yields an age only when the local sequence is long enough to be tied to that scale without ambiguity. Polarity boundaries are few and far apart, so the method places a deposit within an interval rather than on a date. In 2026 it was used this way for the Thomas Quarry remains, put at about 773,000 years, which is the Matuyama to Brunhes transition.
- Molecular clock Range: It estimates when two lineages separated, not when a single individual lived. It works on comparisons between genomes and returns divergence times; applying it requires an assumed mutation rate and generation interval, and those assumptions set the scale. Change the assumed rate and the dates move by millions of years on deep divergences and by tens of thousands on recent ones. Estimates are therefore published as ranges and should be read as ranges: the split between our lineage and the chimpanzee lineage is the best known case.
- Palaeoproteomics Range: Identification, not dating. It reads protein sequences surviving in enamel and bone where DNA has degraded away, allowing a find to be assigned to a group or an individual to be sexed. It works over timespans and in climates where ancient DNA can no longer be recovered. Surviving proteins are few and vary less than the genome, so they resolve relationships between groups in far less detail. In 2025 they placed the Harbin cranium within the Denisovan group, an attribution later confirmed independently by mitochondrial DNA from its dental calculus.
Primary references
- Lewontin, R. C. (1972) The Apportionment of Human Diversity Evolutionary Biology, 6, 381-398 doi:10.1007/978-1-4684-9063-3_14 ↗
- Rosenberg, N. A., Pritchard, J. K., Weber, J. L., et al. (2002) Genetic Structure of Human Populations Science, 298(5602), 2381-2385 doi:10.1126/science.1078311 ↗
- Allentoft, M. E., Sikora, M., Refoyo-Martínez, A., et al. (2024) Population genomics of post-glacial western Eurasia Nature, 625, 301-311 doi:10.1038/s41586-023-06865-0 ↗
- Haak, W., Lazaridis, I., Patterson, N., et al. (2015) Massive migration from the steppe was a source for Indo-European languages in Europe Nature, 522, 207-211 doi:10.1038/nature14317 ↗
- Olalde, I., Brace, S., Allentoft, M. E., et al. (2018) The Beaker phenomenon and the genomic transformation of northwest Europe Nature, 555, 190-196 doi:10.1038/nature25738 ↗
- Narasimhan, V. M., Patterson, N., Moorjani, P., et al. (2019) The formation of human populations in South and Central Asia Science, 365(6457), eaat7487 doi:10.1126/science.aat7487 ↗
- Fortes-Lima, C. A., Burgarella, C., Hammarén, R., et al. (2024) The genetic legacy of the expansion of Bantu-speaking peoples in Africa Nature, 625, 540-547 doi:10.1038/s41586-023-06770-6 ↗
- Cooke, N. P., Mattiangeli, V., Cassidy, L. M., et al. (2021) Ancient genomics reveals tripartite origins of Japanese populations Science Advances, 7(38), eabh2419 doi:10.1126/sciadv.abh2419 ↗
- Moreno-Mayar, J. V., Potter, B. A., Vinner, L., et al. (2018) Terminal Pleistocene Alaskan genome reveals first founding population of Native Americans Nature, 553, 203-207 doi:10.1038/nature25173 ↗
- Raghavan, M., Skoglund, P., Graf, K. E., et al. (2014) Upper Palaeolithic Siberian genome reveals dual ancestry of Native Americans Nature, 505, 87-91 doi:10.1038/nature12736 ↗
- Durvasula, A., Sankararaman, S. (2020) Recovering signals of ghost archaic introgression in African populations Science Advances, 6(7), eaax5097 doi:10.1126/sciadv.aax5097 ↗
- Skoglund, P., Thompson, J. C., Prendergast, M. E., et al. (2017) Reconstructing Prehistoric African Population Structure Cell, 171(1), 59-71.e21 doi:10.1016/j.cell.2017.08.049 ↗
- Schlebusch, C. M., Malmström, H., Günther, T., et al. (2017) Southern African ancient genomes estimate modern human divergence to 350,000 to 260,000 years ago Science, 358(6363), 652-655 doi:10.1126/science.aao6266 ↗
- Lazaridis, I., Nadel, D., Rollefson, G., et al. (2016) Genomic insights into the origin of farming in the ancient Near East Nature, 536, 419-424 doi:10.1038/nature19310 ↗
- Skoglund, P., Posth, C., Sirak, K., et al. (2016) Genomic insights into the peopling of the Southwest Pacific Nature, 538, 510-513 doi:10.1038/nature19844 ↗
- Jacobs, G. S., Hudjashov, G., Saag, L., et al. (2019) Multiple Deeply Divergent Denisovan Ancestries in Papuans Cell, 177(4), 1010-1021.e32 doi:10.1016/j.cell.2019.02.035 ↗
- Llorente, M. G., Jones, E. R., Eriksson, A., et al. (2015) Ancient Ethiopian genome reveals extensive Eurasian admixture in Eastern Africa Science, 350(6262), 820-822 doi:10.1126/science.aad2879 ↗
- Posth, C., Yu, H., Ghalichi, A., et al. (2023) Palaeogenomics of Upper Palaeolithic to Neolithic European hunter-gatherers Nature, 615, 117-126 doi:10.1038/s41586-023-05726-0 ↗
- Villmoare, B., Kimbel, W. H., Seyoum, C., et al. (2025) New discoveries of Australopithecus and Homo from Ledi-Geraru, Ethiopia Nature, 650, 374-380 (online 13 agosto 2025) doi:10.1038/s41586-025-09390-4 ↗
- Alemseged, Z., Wynn, J. G., Geraads, D., et al. (2026) Afar fossil shows broad distribution and versatility of Paranthropus Nature, 650, 381-388 doi:10.1038/s41586-025-09826-x ↗
- Williams, S. A., Prang, T. C., Meyer, M. R., et al. (2026) Earliest evidence of hominin bipedalism in Sahelanthropus tchadensis Science Advances, 12(1), eadv0130 doi:10.1126/sciadv.adv0130 ↗
- Fu, Q., et al. (2025) The proteome of the late Middle Pleistocene Harbin individual Science, 389(6761), 704-707 doi:10.1126/science.adu9677 ↗
- Fu, Q., et al. (2025) Denisovan mitochondrial DNA from dental calculus of the >146,000-year-old Harbin cranium Cell, 188(15), 3919-3926.e9 doi:10.1016/j.cell.2025.05.040 ↗
- Sümer, A. P., Rougier, H., Villalba-Mouco, V., et al. (2025) Earliest modern human genomes constrain timing of Neanderthal admixture Nature, 638, 711-717 doi:10.1038/s41586-024-08420-x ↗
- Iasi, L. N. M., Chintalapati, M., Skov, L., et al. (2024) Neanderthal ancestry through time: Insights from genomes of ancient and present-day humans Science, 386(6727), eadq3010 doi:10.1126/science.adq3010 ↗
- Oktaviana, A. A., Joannes-Boyau, R., Hakim, B., et al. (2024) Narrative cave art in Indonesia by 51,200 years ago Nature, 631, 814-818 doi:10.1038/s41586-024-07541-7 ↗
- Holliday, V. T., Windingstad, J. D., Mayer, J. H., et al. (2025) Paleolake geochronology supports Last Glacial Maximum (LGM) age for human tracks at White Sands, New Mexico Science Advances, 11(25), eadv4951 doi:10.1126/sciadv.adv4951 ↗
- Hublin, J.-J., Ben-Ncer, A., Bailey, S. E., et al. (2017) New fossils from Jebel Irhoud, Morocco and the pan-African origin of Homo sapiens Nature, 546, 289-292 doi:10.1038/nature22336 ↗
- Vidal, C. M., Lane, C. S., Asrat, A., et al. (2022) Age of the oldest known Homo sapiens from eastern Africa Nature, 601, 579-583 doi:10.1038/s41586-021-04275-8 ↗
- Harmand, S., Lewis, J. E., Feibel, C. S., et al. (2015) 3.3-million-year-old stone tools from Lomekwi 3, West Turkana, Kenya Nature, 521, 310-315 doi:10.1038/nature14464 ↗
- Villmoare, B., Kimbel, W. H., Seyoum, C., et al. (2015) Early Homo at 2.8 Ma from Ledi-Geraru, Afar, Ethiopia Science, 347(6228), 1352-1355 doi:10.1126/science.aaa1343 ↗
- Plummer, T. W., Oliver, J. S., Finestone, E. M., et al. (2023) Expanded geographic distribution and dietary strategies of the earliest Oldowan hominins and Paranthropus Science, 379(6632), 561-566 doi:10.1126/science.abo7452 ↗
- Higham, T., Douka, K., Wood, R., et al. (2014) The timing and spatiotemporal patterning of Neanderthal disappearance Nature, 512, 306-309 doi:10.1038/nature13621 ↗
- Mylopotamitaki, D., Weiss, M., Fewlass, H., et al. (2024) Homo sapiens reached the higher latitudes of Europe by 45,000 years ago Nature, 626, 341-346 doi:10.1038/s41586-023-06923-7 ↗
- Clarkson, C., Jacobs, Z., Marwick, B., et al. (2017) Human occupation of northern Australia by 65,000 years ago Nature, 547, 306-310 doi:10.1038/nature22968 ↗
- Henshilwood, C. S., d'Errico, F., van Niekerk, K. L., et al. (2018) An abstract drawing from the 73,000-year-old levels at Blombos Cave, South Africa Nature, 562, 115-118 doi:10.1038/s41586-018-0514-3 ↗
- Sehasseh, E. M., Fernandez, P., Kuhn, S., et al. (2021) Early Middle Stone Age personal ornaments from Bizmoune Cave, Essaouira, Morocco Science Advances, 7(39), eabi8620 doi:10.1126/sciadv.abi8620 ↗
- Détroit, F., Mijares, A. S., Corny, J., et al. (2019) A new species of Homo from the Late Pleistocene of the Philippines Nature, 568, 181-186 doi:10.1038/s41586-019-1067-9 ↗
- Sutikna, T., Tocheri, M. W., Morwood, M. J., et al. (2016) Revised stratigraphy and chronology for Homo floresiensis at Liang Bua in Indonesia Nature, 532, 366-369 doi:10.1038/nature17179 ↗
- Berger, L. R., Hawks, J., de Ruiter, D. J., et al. (2015) Homo naledi, a new species of the genus Homo from the Dinaledi Chamber, South Africa eLife, 4, e09560 doi:10.7554/eLife.09560 ↗
- Zhu, Z., Dennell, R., Huang, W., et al. (2018) Hominin occupation of the Chinese Loess Plateau since about 2.1 million years ago Nature, 559, 608-612 doi:10.1038/s41586-018-0299-4 ↗
- Hoffmann, D. L., Standish, C. D., García-Diez, M., et al. (2018) U-Th dating of carbonate crusts reveals Neandertal origin of Iberian cave art Science, 359(6378), 912-915 doi:10.1126/science.aap7778 ↗
Data provenance
- Natural Earth, vettoriali 110m ↗ Licence: Public domain, in every raster and vector version. No permission is required to use it, and crediting the authors is not obligatory. The authors disclaim responsibility for accuracy, content and use of the data; political boundaries imply no recognition. Made with Natural Earth (the short form suggested by the authors, not obligatory).