Appearance
3.5 — Human Evolution
For most of the last few million years there were several kinds of human alive at the same time. As recently as 50,000 years ago there were at least four: our own species, Neanderthals in Europe and western Asia, Denisovans in eastern Asia, and the very small Homo floresiensis on the Indonesian island of Flores. There is decent evidence that Homo luzonensis was in the Philippines too.
Being the only human species on Earth is a recent and unusual situation, and it is worth starting there, because the mental picture most people carry — a single line marching from ape to man — is wrong in the way that matters most. It was a bush, and every other branch is gone.
The split from other apes
Genetic comparison puts the human–chimpanzee divergence at roughly 6 to 8 million years ago, with real uncertainty for the reasons in Chapter 3.4. Chimpanzees and bonobos are our closest living relatives; gorillas branched off earlier, around 8 to 9 million years ago, and orangutans earlier still.
We did not descend from chimpanzees. Chimpanzees have been evolving for exactly as long as we have. The common ancestor was neither, and probably resembled neither closely.
The genetic difference is about 1.2 percent in protein-coding sequence, rising to around 4 percent if insertions and deletions are counted. That is a small number carrying a very large phenotypic difference, and the resolution is that most of the difference is regulatory rather than in the proteins themselves — when genes are switched on, in which tissue, and for how long (Chapter 2.4). Human and chimpanzee brains are built largely from the same proteins on a different schedule.
Hominin is the word for humans and all species closer to us than to chimpanzees.
Walking came first
The old assumption was that a big brain came first and the rest followed. The fossil record shows the exact opposite, and the evidence is not subtle.
Sahelanthropus (around 7 million years ago, Chad) and Ardipithecus (around 4.4 million years ago, Ethiopia) both show signs of upright walking with a brain no larger than a chimpanzee's.
Australopithecus afarensis — of which "Lucy" is the famous specimen, found in Ethiopia in 1974 — lived 3 to 4 million years ago with a brain of about 400 to 500 cm³, roughly a third of ours, and a skeleton unambiguously adapted for walking upright.
And there is direct evidence that does not depend on interpreting bones. At Laetoli in Tanzania, a volcanic ash fall about 3.66 million years ago was rained on, then walked across, then buried. The Laetoli footprints show a modern-looking human gait: a deep heel strike, a transfer of weight along the outside of the foot, and a push-off from a big toe that is in line with the others rather than sticking out sideways like a chimpanzee's grasping toe. That is not an ape's foot, and the brain that owned it was ape-sized.
Why walking upright? Several hypotheses, none proven, several probably true together: it frees the hands to carry food and tools; it is more energy-efficient over long distances, though less efficient for sprinting; it raises the eyes above tall grass; it reduces the body surface exposed to overhead sun at midday and increases exposure to wind. What is clear is that it came first and cost a great deal — Chapter 3.6 covers the bill.
The genus Homo
Homo habilis, around 2.4 to 1.4 million years ago, with a brain of 550 to 690 cm³ and an association with the first stone tools, which is where the name — handy man — came from.
Homo erectus, from about 1.9 million years ago, brain 850 to 1,100 cm³. This is the first hominin to leave Africa, reaching Georgia by 1.8 million years ago and Java and China soon after. It persisted for well over a million years — far longer than our own species has existed — and it is the species associated with the controlled use of fire.
Fire matters more than it sounds. Cooking gelatinises starch and denatures protein (Chapter 1.3), which makes food far easier to digest and releases substantially more usable energy per mouthful. Richard Wrangham's argument is that this is what paid for the brain: human brains use about 20 percent of resting energy, which is an enormous share, and the human gut is unusually small for a primate of our size. Cooking outsources part of digestion, letting the gut shrink and the brain grow. The evidence for when fire was controlled is contested — good evidence exists from around 1 million years ago and disputed evidence earlier — but the logic is well supported.
Homo heidelbergensis, roughly 700,000 to 200,000 years ago, brain around 1,100 to 1,400 cm³, generally regarded as ancestral to both Neanderthals and us.
Homo sapiens. The oldest fossils currently assigned to our species are from Jebel Irhoud in Morocco, about 300,000 years old — a date that pushed the origin back by roughly 100,000 years when published in 2017, and that also moved the geography, since the finds are in North Africa rather than East Africa. The current picture is not a single origin point but a pan-African process, with populations across the continent exchanging genes and contributing to what became us.
Neanderthals, and what we got from them
Neanderthals were not brutish failures. Their brains were on average slightly larger than ours. They made complex tools, controlled fire, hunted large game cooperatively, used pigments, cared for injured group members over long periods — several skeletons show healed injuries that would have made the individual dependent on others for years — and buried at least some of their dead. They lived through European ice ages that would kill an unequipped modern human in a day.
They disappeared around 40,000 years ago, and why is unresolved. Competition with expanding modern humans, climate instability, small population sizes and inbreeding, and absorption by interbreeding all likely contributed. "Extinction" is partly the wrong word, because of what the genetics showed.
In 2010 the Neanderthal genome was sequenced, and the comparison was decisive: people of non-African ancestry carry roughly 1.5 to 2 percent Neanderthal DNA. Africans carry very little, and what they do carry appears to come from back-migration. Svante Pääbo received the 2022 Nobel Prize in Medicine for this work.
Denisovans are stranger still. They were identified in 2010 from a single finger bone in a Siberian cave, sequenced before anyone knew what the species looked like. A whole human population was discovered from DNA alone. Melanesians and Aboriginal Australians carry 3 to 5 percent Denisovan ancestry.
And the inherited sequences are not inert. Several have measurable effects.
Tibetans carry a Denisovan version of a gene called EPAS1 that regulates the response to low oxygen. Most people at high altitude respond by thickening the blood, which raises viscosity and causes problems; the Denisovan variant produces a different response that avoids this. It is present in about 80 percent of Tibetans and almost nobody else, and it is one of the strongest signals of recent selection in any human population — a gene inherited from an extinct human species that made living at 4,000 metres possible.
Neanderthal variants affect skin, hair, immune function and sleep timing. A Neanderthal-derived region on chromosome 3 was found to be associated with more severe COVID-19 outcomes, and a different Neanderthal segment with protection — a 40,000-year-old inheritance showing up in a 2020 pandemic.
Out of Africa, and what "race" means genetically
Modern humans expanded out of Africa in a major wave around 60,000 to 70,000 years ago, though earlier excursions occurred and left fossils in the Levant. That founding group was small — the bottleneck of Chapter 3.3 — which is why all non-African populations together carry less genetic diversity than the populations within Africa.
The route ran through the Middle East, then east along the coast to South Asia, reaching Australia by around 65,000 years ago, and into Europe around 45,000 years ago. The Americas were reached via Beringia, with dates still under active revision but at least 15,000 years ago and possibly considerably earlier.
The genetics of human variation, stated plainly.
About 85 percent of all human genetic variation exists within any single local population. Only around 15 percent distinguishes continental groups. Two randomly chosen people from the same village typically differ almost as much as two people from different continents.
Variation is clinal, not categorical — it changes gradually with distance, with no sharp boundaries anywhere, because humans have moved and interbred continuously. The apparent boundaries in traditional racial categories mostly reflect which populations happened to be sampled, and oceans and deserts that made travel harder.
Visible traits are unrepresentative. Skin colour tracks ultraviolet intensity by latitude and has evolved lighter and darker independently several times — the light skin of northern Europeans and of northern East Asians arises from different genes. Skin colour therefore tells you almost nothing about the rest of someone's genome, which is the biological reason race is a poor proxy for anything medical.
None of this means population differences are irrelevant to medicine. They are real and sometimes important — the drug metabolism and carrier-frequency examples in Chapter 3.3 are genuine. The distinction is that those are specific, testable genetic facts about particular variants, not properties of a racial category, and medicine is steadily replacing ancestry as a proxy with direct testing of the variant in question.
Recent evolution: we have not stopped
Lactase persistence arose independently at least four times — once in Europe, at least three times in Africa, each with a different mutation — in the last 10,000 years, following the domestication of cattle. This is one of the strongest signals of recent selection in the human genome, and it is a direct response to a cultural change (Chapter 1.2).
Malaria resistance alleles — sickle, thalassaemia, G6PD deficiency, Duffy negativity — all arose within the last several thousand years, following the spread of agriculture, which created the standing water and dense settlement that malaria needs.
High-altitude adaptation in Tibetans, Andeans and Ethiopians, by three different genetic routes.
Adult height and the timing of puberty show measurable selection signals in historical data.
And selection pressures now are not the ones people assume. Modern medicine has largely removed selection for surviving infection in wealthy populations. What remains is selection acting through fertility — how many children people have, and at what age — which is a very different filter and one that operates on behaviour and physiology in ways nobody can currently predict.
Where this matters clinically
Evolutionary medicine reframes several conditions as mismatch problems. Human physiology was tuned by an environment of intermittent food, high physical activity, constant infectious challenge, and no refined sugar. Type 2 diabetes, obesity, hypertension and myopia are all far commoner now than in populations living traditionally, and the difference is environmental, not genetic. The genes did not change; the world did.
Salt handling is the clearest example. Ancestral human diets were low in sodium, and the kidney evolved to retain it aggressively (Chapter 10.3). In an environment of abundant salt, that same machinery drives blood pressure up. Populations with the strongest salt-retention adaptations have the highest hypertension rates when they adopt a modern diet.
And the hygiene hypothesis proposes that an immune system tuned by constant parasitic and microbial exposure misbehaves without it, contributing to the rise in allergy and autoimmune disease. The evidence is suggestive rather than settled, and Chapter 13.6 gives it a fair hearing rather than a headline.
What the next page fixes
Standing up freed our hands and cost us a great deal. The human body carries a long list of arrangements that no engineer would choose, all of them explicable as modifications of something that was doing a different job. Chapter 3.6 goes through the bad design in detail — the back, the birth canal, the eye's backwards retina, the shared airway and food pipe, the nerve that goes the wrong way — and answers the question the reader will already have asked: why is a body that survived four billion years so easy to kill with a blade or a bullet.