A diet-centered history of human evolution, from 60-million-year-old fruit-eaters to the present, and why “we come from carnivores” doesn’t survive contact with the evidence.

Stand in the Hall of the Bulls at Lascaux, in southwestern France, and you are surrounded by animals. Horses gallop across the ceiling by the hundreds. Aurochs, bison, stags, a bear, a rhino, a handful of big cats. It is easy to come away thinking you have just seen a 17,000-year-old menu.

You haven’t. The people who painted Lascaux lived on reindeer. At Upper Paleolithic sites across this region, reindeer bones make up something like 80 to 90 percent of the food bones archaeologists dig up. At Lascaux itself, reindeer are almost entirely absent from the walls (depending on who is counting, there is either one or none). The painters even used reindeer antler to grind their pigments. The animal that fed them is present as a tool and missing as a subject.1

Hold onto that paradox, because it captures the single most important idea in this whole story: what survives is not the same as what was eaten. Bones fossilize. Stone tools last forever. Paintings of horses endure on cave walls. Fruit, tubers, leaves, nuts, honey, and greens rot within days. For most of the history of archaeology, the record we could actually read was a record of meat, not because meat dominated the diet but because meat leaves durable garbage.

And every time scientists have invented a new way to see the plants, they have found them. In the tartar on Neanderthal teeth. In the starch grains clinging to 30,000-year-old grinding stones. In 50,000-year-old fossilized feces. In the nitrogen locked inside 3.5-million-year-old tooth enamel. In the genes we carry today.

So this is a post about how humans evolved, told through the food we ate and how we know. It goes back as far as the evidence allows, to the first fruit-eating primates, through our cousins the chimpanzees and bonobos, through the upright australopiths and the fire-wielding Homo erectus, through Neanderthals and Ice Age hunters and the first farmers, all the way to the Hadza collecting honey today. Along the way it will make an argument that I think the evidence supports overwhelmingly: we did not “come from carnivores.” We came from fruit-eaters, and we became the most flexible omnivore the planet has ever produced. Meat mattered. Fat and marrow mattered. But so did roots and nuts and honey and cooked starch, at every step, and our bodies say so.

I’ll try to be honest about the places where meat genuinely mattered, because a debunk that overstates its case is not a debunk, it’s just a different myth. And before the story starts, we need to deal with a deeper problem that both sides of the diet wars tend to skip: even if we knew exactly what our ancestors ate, would that tell us what is healthiest for us today? It wouldn’t, and understanding why changes the whole conversation.

A note before we start: evolution is a matchmaker, not a doctor

The carnivore argument, the paleo argument, and, to be fair, the mirror-image vegan argument all share a hidden assumption: that whatever our ancestors evolved eating must be what is best for our health. It feels intuitive. It is also wrong, and the reason is one of the most important and least appreciated ideas in biology.

Natural selection grades on one exam: did you survive long enough to reproduce, and did your offspring do the same? Everything that happens to your body after that is, from evolution’s point of view, extra credit it cannot see. In the 1950s Peter Medawar and George Williams worked out the consequence. Because the force of selection fades with age and effectively vanishes once reproduction is over, a gene that helps you in your twenties can be favored even if it wrecks you in your sixties.2 Williams called this antagonistic pleiotropy: one gene, two effects, pulling in opposite directions at different ages. Thomas Kirkwood later framed the same trade-off as the “disposable soma”: the body invests in reproduction first and repairs itself only as much as is needed to get there.3

For decades this was theory. It isn’t anymore. Williams laid out nine testable predictions in 1957; a 2017 reappraisal judged six of them supported by the evidence.4 A 2018 review concluded that such trade-offs are common, perhaps ubiquitous.5 And it is written in our own genomes. In 2017 researchers found that the gene variants that raise the risk of coronary artery disease, the leading cause of death on Earth, are significantly enriched for lifetime reproductive success: the same variants that clog arteries late in life are linked to having more children early in life.6 In 2023 a study of 276,406 people in the UK Biobank found a strong negative genetic correlation between reproduction and lifespan: variants that favor reproduction tend to shorten life.7 A 2026 analysis of 62 diseases went further, finding that disease-risk alleles are, on average, tied to higher fertility and lower longevity and appear to have been favored by selection over the last fifty thousand years.8

You have already met one of the cleanest examples in this post’s own story. Around 17 million years ago our ape ancestors lost the enzyme uricase, after a gradual decline in its activity over millions of years.9 The leading hypothesis is that this helped fruit-eating apes convert fruit sugar into stored fat during lean seasons, a survival advantage in a world of famine. In a world of abundance, the same loss leaves us with high uric acid: today 3.9 percent of American adults, about 9.2 million people, have gout, and roughly one in five have elevated uric acid.10 A fruit adaptation that helped our ancestors reproduce now hurts us in the decades evolution never optimized.

The costs are not hypothetical, and the clearest examples are the very adaptations usually held up as proof that a population was built for its diet. Arctic peoples carry a variant of CPT1A, the enzyme that feeds long-chain fats into the mitochondria, at frequencies near 70 percent and almost nowhere else on Earth. Selection drove it hard. It is also associated with impaired fasting ketogenesis, hypoglycemia in infants, and elevated infant mortality.11, 12 The Greenlandic Inuit fat-processing variants come with their own price, on the order of two centimeters of height and four kilograms of weight.13 These populations really were adapted to their diets. The adaptations were also expensive. Selection is not a quality-control department; it is a filter that keeps whatever reproduces, side effects included.

A caution about how far this idea can be pushed. Antagonistic pleiotropy is a framework, not a finding, and it cannot be used to prove that any particular food is bad for you. That is not its job. Its job is to dismantle a single move: humans evolved eating X, therefore X is good for us. Evolution has no stake in your seventh decade. Whatever our ancestors ate got them to reproduction in the environments they lived in. That is the entire claim the evidence supports, in any direction.

So here is the logical problem with “we evolved to eat it, so it’s good for us.” Evolution did not design a diet for a long, healthy life. It settled for whatever got enough of us to the next generation. “Our ancestors ate this” and “our ancestors could eat this” are historical facts. “This is optimal for a 75-year-old today” is a medical claim, and the first two do not deliver the third. Philosophers call this the appeal to nature. Biologists call it forgetting the selection shadow.

To be fair, evolutionary thinking is not useless here. It is very good at generating hypotheses about mismatch: the idea that ultra-processed food, refined sugar, and sedentary life sit badly with bodies shaped for a different world.14 The uricase story is itself a mismatch story. What evolution cannot do is tell you the optimum. Only outcome data on real, living, aging humans can do that, and that evidence is now substantial. In a 2025 analysis of 105,015 people followed for up to 30 years, only 9.3 percent reached age 70 free of major chronic disease with intact memory, mobility, and mental health, and every one of eight healthy dietary patterns raised the odds of getting there. The strongest, the Alternative Healthy Eating Index, was associated with 86 percent greater odds at 70 and more than double the odds at 75. In fairness, the index that most strictly favored plant foods over animal foods was the weakest of the eight, a reminder that food quality matters more than category: the winning patterns were rich in vegetables, fruit, whole grains, nuts, legumes, and healthy fats, low in red and processed meat and sugary drinks, with modest amounts of healthy animal foods.15 In three cohorts totaling 206,404 people, those who most closely followed the plant-forward “planetary health diet” had 23 percent lower mortality than those who followed it least.16 And in 2015 the World Health Organization’s cancer agency classified processed meat as a Group 1 carcinogen and red meat as probably carcinogenic, a classification of hazard rather than of how big the risk is, but not one it has ever applied to lentils.17

Hold this frame for the rest of the post. The evolutionary story that follows is fascinating and it demolishes the carnivore myth on its own terms. But its deepest lesson is not “eat like a caveman.” It is that we are the first ape flexible enough to choose its diet on purpose, and the first with the science to choose well.

Part 1: How we know what we know

Before the story, a word about the evidence, because a scientifically minded reader deserves to know where each claim comes from, and a lay reader deserves to know why “we found bones” isn’t the end of the argument.

Paleoanthropologists reconstruct ancient diets from roughly a dozen independent lines of evidence. The beauty of this is that they don’t share the same weaknesses, so where they agree, we can be confident.

Comparative anatomy. Living primates give us a baseline. If every one of our closest relatives is a plant-eater with a particular kind of gut and teeth, and we share those features, that tells us something about the ancestor we all descend from.

Fossils. Teeth are the hardest thing in the body and survive best. Their size, shape, enamel thickness, and cusp patterns are tuned to diet over evolutionary time. Jaws, rib cages, and pelves tell us about chewing muscles and gut size. Skulls tell us brain size.

Dental microwear. Under a microscope, the surface of a tooth carries scratches and pits left by food in the last weeks of an animal’s life. Hard seeds pit; tough leaves scratch. Peter Ungar and colleagues have used this to overturn some famous assumptions, as we’ll see.18

Stable isotopes. This is the technique carnivore-diet advocates lean on hardest, so it’s worth understanding. Atoms of carbon and nitrogen come in slightly different weights (isotopes). Plants that photosynthesize with the “C3” pathway (most trees, shrubs, fruits) have a different carbon signature than “C4” plants (tropical grasses and sedges). Animals carry the signature of what they ate. Nitrogen isotopes get heavier as you go up the food chain, so a wolf has higher nitrogen than a deer, which is higher than grass. Measure the collagen in a fossil bone or the enamel of a tooth and you can estimate where the animal sat in the food web.19

But isotopes have two limitations that matter enormously here. First, bone collagen only survives about 100,000 to 200,000 years, so classic isotope work has been concentrated on Neanderthals and Ice Age humans, mostly in cold Europe where plants were scarce. Second, and this is the big one, nitrogen isotopes in collagen mostly track where your protein came from. They are largely blind to the calories from fruit sugar, tuber starch, honey, or fat. A person who ate a moderate amount of meat plus a huge amount of cooked tubers would look, isotopically, a lot like a meat-eater. Any honest use of isotope data has to carry that caveat.20

Dental calculus. Tartar, the hardened plaque a dentist scrapes off your teeth, is a time capsule. It traps starch grains, plant microfossils called phytoliths, proteins, and DNA from food and from the mouth’s microbes. Christina Warinner, Laura Weyrich, Amanda Henry and others have pulled Neanderthal meals out of it.21, 22

Coprolites. Fossilized or desiccated feces. They preserve pollen, phytoliths, starch, seeds, fibers, parasite eggs, DNA, and chemical biomarkers: coprostanol from digested animal cholesterol versus stigmastanol from plant sterols. If you want to know what someone actually ate, as opposed to what they painted, there is nothing better than their poop.23

Stone tool residues and use-wear. Starch grains and plant polish on a tool edge, versus the wear from slicing meat or scraping hide.

Cut marks and taphonomy. How bones were broken and marked distinguishes human butchery from hyena gnawing, and reveals marrow extraction.

Archaeobotany. Charred seeds, nuts, and tubers recovered by floating sediment in water. Plants preserve when burned, waterlogged, or desiccated, and almost never otherwise, which is exactly why the record is biased.

Genetics. The genes for color vision, vitamin C synthesis, starch digestion, lactose digestion, fat metabolism, jaw muscles, and taste all carry a history.

Ethnography. Living hunter-gatherers, used carefully, as analogies rather than fossils.

Physiology. Some diets are simply impossible for a human body, and knowing the limits constrains the past.

Keep this toolkit in mind. When one line of evidence points one way and eleven point the other, the eleven win.

Part 2: The fruit in our genes (65 to 25 million years ago)

Our story starts long before anything you’d call human, with small, big-eyed mammals scampering through the canopy of forests that had just been transformed by a new kind of plant.

Flowering plants, angiosperms, spread across the globe in the late Cretaceous and exploded in the Paleocene. They came with an innovation: fleshy fruit, sugar wrapped around seeds, a bribe to get animals to carry those seeds away. The earliest primates appear in the fossil record right around this time, roughly 65 to 55 million years ago, and one long-standing hypothesis is that primates and fruiting angiosperms co-evolved. Grasping hands, forward-facing eyes, and good color vision are exactly what you’d want for finding and picking ripe fruit at the ends of thin branches.24

Whether or not that co-evolution story is exactly right (a competing idea emphasizes hunting insects among the leaves), we carry three pieces of genetic evidence that our lineage was steeped in fruit for a very long time.

We can’t make vitamin C

Nearly every mammal synthesizes its own vitamin C. Your dog does. A cow does. We don’t, and neither does any monkey or ape. The gene responsible, GULO (L-gulonolactone oxidase), is broken in our lineage. Estimates place its inactivation around 61 million years ago, in the common ancestor of the anthropoid primates.25, 26

Think about what that means. A mutation that would cause fatal scurvy in most mammals persisted in our ancestors without being weeded out by natural selection. That is only possible if the diet was so reliably saturated with vitamin C that making your own was pointless. In other words, fruit and fresh plant food. Tellingly, the same gene was lost independently in fruit bats and guinea pigs, two other lineages that live on plants. A broken GULO gene is a fingerprint left on our genome by tens of millions of years of eating fruit.26

One honest complication, because it matters later. Raw animal tissue is not entirely devoid of vitamin C. Arctic peoples avoided scurvy without fruit, and the reason is that they ate organs and muktuk, the skin and blubber of whales, raw and in quantity; cooking destroys the vitamin. That is a real workaround, and it is worth noticing what it required: eating whole animals raw, in a place where nothing else grew. It is the exception that shows the rule. A lineage can only afford to break the vitamin C gene if its food supplies the vitamin without fail, and for the fifty million years before anyone reached the Arctic, that food was plants.

We see red

Most mammals see the world in two colors. Old World monkeys and apes, including us, see in three, because an ancestral gene for a light-sensing opsin duplicated on the X chromosome roughly 30 to 40 million years ago and the copies diverged, giving us separate red and green receptors.27

Why? The leading explanation is that trichromacy lets you spot a ripe reddish or yellow fruit against a wall of green leaves, and possibly young, nutritious reddish leaves too. In South American monkeys, where color vision varies between individuals, trichromatic animals are better at finding ripe fruit. The story isn’t perfectly tidy, since some leaf-eating monkeys are also trichromatic, so “young red leaves” may be part of the answer. But whatever the exact selective pressure, our color vision is a plant-foraging adaptation. Cats, wolves, and hyenas, actual carnivores, are dichromats. They don’t need to tell a ripe fig from a green one.27

We taste sweet, and we handle fructose strangely

Primates have a functional sweet receptor tuned to sugars, and a battery of bitter receptors (the TAS2R family) that warn against plant toxins. Cats have lost their sweet receptor entirely. An animal that never eats sugar has no use for it. We kept ours.28

There’s a more speculative chapter here too. Apes lost a working copy of the uricase enzyme around 15 to 20 million years ago. Richard Johnson and colleagues have proposed that this made our ancestors better at converting fruit fructose into stored fat during the seasonal famines of the cooling Miocene, a survival advantage then and a metabolic liability now.9 It’s a hypothesis, not a settled finding, but the loss itself is real and it fits a lineage eating a lot of fruit.

Teeth and guts of a fruit-eater

Frugivore teeth have low, rounded molar cusps for crushing soft fruit and broad incisors for scraping flesh from a rind (look at a chimpanzee’s front teeth). Carnivore teeth are blades: carnassials that shear meat, with canines for gripping and killing. Frugivore guts are generous, with a capacious colon for fermenting fiber. A cat’s gut is short and simple; meat is easy to digest and doesn’t need fermenting. Our ancestors had the former, not the latter.29

Bottom line for Part 2: before there was anything resembling a human, our lineage spent tens of millions of years as fruit specialists, and it left permanent marks: color vision, sweet taste, and a broken vitamin C gene that would kill a carnivore.

Part 3: The cousins (what chimpanzees and bonobos actually eat)

Our closest living relatives, chimpanzees and bonobos, share a common ancestor with us roughly 6 to 8 million years ago. They aren’t our ancestors, but they are the best living window we have onto what that ancestor was probably like.

Here is what more than sixty years of field observation at Gombe, Taï, Ngogo, Kibale, and Budongo have found. Chimpanzees are ripe-fruit specialists. Fruit typically accounts for half to two-thirds of their feeding time, and considerably more when fruit is abundant. The rest is leaves, pith, seeds, flowers, bark, and insects like termites and ants.30

And meat? The Jane Goodall Institute puts it at about 2 to 5 percent of the Gombe diet. Some sites report a bit more, but nowhere does meat approach a staple.31

What makes this especially telling is that chimpanzees are good hunters. At Ngogo in Uganda, the largest chimpanzee community ever studied, David Watts and John Mitani documented hunts of red colobus monkeys that succeed more than 80 percent of the time, averaging several kills per successful hunt. Lions succeed roughly a quarter of the time. Chimpanzees are, pound for pound, far more effective predators than the great cats.32

Yet with all that skill, they eat very little meat, and (here’s the part that should stop a carnivore-diet advocate cold) they hunt more when other food is plentiful, not when they’re hungry. Hunting at Ngogo peaks when fruit is abundant and large groups gather. Meat for chimpanzees is a prized social currency, shared to build alliances and court mates, not a dietary foundation.32

Bonobos, our other closest cousins, eat even less meat. They live on ripe fruit and the soft stems and shoots of terrestrial herbs, with occasional documented hunts of small antelope and monkeys. Food sharing is often led by females.33

Widen the lens to all the great apes and the pattern is unanimous. Mountain gorillas eat leaves; lowland gorillas eat fruit when they can get it. Orangutans are so devoted to fruit that they gorge and store fat during the irregular “mast” fruiting years of Southeast Asian forests and live off it during lean periods.34 Not one great ape is a carnivore. Not one is even a serious omnivore in the way a bear or a pig is.

Chimpanzees also foreshadow the human toolkit. They fish for termites with stems, crack nuts with stone hammers and anvils at Taï, extract honey, and at Fongoli in Senegal, Jill Pruetz documented them sharpening sticks to spear bushbabies in tree hollows and using digging tools to pull up roots.35 The whole flexible foraging repertoire our ancestors would build on, including the sticks for digging tubers, is already there in a fruit-eating ape.

The most reasonable reconstruction of the last common ancestor of humans and chimpanzees is a woodland ape eating mostly ripe fruit and other plant foods, some insects, and an occasional small animal. A frugivore that dabbled. That is the starting line for the human story.

Part 4: Hard times, hard foods (the Miocene apes and the first hominins)

Around 10 million years ago the world began to cool and dry. Forests fragmented into woodland mosaics and eventually grasslands. Fruit became seasonal and patchy. Apes had to cope, and the fossils show how: thick molar enamel.

Miocene apes like Afropithecus, Sivapithecus, and Ouranopithecus evolved thick enamel on their molars. This has sometimes been read as a step toward a “harder” diet, and it was, but the hard things were nuts, seeds, hard-shelled fruit, and tough roots, the “fallback foods” an ape turns to when the ripe figs run out. Thick enamel resists the cracking and grinding of hard plant parts. It has nothing to do with meat, which is soft and needs no enamel at all.36

Then, between 7 and 4 million years ago, some of these apes started walking upright. Sahelanthropus (about 7 million years), Orrorin (about 6 million), and then the wonderfully preserved Ardipithecus ramidus, “Ardi,” at 4.4 million years. Ardi lived in woodland, had small canines, and carbon isotopes dominated by C3 forest plants. She was an omnivorous forager of fruit, plants, and probably insects. Nothing about her says hunter.37

Why walk upright? Hypotheses include carrying food back to share, foraging efficiently between scattered trees in a patchy woodland, reaching low-hanging fruit while standing in the branches, and staying cool. Notice that several of these are about plant foraging in an increasingly open landscape.

The australopiths and the grass revolution

Australopithecus (roughly 4.2 to 2 million years ago) is where the isotopes start telling a fascinating story. Starting around 3.5 to 3.8 million years ago, hominin tooth enamel shows a growing “C4” signal: they were eating food from the tropical grass and sedge world, not just the forest.38, 39

Now, here’s where a careless reader goes wrong. C4 does not mean meat. A C4 signal could come from eating grasses and sedges directly, from eating the starchy underground parts of sedges (tiger nuts and similar corms are a C4 food), or from eating animals that ate grass. The isotopes alone can’t tell you which. Dental microwear and anatomy have to break the tie.

In 2025 they got a spectacular assist. Tina Lüdecke and colleagues at the Max Planck Institute for Chemistry measured nitrogen isotopes not in collagen (long gone after 3.5 million years) but locked inside tooth enamel, using a new technique. They analyzed seven Australopithecus individuals from Sterkfontein, South Africa, and compared them to contemporary herbivores and carnivores from the same cave. The result: australopith nitrogen values sat down among the herbivores, far below the carnivores. Their diet was variable but plant-based, with no sign of regular meat consumption.40

Let that sink in. Three and a half million years ago, the direct predecessors of our genus were, in any practical sense, vegetarians.

Australopithecus sediba, at about 2 million years, adds texture: phytoliths trapped in its dental calculus show it ate bark, leaves, fruit, and wood, a forest-edge forager’s menu.41

Nutcracker Man, the hominin cow

If you want a single story that shows how badly a skull can mislead you about diet, it is Paranthropus boisei.

Found by Mary Leakey at Olduvai in 1959, the skull known as OH 5 has the most massive chewing apparatus of any hominin: enormous flat molars, a huge jaw, flaring cheekbones, a bony crest on top of the skull to anchor giant chewing muscles. It was nicknamed “Nutcracker Man,” and for decades everyone assumed it cracked hard nuts and seeds.

Then Peter Ungar looked at the microwear and found almost no pitting, the signature of hard objects, and mostly the fine scratches of tough, fibrous food.18 And in 2011 Thure Cerling’s team measured carbon isotopes in Paranthropus boisei teeth and found the diet was around 77 percent C4 on average, ranging from 61 to 91 percent: grasses and sedges, the food of zebras and warthogs. It was, in the memorable phrase, the “cow of the hominins,” using those giant molars to grind vast quantities of low-quality grass and sedge.42

The hominin with the scariest-looking jaw was a grazer. Remember that the next time someone tells you a body part “proves” a meat diet.

Roots, tubers, and digging sticks

Somewhere in this period a food that would carry hominins through millions of years came into its own: underground storage organs. Roots, tubers, corms, bulbs, rhizomes. Plants stash starch underground to survive the dry season, and that starch is there year-round if you can dig for it. Greg Laden and Richard Wrangham argued in 2005 that these underground foods were the key fallback that let hominins survive in seasonal savannas where fruit disappears for months.43 The Hadza of Tanzania still dig tubers with sharpened sticks every day, and women’s tuber gathering is a backbone of their diet.44

We have direct evidence of the technology, too: boxwood digging sticks made by Neanderthals in Italy 171,000 years ago, shaped and fire-hardened for pulling up roots.45 Chimpanzees at Fongoli use digging tools today. The digging stick is as old and as fundamental a human tool as the spear, and it has never once appeared in a cave painting.

The first stone tools were food processors, not just butcher knives

Stone tools long predate our genus. The oldest known, from Lomekwi in Kenya, are 3.3 million years old, older than Homo.46 The Oldowan tradition appears by about 2.6 million years ago.

At Nyayanga, Kenya, Tom Plummer’s team found Oldowan tools about 2.9 million years old alongside butchered hippo bones and, crucially, evidence from wear analysis that the same tools were used to cut and scrape animal tissue and to pound plant material. The only hominin fossils at the site are Paranthropus teeth.47 Rick Potts of the Smithsonian described stone tools as “like suddenly evolving a brand-new set of teeth outside your body.” Those new teeth chewed roots as readily as they cut meat.

Part 5: Becoming Homo (meat, marrow, fire, and starch)

Now I have to be honest about the part of the story where meat really does step onto the stage, because it does.

Around 2.8 million years ago the genus Homo appears, and by 1.9 million years ago Homo erectus is a different kind of creature: bigger body, much bigger brain, smaller teeth and jaws, a narrower rib cage suggesting a smaller gut, and long legs built for walking and perhaps running enormous distances.48 Something changed in the diet to make all of that possible, and animal foods were part of it.

Marrow first, then hunting

At Kanjera South in Kenya, about 2 million years ago, hominins were repeatedly bringing whole small antelopes and parts of larger ones back to the same spot and butchering them, the earliest evidence of persistent meat eating.49 Jessica Thompson and colleagues have argued that the first step wasn’t hunting at all but scavenging what other predators couldn’t get at: smashing open bones for marrow and skulls for brains, both dense in fat.50 Whether early Homo were mostly scavengers or mostly hunters was one of the great debates of the field, and the answer is probably “both, depending on time and place.”

Why did animal fat and protein matter so much right then? Because brains are expensive. A modern human brain is 2 percent of body weight but consumes about 20 percent of resting energy in adults, and far more in children. Leslie Aiello and Peter Wheeler’s “Expensive Tissue Hypothesis” (1995) proposed a trade: a higher-quality, more digestible diet let the gut shrink, freeing energy for a bigger brain.51 A later analysis across mammals found no clean gut-for-brain trade-off and pointed to fat storage and other factors instead.52 The specific swap is contested; the broader point, that a richer diet was necessary to afford a larger brain, is not.

So: concession made. Marrow, fat, meat, and brains were real, important, high-value foods for early Homo. Anyone who tells you meat was irrelevant to human evolution is also wrong.

But brains run on sugar

Here is the counterweight. The brain doesn’t burn protein or fat directly under normal conditions. It burns glucose. And it burns a staggering amount. Christopher Kuzawa’s team showed in 2014 that in a five-year-old child, the brain’s glucose use is equivalent to about 66 percent of the body’s resting metabolism and about 43 percent of total daily energy needs.53 Humans have the longest, most brain-hungry childhood of any animal.

That is why Karen Hardy, Jennie Brand-Miller, and colleagues argued in 2015 that dietary carbohydrate, especially cooked starch from tubers and other plants, was essential to human brain evolution, and that the multiplication of our salivary amylase genes co-evolved with cooking.54 The genetic timing details have been debated, but the physiological logic is hard to escape. You can’t easily feed a child’s brain on lean meat.

Nature left one more clue. Human breast milk is unusually low in protein and high in lactose, a sugar. Our infants, in their most rapid period of brain growth, are fed a high-carbohydrate diet by design.55

Fire

Richard Wrangham’s cooking hypothesis, laid out in Catching Fire (2009), argues that cooking is the defining human adaptation. Cooking gelatinizes starch and denatures protein, making both far more digestible; it kills pathogens; it softens food so we can chew it in an hour a day instead of six; and it let our teeth, jaws, and guts shrink while our brains grew.56

When did we tame fire? This is genuinely unresolved. There are tantalizing traces at Wonderwerk Cave in South Africa around a million years ago, and firm evidence of repeated hearths at Gesher Benot Ya’aqov in Israel around 790,000 years ago. Habitual fire use in Europe is only solidly documented from about 400,000 years ago.56 Wrangham thinks cooking began with Homo erectus nearly 2 million years ago; many archaeologists think the evidence supports a later date. Either way, for at least the last several hundred thousand years, we have been the cooking ape.

And what were they cooking at Gesher Benot Ya’aqov? Fish, for one; the site preserves the earliest evidence of cooked fish, about 780,000 years old.57 But the waterlogged sediments also preserved something almost never seen from this period: a plant assemblage of 55 species, including nuts, fruits, seeds, water chestnuts, and tubers. Yoel Melamed and colleagues described a diet built on plants and aquatic foods alongside meat.58 When the preservation is good enough to see the plants, the plants are everywhere.

Even before fire, tools shrank our jaws

Katherine Zink and Daniel Lieberman ran a lovely experiment in 2016: they fed volunteers raw goat meat and raw tubers, measured chewing effort, and then tested what happened when the meat was sliced and the tubers pounded with simple stone tools. Slicing and pounding cut the number of chews needed by tens of thousands per year. Tool use alone, before any cooking, could explain much of the reduction in teeth and jaws seen in early Homo.59 Around the same time, the gene MYH16, which builds heavy jaw muscle in other apes, was inactivated in our lineage.60 Note that the experiment needed both foods: meat was hard to chew raw and tubers were hard to chew raw, and tools helped with both.

Homo erectus walked out of Africa and was in Georgia by 1.8 million years ago, then across Asia. You cannot do that on a specialist’s diet. You do it by eating whatever the new place offers.

Part 6: Neanderthals (the carnivores who weren’t only carnivores)

If there is one piece of evidence carnivore-diet proponents cite most, it’s Neanderthal isotopes. So let’s look at it squarely.

Nitrogen isotopes in Neanderthal bone collagen from Ice Age Europe are high. As high as wolves and hyenas. Michael Richards, Hervé Bocherens, and others concluded that European Neanderthals were “top-level carnivores,” eating reindeer, horse, mammoth, and rhino.20 In 2019 Klervia Jaouen’s team sharpened the picture by measuring nitrogen in individual amino acids of the collagen, a more refined method, and reached the same conclusion,61 and in 2022 a completely different technique, zinc isotopes in tooth enamel from a Neanderthal in Spain, agreed again.62 And in 2025, Lutz Kindler and colleagues published one of the most striking sites in Neanderthal archaeology: at Neumark-Nord in Germany, 125,000 years ago, Neanderthals processed the bones of at least 172 large mammals, smashing them into over a hundred thousand fragments and heating them in water to render the fat. A “fat factory.” Intensive, organized, strategic.63

I’m not going to pretend this away. In glacial Europe, Neanderthals were highly carnivorous, skilled big-game hunters who understood fat as the critical resource it was. Concession made, fully.

Now the caveats, which are not small.

First, isotopes see protein, not calories. Bone collagen nitrogen tells you about the protein you ate. If a Neanderthal got most of her protein from reindeer and a large share of her calories from cooked roots, nuts, and berries, the collagen would say “carnivore.” Carbohydrates are nearly invisible to this method.20

Second, the sample is skewed. Collagen survives best in cold, dry caves, so the isotope record over-represents glacial Europe, the coldest, least plant-rich environment any human ever inhabited. It’s like sampling only Inuit and concluding humans are seal-eaters.

Third, Neanderthals are not our ancestors. They are a sister lineage, adapted to a specific brutal climate. Their cold-steppe diet is not the human baseline.

Fourth, and decisively: when we look at Neanderthals with methods that can see plants, we find plants. Amanda Henry, Alison Brooks, and Dolores Piperno found cooked starch grains from date palm, legumes, and grass seeds in the dental calculus of Neanderthals from Shanidar in Iraq and Spy in Belgium (yes, Belgium, in the cold north).21 Laura Weyrich’s team sequenced ancient DNA from Neanderthal calculus in 2017: the Spy individuals had eaten woolly rhino and wild sheep, but the El Sidrón Neanderthals in Spain showed mushrooms, pine nuts, and moss with no meat signal at all, and one sick individual had been chewing poplar bark (source of salicylic acid, the active ingredient in aspirin) and a Penicillium mold.22 Karen Hardy found yarrow and chamomile, bitter plants of no caloric value and likely medicinal, in El Sidrón calculus.64 At Shanidar, Ceren Kabukcu’s team found 70,000-year-old charred remains of cooked pulses, wild mustard, and nuts, soaked and pounded and combined: not just food but cuisine.65

And then there’s the poop. At El Salt in Spain, Ainara Sistiaga’s team found 50,000-year-old Neanderthal coprolites, the oldest known hominin feces. The biomarkers showed a diet dominated by meat, as expected, but also a substantial, unambiguous plant component from the sterols only plants produce.23 Even in the fecal record, even in a meat-eating population, the plants show up.

The fair summary: Neanderthals were flexible too. In the cold north they leaned heavily on meat and fat because that’s what there was. In Iberia and the Levant they ate a much broader diet. Nowhere were they pure carnivores, and their tartar proves it.

Part 7: Our own species and the hidden plant record (300,000 to 12,000 years ago)

Now to Homo sapiens. This is where the “bones survive, plants rot” bias has done the most damage to popular understanding, and where the last two decades of archaeobotany have been most revealing.

170,000 years ago, Border Cave, South Africa. Lyn Wadley’s team found 55 charred, whole rhizomes of Hypoxis, a starchy plant still eaten in the region, roasted in hearths. This is the oldest direct evidence of cooking underground storage organs. Hypoxis rhizomes are nutritious, roughly 500 kilojoules per 100 grams, available year-round, and easy to dig with a stick. A reliable, cooked carbohydrate staple, at the dawn of our species.66

120,000 years ago, Klasies River, South Africa. Roasted tubers and rhizomes again, in hearths.67

65,000 years ago, Madjedbebe, Australia. Among the earliest traces of people in Australia, grinding stones and charred plant remains showing the processing of fruits, nuts, roots, and palm stems.68 At Niah Cave in Borneo, around the same time, people were detoxifying poisonous wild yams, which requires sophisticated botanical knowledge.

30,000 years ago, across Europe. Anna Revedin’s team found starch grains from cattail and fern rhizomes on grinding stones at Gravettian sites in Italy, Russia, and the Czech Republic. Ice Age Europeans, the very people whose isotopes look so carnivorous, were making flour.69

23,000 years ago, Ohalo II, Israel. A waterlogged camp on the Sea of Galilee preserved about 150,000 plant remains from roughly 140 species, including wild wheat, barley, and oats, along with a grinding stone bearing cereal starch. Ainit Snir and colleagues even found “proto-weeds,” suggesting small-scale trial cultivation 11,000 years before farming.70

14,400 years ago, Shubayqa 1, Jordan. The oldest bread yet found, made by Natufian hunter-gatherers from wild cereals and tubers, thousands of years before agriculture.71

15,000 years ago, Grotte des Pigeons, Morocco. Louise Humphrey’s team found the earliest population with widespread tooth decay, over half the adult teeth affected, from a diet heavy in acorns and pine nuts. Cavities are a signature of a starchy diet.72

And the coprolites. From Hinds Cave in Texas to the Paisley Caves in Oregon, where Eske Willerslev and Thomas Gilbert’s team recovered human DNA from 14,000-year-old feces,73 preserved hunter-gatherer coprolites are packed with seeds, fibers, pollen, and plant tissue. Some imply fiber intakes well over 100 grams a day; the modern Western average is around 15. Whatever these people were painting, they were eating an enormous quantity of plants.

None of this means Ice Age hunters didn’t hunt. They did, brilliantly. Mammoth, reindeer, horse, bison. Fish from the rivers (the isotopes at Kostenki and Dolní Věstonice show increasing freshwater fish). They rendered fat and stored it. Meat and fat were essential in glacial climates. But the plant record, wherever preservation allows us to see it, is rich and deep, and it stretches back to the very beginning of our species.

Part 8: The cave paintings are not a menu

Let me come back to Lascaux, and to the argument you’ll hear: “Look at the cave paintings. They only painted animals. That’s what they ate.”

Three problems.

They painted the wrong animals. At Lascaux, horses dominate with well over 300 images, followed by stags, aurochs, and bison. Reindeer, the dietary staple of the region, are essentially absent.1 At Chauvet, 36,000 years old, the walls are covered with lions, mammoths, rhinos, and cave bears, which together make up nearly two thirds of the figures. People did not live on cave lions. The Metropolitan Museum of Art’s own summary notes that these images “are thus not simple depictions of daily life.”74 Predators outnumber prey. It’s the opposite of a food log.

They painted people too. Hand stencils, the “Venus” figurines, the human figure in the Lascaux shaft, the therianthropes (human-animal hybrids) at Chauvet and in Sulawesi. By the “it’s a menu” logic, they were cannibals. Obviously not. They drew what fascinated, frightened, and meant something to them. Animals and people both did.

They never painted plants. Not once, anywhere in the Paleolithic art record, is there a clear depiction of a fruit, a tuber, a nut, or a grain. And yet, as we’ve just seen, we have their flour, their bread, their roasted rhizomes, their cavities, and their plant-stuffed feces. If the absence of plants in the art meant they didn’t eat plants, the physical evidence would be impossible.

Archaeologists abandoned the “hunting magic” reading of cave art (Abbé Breuil’s idea, a century ago) precisely because the painted animals don’t match the hunted ones. Later theories emphasize symbolism, structure, shamanic experience (David Lewis-Williams), and information storage; a recent and heavily contested proposal even reads the dots and lines as a lunar calendar of animal behavior.75, 76 What every serious interpretation agrees on is that the art was about meaning, not menus.

To be fair, art sometimes does line up with food. In Sulawesi, Indonesia, a painting of a warty pig is at least 45,500 years old; a hunting scene at Leang Bulu’ Sipong 4, first dated to at least 43,900 years, was re-dated in 2024 to at least 48,000; and a newly dated scene at Leang Karampuang showing human-like figures with a pig is at least 51,200 years old. These people did hunt pigs and dwarf buffalo.77, 78 So art can reflect prey. It just can’t tell you what fraction of the diet that prey was, and it’s silent on everything that grew from the ground.

What does tell us about diet are the unglamorous artifacts: grinding stones, mortars, digging sticks, fish hooks (23,000 years old in Okinawa; deep-sea tuna fishing 42,000 years ago in Timor79, 80), the impressions of nets and woven plant-fiber textiles at Dolní Věstonice 26,000 years ago, shell middens, roasting pits, and storage pits. And yes, the wooden spears of Schöningen in Germany, evidence of organized big-game hunting by archaic humans, originally dated to 300,000 to 400,000 years and re-dated in 2025 to about 200,000 years.81 Hunting was real. So was everything else.

Part 9: Farming, milk, and the genes that caught up (12,000 years ago to today)

About 11,500 years ago, in the Fertile Crescent, people who had been harvesting wild cereals for millennia began deliberately planting them. Independently, and within a few thousand years, others domesticated rice in China, maize in Mesoamerica, potatoes in the Andes, sorghum and millet and yams in Africa, and taro and bananas in New Guinea. Every one of these founding crops was a carbohydrate.

Agriculture was a mixed blessing for individuals, a triumph for populations. Early farmers were shorter, had more cavities, more anemia, and more infectious disease than the foragers before them.82 But they had more babies, and the world filled up with them.

The transition also left marks in our genes, and one of them is the best single-gene rebuttal to the carnivore story.

The starch gene

Salivary amylase is the enzyme that starts breaking down starch in your mouth. It’s why a cracker turns sweet if you hold it on your tongue. Chimpanzees and bonobos have one copy of the gene, AMY1. Neanderthals had one. Modern humans have anywhere from two to more than a dozen.83

In 2007, George Perry and colleagues showed that populations with high-starch diets, like the Japanese and the Hadza, carry more copies than populations with low-starch diets.83 Two major studies in 2024 pushed the story much deeper. Omer Gokcumen’s team found that the initial duplication of AMY1 happened hundreds of thousands of years ago, quite possibly around 800,000 years, and that hunter-gatherers 45,000 years ago already carried four to eight copies, long before any farming. Then, with agriculture, copy numbers rose further: European farmers went from an average of about four to about seven copies over roughly 4,000 years.84, 85 Our lineage was under selection to digest starch since before our species existed, and again, hard, once we started growing it.

There is a wonderful coda. In 2013 Erik Axelsson’s group sequenced dog and wolf genomes to find what changed during domestication. Among the top hits: dogs had multiplied their amylase gene too, along with other starch-digestion genes. Dogs became better starch-eaters by living off the scraps of a starch-eating species.86 The animal that shared our campfires evolved to share our carbs.

Milk, fat, and other adaptations

Around 5,000 to 7,000 years ago in Europe, and independently in East Africa and Arabia, variants that keep the lactase enzyme switched on into adulthood spread rapidly, letting adults drink milk. Richard Evershed’s 2022 study of pottery lipids showed that Europeans were drinking milk for thousands of years before lactase persistence became common, and that the gene’s spread was probably driven by famine and disease, when milk-induced diarrhea turned deadly.87 That distinction is worth holding onto, because it is the whole lesson of this post in miniature. Lactase persistence did not spread because milk is optimal. It spread because during famine, not being able to digest it could kill you. An adaptation is a record of what a population survived, not an endorsement of what it ate.

Then there is the detail that settles the argument. The FADS gene cluster builds the long-chain fatty acids your brain and cell membranes need. In Greenlandic Inuit, selection favored variants that suppress that pathway, which makes sense when your food already arrives packed with marine omega-3.13 In Europeans, selection favored variants that increase production of the same fats from the short-chain precursors you get from plants.88 The same gene cluster, two populations, opposite directions: tuned in one case to animals, in the other to plants. Arctic peoples also carry that CPT1A fat-processing variant at around 70 percent frequency and essentially zero elsewhere.11 And adult milk digestion in the Maasai runs not on the European lactase variant but on a different one, chiefly C-14010, which arose independently in East Africa on a different genetic background. The selection signal there is the strongest in the entire Maasai genome, and stronger than the equivalent European one.89, 90

This is the real answer to “our ancestors ate this, so should we.” These peoples are not interchangeable bodies eating different foods. They carry different machinery, tuned by different histories, and none of it transfers. A Maasai herder could drink three liters of milk a day because of an allele you almost certainly do not have. Local diets, local genes. There is no single ancestral human diet written in our DNA; there is a genome tuned for flexibility and then fine-tuned, everywhere, to local circumstances.

What living hunter-gatherers actually eat

Because so much of the carnivore argument rests on an imagined Paleolithic, it’s worth looking at the real foragers we can still observe.

The Hadza of northern Tanzania, among the best-studied, eat honey, baobab fruit, berries, tubers, and meat. Honey is their single favorite food and can supply around 15 percent of their calories.91, 92 Women’s gathering of tubers and fruit provides a large share of the daily food; men hunt, with wildly variable success. The Ju/’hoansi of the Kalahari, studied by Richard Lee in the 1960s, got the majority of their calories from plants, above all the mongongo nut.93

The Inuit, the perennial example, ate a diet that was overwhelmingly animal because the Arctic offers almost nothing else, and even they ate berries, roots, seaweed, and the partially digested plant contents of caribou stomachs. They are the extreme end of human flexibility, not the human norm. And the famous claim that they were protected from heart disease has not survived scrutiny: it came from incomplete Greenlandic death certificates rather than from anyone examining arteries, and specialists retired it two decades ago.94, 95 When CT scanners were finally pointed at preserved bodies from four preindustrial populations, atherosclerosis turned up in a third of them, including three of the five Unangan hunter-gatherers of the Aleutian Islands, who lived on marine mammals.96 Five bodies cannot establish a rate. They are enough to establish that a diet of seal and whale did not prevent the disease. I take that argument apart in full elsewhere, along with the Maasai version of it.

A frequently cited 2000 analysis by Loren Cordain estimated that hunter-gatherers averaged about 65 percent animal foods.97 That number came from an ethnographic database whose coding lumps fishing and hunting into broad categories and systematically under-records gathering, especially by women. When Herman Pontzer and Brian Wood reviewed the evidence in 2021, they found the plant-to-animal balance tracks climate and economy rather than anything about human nature: a roughly even split among warm-climate hunter-gatherers, more meat in cold climates and in fishing and herding societies, more plants among warm-climate foragers and part-time farmers. Paleolithic diets, they concluded, would have been remarkably variable across time and space, and there is no single “paleo diet.”98 What foragers share is not a macronutrient ratio but a pattern: whole foods, high fiber, enormous variety, and a lot of physical work.

And a plant-heavy version of that pattern looks remarkably good in the body. In a landmark 2017 study, Hillard Kaplan’s team CT-scanned the hearts of 705 Tsimané adults in the Bolivian Amazon, a population living on farmed plantains, rice, manioc, and corn plus fish and game, a diet about 72 percent carbohydrate and 14 percent fat. Eighty-five percent had no detectable coronary artery calcium, compared with 14 percent of American adults of similar age, and their arteries were estimated to be roughly 28 years younger than those of a comparable US population.99

Now the honesty this post owes its own argument. I have just spent nine sections warning against reading a single population as proof of what humans should eat, so I am not going to turn around and do it. The Tsimané are extraordinarily physically active, lean, largely non-smoking, and heavily burdened by infection. Every one of those independently lowers cardiovascular risk, and none of them is the diet. The paper’s own conclusion leads not with any particular food but with a lifetime of very low LDL cholesterol, averaging about 91 mg/dL. That is the useful lesson, and it happens to be a mechanism rather than a menu: whatever keeps LDL low across a whole life protects arteries. Pick any traditional population and you can build a dietary story on a confounded foundation. The Maasai have been used this way for sixty years, and so have the Inuit. The ancestral appeal cannot be settled by choosing a better population, because there is always another one. It can only be settled by evidence that controls for the confounders, which is a different kind of study and a different discussion.

Part 10: Taking the carnivore argument apart, claim by claim

You’ve now got the whole story. Here’s how it bears on the specific claims you’ll hear.

“Humans evolved as apex predators for two million years.” This comes mainly from a 2021 review by Miki Ben-Dor, Raphael Sirtoli, and Ran Barkai, which assembled 25 lines of indirect evidence (from stomach acidity to gut proportions to fat storage) to argue that Pleistocene Homo was hypercarnivorous until the megafauna declined around 85,000 years ago.100 It is a genuine scholarly hypothesis and deserves to be treated as one, not dismissed as fringe. But it’s a synthesis of indirect arguments, not a body of direct dietary evidence, and it runs into the problems we’ve covered: the taphonomic bias that inflates meat in the record, the isotopes’ blindness to carbohydrate, the 2025 enamel study showing australopiths ate no regular meat, the amylase duplications hundreds of thousands of years old, the cooked rhizomes at 170,000 years, and the plant-rich coprolites and calculus. Many anthropologists read the same evidence and arrive at flexible omnivory.98 Say “contested,” not “debunked,” and then lay out why the balance of evidence is against it.

“Our stomach acid is as strong as a scavenger’s.” True, and interesting. A 2015 study found human stomach pH around 1.5, comparable to carrion feeders.101 But the authors’ own explanation is that acidity is a pathogen filter, not a dietary readout, and their data show acidity doesn’t sort neatly by diet. Strong acid is consistent with a species that ate risky foods, scavenged carcasses included. It doesn’t tell you what fraction of the diet those foods were.

“Our short colon means we’re not built for plants.” Katharine Milton’s comparative work shows the human gut is a textbook omnivore’s: a long small intestine for absorbing high-quality, digestible food, and a colon that’s smaller than a chimp’s but still ferments fiber into useful short-chain fatty acids.29 It shrank because we cook and process food, not because we became cats. Actual carnivores have short, simple guts throughout, no salivary amylase, and can’t taste sweet.

“We have canines.” Our canines are tiny, blunt, and have been shrinking since Ardipithecus. Our molars are flat millstones. Our jaws move side to side to grind. A lion’s teeth shear and its jaw only opens and closes. We chew; carnivores gulp. If our teeth prove anything, they prove omnivory with a strong plant lean.

The anatomy charts

A whole genre of infographic sets a human beside a gorilla or a cow, lists the plant-processing equipment we lack, and concludes we were built for meat. The format is persuasive and the errors are severe enough to be worth taking one at a time.

“A gorilla makes cellulase and you don’t.” Gorillas do not make cellulase. Neither do cows. No mammal does. Mammals possess no enzyme capable of breaking the bonds in plant cell walls; every cellulose-digesting animal on Earth outsources the job to symbiotic microbes, and so do you. When a chart assigns a nonexistent enzyme to the wrong animal, the rest of its rows deserve the same scrutiny.

“Your gut bacteria do nothing with fiber.” You carry roughly 38 trillion bacteria, and the dominant fiber-fermenting genera include one, Ruminococcus, named for the rumen itself. They turn fiber into short-chain fatty acids, and butyrate is the preferred fuel of your own colon cells. That is the mechanism behind most of fiber’s measured benefit.

“Compare us to a gorilla.” Gorillas are folivores, leaf specialists, an endpoint every bit as extreme as the Arctic. Our closest relatives are chimpanzees and bonobos, ripe-fruit specialists who get 2 to 5 percent of their diet from meat. Picking the one great ape that went furthest into leaves, noting we don’t match it, and concluding we are carnivores is a comparison problem, not a finding.

“We lack the jaw muscles, the sagittal crest, and the bite force of a plant eater.” Lions and hyenas have prominent sagittal crests and formidable bites; neither feature marks a plant eater. And the hominin with the heaviest chewing apparatus that ever existed, Paranthropus boisei, was eating grasses and sedges. What actually shrank our jaws is extraoral processing: slicing and pounding with stone tools, then cooking, and the chewing experiments show it worked on raw tubers just as much as raw meat.

“A gallbladder, bile, lipase and lacteals are meat-eating equipment.” Cattle, sheep and goats have gallbladders. Every mammal has lacteals, including strict herbivores. Bile is what lets you absorb the fat-soluble vitamins and carotenoids in plants, and fiber lowers cholesterol partly by binding bile acids. Lipase is universal, because nuts, seeds, olives and avocados are fat. All of it is real equipment. None of it is diagnostic of anything.

“A cow manufactures B12 and you manufacture none.” Neither does the cow. B12 is synthesized only by bacteria and archaea; no plant and no animal makes it. The cow’s rumen microbes make hers, and only if she gets enough cobalt, which is why cobalt is a routine feed supplement and why deficient cattle are treated with direct B12 injections. Both routes end at bacteria. One of them just takes eighteen months and an animal.

“Humans are trophic level 13, above the lion and the wolf.” Trophic levels run from 1 to 5. Primary producers are 1 and apex predators top out around 5, because each level loses roughly 90 percent of the energy below it; a level-13 animal would need more energy than the biosphere produces. The actual measurement exists: analyzing the food supply of 176 countries, researchers put the human trophic level at 2.21, about that of an anchovy or a pig, with national values spanning only 2.04 to 2.57, and framed the result as a challenge to “the perception of humans as top predators.”102 Lions and wolves sit near 4. The most carnivorous nation on Earth does not reach 2.6. Even the strongest academic case for Pleistocene hypercarnivory argues for a position near the top of a five-point scale, not past the end of it.

What these charts share with the Maasai and Inuit arguments is the same hidden step. Each takes a real anatomical fact and infers optimality from it. Even where the facts are right, and some are, the conclusion does not follow. A gallbladder tells you what an animal can process. It never tells you what it should eat.

“Neanderthal isotopes show hypercarnivory.” Covered in Part 6. Protein-only signal, cold-climate sample, sister lineage, and their own tartar is full of cooked plants.

“The cave paintings show only animals.” Covered in Part 8. Wrong animals, plus people, minus plants.

“Meat made us human.” Partly. Marrow, fat, and meat were essential contributors to the brain-growth story, alongside cooked starch, tubers, honey, aquatic foods, tools, fire, and social cooperation. Brains run on glucose. Human milk is sugar-rich. “Meat made us human” is one thread in a rope.

“We can’t survive on plants; we need meat.” We can’t thrive on lean meat either. Humans have a protein ceiling around 35 to 40 percent of calories; beyond it, “rabbit starvation” sets in, with nausea, diarrhea, and eventually death, a phenomenon well known to Arctic explorers.103, 104 Any all-meat human diet has to be very high in fat precisely because our physiology rejects a high-protein one. Meanwhile we digest cooked starch superbly, we ferment fiber, and unlike cats we synthesize taurine and don’t need preformed vitamin A from liver. We do need vitamin B12, which is easiest to get from animal foods or a supplement, which is one reason “omnivore” is the right word (here is the full rundown of essential nutrients).

Part 11: The future is in our hands (what you eat reaches forward)

If evolution doesn’t dictate our diet, what does the choice actually affect? More than your own body, it turns out. This is the part of the story where I want to be especially careful, because it is both genuinely remarkable and genuinely overhyped.

You once lived inside your grandmother

Start with a fact of embryology that sounds like a riddle. A woman is born with every egg she will ever have. Those eggs form while she is still a fetus: the classic count, from a 1963 study, puts the total across both ovaries at about 6.8 million germ cells by the fifth month of gestation, falling to roughly 2 million by birth and dwindling from there.105 (Those textbook numbers rest on remarkably few specimens and are debated, but the timing is not.) Follow the logic. When your grandmother was pregnant with your mother, the cell that would become half of you was already forming inside your mother’s tiny ovaries. Your grandmother’s nutrition, stress, and environment during that pregnancy reached three generations at once: herself, her daughter, and the cell that became you. Your DNA is still a fresh shuffle of both parents. But the cellular starting conditions were laid down a generation earlier than most people imagine.

Scientists draw a line here that the popular press often blurs. Effects on a generation that was directly exposed, a fetus in the womb or the egg and sperm cells inside that fetus, are called intergenerational. Effects that persist into a generation that was never exposed at all are transgenerational, and those are much harder to prove, because the marks on our DNA are largely wiped clean and reset when eggs and sperm are made and again in the early embryo.

What the evidence actually shows

The intergenerational evidence is strong. In the Dutch Hunger Winter of 1944 to 1945, people conceived during the famine were compared, six decades later, with their own same-sex siblings conceived before or after it. In 60 sibling pairs, those exposed around conception carried about 5 percent less methylation on the imprinted growth gene IGF2, a chemical mark that regulates how the gene is read, and the effect was specific to exposure around conception rather than late pregnancy.106 Half a century of research on the “developmental origins of health and disease” says the same thing in a hundred ways: nutrition in the womb sets metabolic risk for life.

Fathers matter too. In rats, a father’s chronic high-fat diet programmed dysfunction in the insulin-producing cells of his daughters, the first demonstration in mammals that a father’s diet could reach his offspring without any shared womb.107 In humans, the sperm of obese men carries a different pattern of epigenetic marks than that of lean men, and the pattern shifts within a year of weight loss after bariatric surgery.108 Whether those changes alter children’s health is still being worked out, but the mechanism is real.

Grandparents are where it gets interesting and where the caution matters. In the isolated Swedish parish of Överkalix, with meticulous harvest records going back to the 1800s, researchers found that if a boy had experienced a glut of food around ages nine to twelve, his grandchildren died of diabetes at four times the rate of others.109 The pattern was strangely sex-specific, running from paternal grandfathers to grandsons and from paternal grandmothers to granddaughters.110 A far larger Swedish study in 2018 tested the claim: grandsons of grandfathers who had abundant food before puberty did have higher mortality, especially from cancer, but the diabetes and heart-disease findings did not replicate.111 The Överkalix researchers themselves were careful to say they were not claiming the mechanism was epigenetic, only that it was a good candidate. And a 2018 review in Nature Communications cautioned that true transgenerational epigenetic inheritance, confirmed in plants, worms, and flies, remains a matter of controversial debate in humans, because it is nearly impossible to separate from what families also pass down: genes, environments, and habits, including what they cook.112

So the honest ladder looks like this. That a mother’s diet in pregnancy shapes her child’s lifelong risk: established. That a father’s diet and weight alter his sperm and can affect his children: strong in animals, emerging in humans. That grandparents’ diets echo in grandchildren: suggestive in some human populations, with a contested mechanism. That epigenetic marks pass unaltered through multiple human generations: not established. (One aside, since the phrase gets thrown around: the “Mother’s Curse” is a separate idea entirely. Mitochondria carry their own small loop of DNA inherited only from your mother, so a mutation that harms only sons is nearly invisible to selection, because it never passes through a son to the next generation. Researchers traced exactly that pattern through French-Canadian genealogies across three centuries.113 It is real, it has nothing to do with diet, and no food changes it.)

The marks can move, but the inheritance runs elsewhere

What makes this hopeful rather than fatalistic is that epigenetic marks are not fixed. They respond to how you live now. In 2023 Stanford researchers randomized 22 pairs of identical twins, one twin to a healthy vegan diet and the other to a healthy omnivorous diet, for eight weeks. The vegan twins ended with lower LDL cholesterol, lower fasting insulin, and about 1.9 kilograms more weight loss than their siblings.114 A companion analysis of the same twins reported that measures of “epigenetic age” derived from DNA methylation fell in the vegan twins but not the omnivorous ones.115 Treat that finding as preliminary: 21 twin pairs for eight weeks is a tiny, short study, methylation clocks are surrogate markers rather than disease, the vegan twins also lost more weight, which by itself could move the clocks, and several authors work for the company that sells the clock tests.

Now the claim I wanted to make in this section, and can’t. It would be satisfying to say that what you eat reprograms your grandchildren’s biology. The human evidence does not support it. The developmental point about the eggs is real and requires no epigenetics at all, and it reaches exactly one generation past you. Everything beyond that is suggestive at best and confounded by the ordinary things families pass down: genes, money, neighborhoods, and recipes.

None of this is a reason to blame your parents or grandparents, who ate what they had in the world they lived in. But it does mean the channel between what we do now and the people who come after us, the one we can actually measure and act on, does not run through the genome.

It runs through the world we hand them. Which is measurable, and which is where this ends.

Part 12: The flexible ape

Here’s the whole arc in one breath.

Sixty million years ago our ancestors were small fruit-eaters in the canopy, and the fruit rewired us: color vision to find it, a sweet tooth to want it, and a broken vitamin C gene that made us dependent on it. Our closest cousins, chimpanzees and bonobos, are still fruit specialists who treat meat as a rare luxury. When the forests thinned, our ancestors stood up and started digging, and the australopiths lived on plants so thoroughly that their enamel reads like an herbivore’s; the hominin with the biggest jaw on Earth was eating grass. Stone tools, from the very beginning, pounded roots as well as cutting hippo. Then Homo arrived and did something new: it added marrow, fat, and meat to the plant base, learned to cook everything, and grew a brain so hungry for glucose that it had to be fed cooked starch through a decade-long childhood. Neanderthals in the Ice Age leaned hard on meat because the tundra gave them little else, and still cooked pulses and chewed medicinal herbs. Our own species roasted rhizomes 170,000 years ago, ground flour 30,000 years ago, baked bread 14,000 years ago, and then, having eaten wild grains for millennia, planted them. Our genes multiplied the starch enzyme in our saliva before farming and again after it, and our dogs’ did too.

We are not carnivores, and we never were. We are not herbivores either. We are the flexible ape: the animal that can live on seal blubber in the Arctic and plantains in the Amazon, that can dig, crack, gather, scavenge, hunt, fish, ferment, and above all cook, and that has left proof of its plant-eating in every place the ground was kind enough to preserve it.

The next time someone points at a cave painting and tells you we come from carnivores, remember the reindeer that fed Lascaux and never made it onto the wall, and the tubers roasting in a South African hearth 170,000 years ago. The bones lie by omission. The teeth, the genes, the tartar, and the poop tell the truth.

So what do we do with the choice?

Here is where the two halves of this post meet. Our evolutionary history proves we can eat almost anything. It does not tell us what we should eat. Evolution is a matchmaker, not a doctor; it handed us flexibility and then stopped caring. That leaves the decision to us, and it means we get to make it on grounds our ancestors never could: the best current evidence on healthspan, the effect on the world we hand to the next generation, and what we are willing to do to other minds.

The health evidence gives a direction, and it is worth stating precisely rather than conveniently. Shifting your diet toward whole plant foods and away from red and processed meat is well supported; that claim rests on many designs that fail in different ways and agree anyway, and it does most of the work.15, 16 A well-planned fully plant-based diet is a defensible endpoint of that gradient, conditional on B12 without exception and attention to vitamin D, iodine, calcium, iron, and enough protein, more of it rather than less past sixty-five. That support is real but weaker, and it takes the form of an absence: no study identifies a level of animal food whose addition to a well-planned plant-based diet improves a hard outcome. Nothing in the data marks a floor. What the evidence does not support is a scientific mandate that everyone must be vegan, and there are genuine countersignals to weigh, including unresolved findings on stroke and fracture risk and the fact that no trial of a vegan diet has yet measured heart attacks or deaths. I go through all of that in detail elsewhere, including the parts that cut against me.

The ecological evidence points the same way, with far greater confidence. The largest analysis of the world’s food system found that meat, aquaculture, eggs, and dairy occupy about 83 percent of the world’s farmland and generate 56 to 58 percent of food’s emissions while supplying 37 percent of our protein and 18 percent of our calories, and that even the lowest-impact animal products typically exceed the highest-impact plant substitutes.116 That is modeling from production data; the check is to take what people actually ate. Linking the dietary records of 55,504 UK adults to food-level environmental data, researchers found vegan diets came in at roughly a quarter of the greenhouse gas emissions and land use of high-meat diets, with every indicator rising alongside animal-food consumption.117 A separate accounting put animal-based foods at 57 percent of all food-production greenhouse gases, roughly twice the plant-based share, with beef alone at a quarter.118 The 2025 EAT-Lancet Commission estimates that food systems account for roughly 30 percent of global emissions and that a shift toward its plant-rich planetary health diet could prevent on the order of 15 million premature deaths a year, up from the roughly 11 million estimated in its 2019 report; these are modeled projections rather than observed results, but sobering ones.119, 120, 121 Roughly three quarters of all antimicrobials sold in the world are used in animals, seeding the resistance that threatens human medicine.122

And then there is the count. According to the UN’s Food and Agriculture Organization, 83 billion land animals were slaughtered for meat in 2022, the vast majority of them chickens, and the figure reached about 88 billion in 2024.123 Farmed fish add an estimated 124 billion more each year, and wild-caught fish hundreds of billions beyond that.124 In 2012, a group of neuroscientists meeting at Cambridge declared that humans are not unique in possessing the neurological substrates of consciousness, and that mammals, birds, and even octopuses possess them too.125 In 2024, dozens of leading scientists signed the New York Declaration on Animal Consciousness, affirming strong scientific support for conscious experience in mammals and birds and at least a realistic possibility of it in all vertebrates, including fish, and many invertebrates. When that possibility exists, they wrote, it is irresponsible to ignore it.126 The United Kingdom wrote sentience into law in 2022, covering every vertebrate along with octopuses, crabs, and lobsters.127 In the same year nearly 500 moral philosophers from some 40 countries signed a declaration condemning animal exploitation as unjust and calling for the closing of slaughterhouses and the growth of plant-based agriculture.128

Notice that those two costs are not the same kind of cost. If a fully plant-based diet turns out to have been oversold, the price is a supplement, some monitoring, and the unresolved questions above: manageable, reversible, and borne by one person. If the environmental accounting turns out to have been right and we keep expanding animal agriculture, the price is cleared forest, emitted carbon, and species that do not come back. That asymmetry is why the second ledger carries more weight here than the first, and why it survives every uncertainty in the nutrition literature intact.

Put it together. We are the ape that can thrive on plants, whose healthiest documented diets lean on plants, whose planet is being reshaped by the animals we breed to eat, and who now knows, with as much certainty as science offers about other minds, that those animals can suffer. Veganism, as the Vegan Society defines it, is a way of living that seeks to exclude exploitation of and cruelty to animals “as far as is possible and practicable.”129 Nothing in our evolution requires it. That is exactly the point. Nothing in our evolution requires slaughterhouses either. And I should be plain about what is happening in this paragraph: the evidence gives a direction and declines to name a stopping point, so something else has to, and for me that something is an ethical judgment about what I am willing to have done on my behalf for a benefit the evidence cannot demonstrate. It is doing exactly the work the evidence can’t. Everything before this section stands without it. For the first time in the history of our lineage, the choice between them is a real choice, made by an animal that can weigh evidence and imagine the future.

Our ancestors ate what the world offered and lived long enough to hand us their genes. We get to ask a question they never could: what kind of world do we want to hand forward? A future built on compassion instead of killing, on fields instead of feedlots, is not a rejection of our nature. It is the fullest expression of the one thing that has always made us different: we can choose.

Companion pieces

This post covers the deep history. The specific arguments get their own treatment: the full evidence review, the ancestral diet argument, on what the Inuit and the Maasai actually prove, the future is in our hands, the B12 question, bone density, cholesterol from the ground up, and protein. On land and animals: rewilding the plate, the 86 percent meme, crop deaths, and the pesticide argument.

References

Links are given as DOI (persistent) links wherever possible. Many are open access.

1. Lascaux figure counts and regional reindeer dominance: French Ministry of Culture, Lascaux official site, archeologie.culture.gouv.fr/lascaux/en; see also Aujoulat, N. (2005). Lascaux: Movement, Space and Time. Abrams.
2. Williams, G. C. (1957). Pleiotropy, natural selection, and the evolution of senescence. Evolution, 11(4), 398–411. https://doi.org/10.1111/j.1558-5646.1957.tb02911.x; Medawar, P. B. (1952). An Unsolved Problem of Biology. H. K. Lewis.
3. Kirkwood, T. B. L. (1977). Evolution of ageing. Nature, 270, 301–304. https://doi.org/10.1038/270301a0
4. Gaillard, J.-M., & Lemaître, J.-F. (2017). The Williams’ legacy: A critical reappraisal of his nine predictions about the evolution of senescence. Evolution, 71(12), 2768–2785. https://doi.org/10.1111/evo.13379
5. Austad, S. N., & Hoffman, J. M. (2018). Is antagonistic pleiotropy ubiquitous in aging biology? Evolution, Medicine, and Public Health, 2018(1), 287–294. https://doi.org/10.1093/emph/eoy033
6. Byars, S. G., Huang, Q. Q., Gray, L.-A., et al. (2017). Genetic loci associated with coronary artery disease harbor evidence of selection and antagonistic pleiotropy. PLOS Genetics, 13(6), e1006328. https://doi.org/10.1371/journal.pgen.1006328
7. Long, E., & Zhang, J. (2023). Evidence for the role of selection for reproductively advantageous alleles in human aging. Science Advances, 9(49), eadh4990. https://doi.org/10.1126/sciadv.adh4990
8. Brigos-Barril, E., Vasallo, C., Farré, X., et al. (2026). Genetic trade-offs in fertility and longevity explain the maintenance of disease-associated alleles in humans. Nature Ecology & Evolution, 10(9), 1774–1785. https://doi.org/10.1038/s41559-026-03140-z
9. Kratzer, J. T., et al. (2014). Evolutionary history and metabolic insights of ancient mammalian uricases. PNAS, 111(10), 3763–3768. https://doi.org/10.1073/pnas.1320393111
10. Chen-Xu, M., Yokose, C., Rai, S. K., Pillinger, M. H., & Choi, H. K. (2019). Contemporary prevalence of gout and hyperuricemia in the United States and decadal trends: The National Health and Nutrition Examination Survey, 2007–2016. Arthritis & Rheumatology, 71(6), 991–999. https://doi.org/10.1002/art.40807
11. Clemente, F. J., Cardona, A., Inchley, C. E., et al. (2014). A selective sweep on a deleterious mutation in CPT1A in Arctic populations. American Journal of Human Genetics, 95(5), 584–589. https://doi.org/10.1016/j.ajhg.2014.09.016
12. Gessner, B. D., Wood, T., Johnson, M. A., Richards, C. S., & Koeller, D. M. (2016). Evidence for an association between infant mortality and homozygosity for the Arctic variant of carnitine palmitoyltransferase 1A. Genetics in Medicine, 18, 933–939. https://doi.org/10.1038/gim.2015.197
13. Fumagalli, M., et al. (2015). Greenlandic Inuit show genetic signatures of diet and climate adaptation. Science, 349(6254), 1343–1347. https://doi.org/10.1126/science.aab2319
14. Corbett, S., Courtiol, A., Lummaa, V., Moorad, J., & Stearns, S. (2018). The transition to modernity and chronic disease: Mismatch and natural selection. Nature Reviews Genetics, 19, 419–430. https://doi.org/10.1038/s41576-018-0012-3; see also Lieberman, D. E. (2013). The Story of the Human Body. Pantheon.
15. Tessier, A.-J., Wang, F., Korat, A. A., et al. (2025). Optimal dietary patterns for healthy aging. Nature Medicine, 31(5), 1644–1652. https://doi.org/10.1038/s41591-025-03570-5
16. Bui, L. P., Pham, T. T., Wang, F., et al. (2024). Planetary Health Diet Index and risk of total and cause-specific mortality in three prospective cohorts. American Journal of Clinical Nutrition, 120(1), 80–91. https://doi.org/10.1016/j.ajcnut.2024.03.019
17. Bouvard, V., Loomis, D., Guyton, K. Z., et al.; IARC Monograph Working Group. (2015). Carcinogenicity of consumption of red and processed meat. The Lancet Oncology, 16(16), 1599–1600. https://doi.org/10.1016/S1470-2045(15)00444-1
18. Ungar, P. S., Grine, F. E., & Teaford, M. F. (2008). Dental microwear and diet of the Plio-Pleistocene hominin Paranthropus boisei. PLOS ONE, 3(4), e2044. https://doi.org/10.1371/journal.pone.0002044
19. Sponheimer, M., et al. (2013). Isotopic evidence of early hominin diets. PNAS, 110(26), 10513–10518. https://doi.org/10.1073/pnas.1222579110
20. Richards, M. P., & Trinkaus, E. (2009). Isotopic evidence for the diets of European Neanderthals and early modern humans. PNAS, 106(38), 16034–16039. https://doi.org/10.1073/pnas.0903821106
21. Henry, A. G., Brooks, A. S., & Piperno, D. R. (2011). Microfossils in calculus demonstrate consumption of plants and cooked foods in Neanderthal diets (Shanidar III, Iraq; Spy I and II, Belgium). PNAS, 108(2), 486–491. https://doi.org/10.1073/pnas.1016868108
22. Weyrich, L. S., et al. (2017). Neanderthal behaviour, diet, and disease inferred from ancient DNA in dental calculus. Nature, 544, 357–361. https://doi.org/10.1038/nature21674
23. Sistiaga, A., Mallol, C., Galván, B., & Summons, R. E. (2014). The Neanderthal meal: A new perspective using faecal biomarkers. PLOS ONE, 9(6), e101045. https://doi.org/10.1371/journal.pone.0101045
24. Sussman, R. W. (1991). Primate origins and the evolution of angiosperms. American Journal of Primatology, 23(4), 209–223. https://doi.org/10.1002/ajp.1350230402
25. Drouin, G., Godin, J.-R., & Pagé, B. (2011). The genetics of vitamin C loss in vertebrates. Current Genomics, 12(5), 371–378. https://doi.org/10.2174/138920211796429736
26. Hornung, T. C., & Biesalski, H.-K. (2019). Glut-1 explains the evolutionary advantage of the loss of endogenous vitamin C-synthesis. Evolution, Medicine, and Public Health, 2019(1), 221–231. https://doi.org/10.1093/emph/eoz024
27. Surridge, A. K., Osorio, D., & Mundy, N. I. (2003). Evolution and selection of trichromatic vision in primates. Trends in Ecology & Evolution, 18(4), 198–205. https://doi.org/10.1016/S0169-5347(03)00012-0
28. Li, X., et al. (2005). Pseudogenization of a sweet-receptor gene accounts for cats’ indifference toward sugar. PLOS Genetics, 1(1), e3. https://doi.org/10.1371/journal.pgen.0010003
29. Milton, K. (1999). Nutritional characteristics of wild primate foods: Do the diets of our closest living relatives have lessons for us? Nutrition, 15(6), 488–498. https://doi.org/10.1016/S0899-9007(99)00078-7
30. Wrangham, R. W., Conklin-Brittain, N. L., & Hunt, K. D. (1998). Dietary response of chimpanzees and cercopithecines to seasonal variation in fruit abundance. I. Antifeedants. International Journal of Primatology, 19, 949–970. https://doi.org/10.1023/A:1020318102257
31. Jane Goodall Institute of Canada. Inside the lives of chimpanzees: Lessons from 65 years at Gombe. janegoodall.ca
32. Watts, D. P., & Mitani, J. C. (2002). Hunting behavior of chimpanzees at Ngogo, Kibale National Park, Uganda. International Journal of Primatology, 23, 1–28. https://doi.org/10.1023/A:1013270606320
33. Surbeck, M., & Hohmann, G. (2008). Primate hunting by bonobos at LuiKotale, Salonga National Park. Current Biology, 18(19), R906–R907. https://doi.org/10.1016/j.cub.2008.08.040
34. Knott, C. D. (1998). Changes in orangutan caloric intake, energy balance, and ketones in response to fluctuating fruit availability. International Journal of Primatology, 19, 1061–1079. https://doi.org/10.1023/A:1020330404983
35. Pruetz, J. D., & Bertolani, P. (2007). Savanna chimpanzees, Pan troglodytes verus, hunt with tools. Current Biology, 17(5), 412–417. https://doi.org/10.1016/j.cub.2006.12.042
36. Ungar, P. S. (2017). Evolution’s Bite: A Story of Teeth, Diet, and Human Origins. Princeton University Press. press.princeton.edu
37. White, T. D., et al. (2009). Ardipithecus ramidus and the paleobiology of early hominids. Science, 326(5949), 64, 75–86. https://doi.org/10.1126/science.1175802
38. Cerling, T. E., et al. (2013). Stable isotope-based diet reconstructions of Turkana Basin hominins. PNAS, 110(26), 10501–10506. https://doi.org/10.1073/pnas.1222568110
39. Wynn, J. G., et al. (2013). Diet of Australopithecus afarensis from the Pliocene Hadar Formation, Ethiopia. PNAS, 110(26), 10495–10500. https://doi.org/10.1073/pnas.1222559110
40. Lüdecke, T., et al. (2025). Australopithecus at Sterkfontein did not consume substantial mammalian meat. Science, 387(6731), 309–314. https://doi.org/10.1126/science.adq7315
41. Henry, A. G., et al. (2012). The diet of Australopithecus sediba. Nature, 487, 90–93. https://doi.org/10.1038/nature11185
42. Cerling, T. E., et al. (2011). Diet of Paranthropus boisei in the early Pleistocene of East Africa. PNAS, 108(23), 9337–9341. https://doi.org/10.1073/pnas.1104627108
43. Laden, G., & Wrangham, R. (2005). The rise of the hominids as an adaptive shift in fallback foods: Plant underground storage organs (USOs) and australopith origins. Journal of Human Evolution, 49(4), 482–498. https://doi.org/10.1016/j.jhevol.2005.05.007
44. Marlowe, F. W. (2010). The Hadza: Hunter-Gatherers of Tanzania. University of California Press. ucpress.edu
45. Aranguren, B., et al. (2018). Wooden tools and fire technology in the early Neanderthal site of Poggetti Vecchi (Italy). PNAS, 115(9), 2054–2059. https://doi.org/10.1073/pnas.1716068115
46. Harmand, S., et al. (2015). 3.3-million-year-old stone tools from Lomekwi 3, West Turkana, Kenya. Nature, 521, 310–315. https://doi.org/10.1038/nature14464
47. Plummer, T. W., et al. (2023). Expanded geographic distribution and dietary strategies of the earliest Oldowan hominins and Paranthropus. Science, 379(6632), 561–566. https://doi.org/10.1126/science.abo7452
48. Bramble, D. M., & Lieberman, D. E. (2004). Endurance running and the evolution of Homo. Nature, 432, 345–352. https://doi.org/10.1038/nature03052
49. Ferraro, J. V., et al. (2013). Earliest archaeological evidence of persistent hominin carnivory. PLOS ONE, 8(4), e62174. https://doi.org/10.1371/journal.pone.0062174
50. Thompson, J. C., Carvalho, S., Marean, C. W., & Alemseged, Z. (2019). Origins of the human predatory pattern: The transition to large-animal exploitation by early hominins. Current Anthropology, 60(1), 1–23. https://doi.org/10.1086/701477
51. Aiello, L. C., & Wheeler, P. (1995). The expensive-tissue hypothesis: The brain and the digestive system in human and primate evolution. Current Anthropology, 36(2), 199–221. https://doi.org/10.1086/204350
52. Navarrete, A., van Schaik, C. P., & Isler, K. (2011). Energetics and the evolution of human brain size. Nature, 480, 91–93. https://doi.org/10.1038/nature10629
53. Kuzawa, C. W., et al. (2014). Metabolic costs and evolutionary implications of human brain development. PNAS, 111(36), 13010–13015. https://doi.org/10.1073/pnas.1323099111
54. Hardy, K., Brand-Miller, J., Brown, K. D., Thomas, M. G., & Copeland, L. (2015). The importance of dietary carbohydrate in human evolution. The Quarterly Review of Biology, 90(3), 251–268. https://doi.org/10.1086/682587
55. Ballard, O., & Morrow, A. L. (2013). Human milk composition: Nutrients and bioactive factors. Pediatric Clinics of North America, 60(1), 49–74. https://doi.org/10.1016/j.pcl.2012.10.002
56. Wrangham, R. (2009). Catching Fire: How Cooking Made Us Human. Basic Books. See also Wrangham, R. (2017). Control of fire in the Paleolithic: Evaluating the cooking hypothesis. Current Anthropology, 58(S16), S303–S313. https://doi.org/10.1086/692113
57. Zohar, I., et al. (2022). Evidence for the cooking of fish 780,000 years ago at Gesher Benot Ya’aqov, Israel. Nature Ecology & Evolution, 6, 2016–2028. https://doi.org/10.1038/s41559-022-01910-z
58. Melamed, Y., et al. (2016). The plant component of an Acheulian diet at Gesher Benot Ya’aqov, Israel. PNAS, 113(51), 14674–14679. https://doi.org/10.1073/pnas.1607872113
59. Zink, K. D., & Lieberman, D. E. (2016). Impact of meat and Lower Palaeolithic food processing techniques on chewing in humans. Nature, 531, 500–503. https://doi.org/10.1038/nature16990
60. Stedman, H. H., et al. (2004). Myosin gene mutation correlates with anatomical changes in the human lineage. Nature, 428, 415–418. https://doi.org/10.1038/nature02358
61. Jaouen, K., et al. (2019). Exceptionally high δ15N values in collagen single amino acids confirm Neandertals as high-trophic level carnivores. PNAS, 116(11), 4928–4933. https://doi.org/10.1073/pnas.1814087116
62. Jaouen, K., et al. (2022). A Neandertal dietary conundrum: Insights provided by tooth enamel Zn isotopes from Gabasa, Spain. PNAS, 119(43), e2109315119. https://doi.org/10.1073/pnas.2109315119
63. Kindler, L., et al. (2025). Large-scale processing of within-bone nutrients by Neanderthals, 125,000 years ago. Science Advances, 11(27), eadv1257. https://doi.org/10.1126/sciadv.adv1257
64. Hardy, K., et al. (2012). Neanderthal medics? Evidence for food, cooking, and medicinal plants entrapped in dental calculus. Naturwissenschaften, 99, 617–626. https://doi.org/10.1007/s00114-012-0942-0
65. Kabukcu, C., et al. (2022). Cooking in caves: Palaeolithic carbonised plant food remains from Franchthi and Shanidar. Antiquity, 97(391), 12–28. https://doi.org/10.15184/aqy.2022.143
66. Wadley, L., et al. (2020). Cooked starchy rhizomes in Africa 170 thousand years ago. Science, 367(6473), 87–91. https://doi.org/10.1126/science.aaz5926
67. Larbey, C., et al. (2019). Cooked starchy food in hearths ca. 120 kya and 65 kya (MIS 5e and MIS 4) from Klasies River Cave, South Africa. Journal of Human Evolution, 131, 210–227. https://doi.org/10.1016/j.jhevol.2019.03.015
68. Florin, S. A., et al. (2020). The first Australian plant foods at Madjedbebe, 65,000–53,000 years ago. Nature Communications, 11, 924. https://doi.org/10.1038/s41467-020-14723-0
69. Revedin, A., et al. (2010). Thirty thousand-year-old evidence of plant food processing. PNAS, 107(44), 18815–18819. https://doi.org/10.1073/pnas.1006993107
70. Snir, A., et al. (2015). The origin of cultivation and proto-weeds, long before Neolithic farming. PLOS ONE, 10(7), e0131422. https://doi.org/10.1371/journal.pone.0131422
71. Arranz-Otaegui, A., et al. (2018). Archaeobotanical evidence reveals the origins of bread 14,400 years ago in northeastern Jordan. PNAS, 115(31), 7925–7930. https://doi.org/10.1073/pnas.1801071115
72. Humphrey, L. T., et al. (2014). Earliest evidence for caries and exploitation of starchy plant foods in Pleistocene hunter-gatherers from Morocco. PNAS, 111(3), 954–959. https://doi.org/10.1073/pnas.1318176111
73. Gilbert, M. T. P., et al. (2008). DNA from pre-Clovis human coprolites in Oregon, North America. Science, 320(5877), 786–789. https://doi.org/10.1126/science.1154116
74. The Metropolitan Museum of Art, Heilbrunn Timeline of Art History. Chauvet Cave (ca. 30,000 B.C.). metmuseum.org
75. Lewis-Williams, D. (2002). The Mind in the Cave: Consciousness and the Origins of Art. Thames & Hudson.
76. Bacon, B., et al. (2023). An Upper Palaeolithic proto-writing system and phenological calendar. Cambridge Archaeological Journal, 33(3), 371–389. https://doi.org/10.1017/S0959774322000415
77. Brumm, A., et al. (2021). Oldest cave art found in Sulawesi. Science Advances, 7(3), eabd4648. https://doi.org/10.1126/sciadv.abd4648
78. Aubert, M., et al. (2019). Earliest hunting scene in prehistoric art. Nature, 576, 442–445. https://doi.org/10.1038/s41586-019-1806-y; re-dating: Oktaviana, A. A., et al. (2024). Narrative cave art in Indonesia by 51,200 years ago. Nature, 631, 814–818. https://doi.org/10.1038/s41586-024-07541-7
79. Fujita, M., et al. (2016). Advanced maritime adaptation in the western Pacific coastal region extends back to 35,000–30,000 years before present. PNAS, 113(40), 11184–11189. https://doi.org/10.1073/pnas.1607857113
80. O’Connor, S., Ono, R., & Clarkson, C. (2011). Pelagic fishing at 42,000 years before the present and the maritime skills of modern humans. Science, 334(6059), 1117–1121. https://doi.org/10.1126/science.1207703
81. Thieme, H. (1997). Lower Palaeolithic hunting spears from Germany. Nature, 385, 807–810. https://doi.org/10.1038/385807a0; re-dating: Hutson, J. M., et al. (2025). Revised age for Schöningen hunting spears indicates intensification of Neanderthal cooperative behavior around 200,000 years ago. Science Advances, 11(19), eadv0752. https://doi.org/10.1126/sciadv.adv0752
82. Larsen, C. S. (2006). The agricultural revolution as environmental catastrophe: Implications for health and lifestyle in the Holocene. Quaternary International, 150(1), 12–20. https://doi.org/10.1016/j.quaint.2006.01.004
83. Perry, G. H., et al. (2007). Diet and the evolution of human amylase gene copy number variation. Nature Genetics, 39, 1256–1260. https://doi.org/10.1038/ng2123
84. Yilmaz, F., et al. (2024). Reconstruction of the human amylase locus reveals ancient duplications seeding modern-day variation. Science, 386(6724), eadn0609. https://doi.org/10.1126/science.adn0609
85. Bolognini, D., et al. (2024). Recurrent evolution and selection shape structural diversity at the amylase locus. Nature, 634, 617–625. https://doi.org/10.1038/s41586-024-07911-1
86. Axelsson, E., et al. (2013). The genomic signature of dog domestication reveals adaptation to a starch-rich diet. Nature, 495, 360–364. https://doi.org/10.1038/nature11837
87. Evershed, R. P., et al. (2022). Dairying, diseases and the evolution of lactase persistence in Europe. Nature, 608, 336–345. https://doi.org/10.1038/s41586-022-05010-7
88. Buckley, M. T., Racimo, F., Allentoft, M. E., et al. (2017). Selection in Europeans on fatty acid desaturases associated with dietary changes. Molecular Biology and Evolution, 34(6), 1307–1318. https://doi.org/10.1093/molbev/msx103
89. Tishkoff, S. A., Reed, F. A., Ranciaro, A., et al. (2007). Convergent adaptation of human lactase persistence in Africa and Europe. Nature Genetics, 39, 31–40. https://doi.org/10.1038/ng1946
90. Schlebusch, C. M., Sjödin, P., Skoglund, P., & Jakobsson, M. (2013). Stronger signal of recent selection for lactase persistence in Maasai than in Europeans. European Journal of Human Genetics, 21, 550–553. https://doi.org/10.1038/ejhg.2012.199
91. Marlowe, F. W., et al. (2014). Honey, Hadza, hunter-gatherers, and human evolution. Journal of Human Evolution, 71, 119–128. https://doi.org/10.1016/j.jhevol.2014.03.006
92. Crittenden, A. N. (2011). The importance of honey consumption in human evolution. Food and Foodways, 19(4), 257–273. https://doi.org/10.1080/07409710.2011.630618
93. Lee, R. B. (1968). What hunters do for a living, or, how to make out on scarce resources. In R. B. Lee & I. DeVore (Eds.), Man the Hunter (pp. 30–48). Aldine.
94. Bjerregaard, P., Young, T. K., & Hegele, R. A. (2003). Low incidence of cardiovascular disease among the Inuit: What is the evidence? Atherosclerosis, 166(2), 351–357. https://doi.org/10.1016/S0021-9150(02)00364-7
95. Fodor, J. G., Helis, E., Yazdekhasti, N., & Vohnout, B. (2014). “Fishing” for the origins of the “Eskimos and heart disease” story. Canadian Journal of Cardiology, 30(8), 864–868. https://doi.org/10.1016/j.cjca.2014.04.007
96. Thompson, R. C., Allam, A. H., Lombardi, G. P., et al. (2013). Atherosclerosis across 4000 years of human history: The Horus study of four ancient populations. The Lancet, 381(9873), 1211–1222. https://doi.org/10.1016/S0140-6736(13)60598-X
97. Cordain, L., et al. (2000). Plant-animal subsistence ratios and macronutrient energy estimations in worldwide hunter-gatherer diets. American Journal of Clinical Nutrition, 71(3), 682–692. https://doi.org/10.1093/ajcn/71.3.682
98. Pontzer, H., & Wood, B. M. (2021). Effects of evolution, ecology, and economy on human diet: Insights from hunter-gatherers and other small-scale societies. Annual Review of Nutrition, 41, 363–385. https://doi.org/10.1146/annurev-nutr-111120-105520
99. Kaplan, H., et al. (2017). Coronary atherosclerosis in indigenous South American Tsimane: A cross-sectional cohort study. The Lancet, 389(10080), 1730–1739. https://doi.org/10.1016/S0140-6736(17)30752-3
100. Ben-Dor, M., Sirtoli, R., & Barkai, R. (2021). The evolution of the human trophic level during the Pleistocene. American Journal of Physical Anthropology, 175(S72), 27–56. https://doi.org/10.1002/ajpa.24247
101. Beasley, D. E., Koltz, A. M., Lambert, J. E., Fierer, N., & Dunn, R. R. (2015). The evolution of stomach acidity and its relevance to the human microbiome. PLOS ONE, 10(7), e0134116. https://doi.org/10.1371/journal.pone.0134116
102. Bonhommeau, S., Dubroca, L., Le Pape, O., et al. (2013). Eating up the world’s food web and the human trophic level. PNAS, 110(51), 20617–20620. https://doi.org/10.1073/pnas.1305827110
103. Speth, J. D., & Spielmann, K. A. (1983). Energy source, protein metabolism, and hunter-gatherer subsistence strategies. Journal of Anthropological Archaeology, 2(1), 1–31. https://doi.org/10.1016/0278-4165(83)90006-5
104. Bilsborough, S., & Mann, N. (2006). A review of issues of dietary protein intake in humans. International Journal of Sport Nutrition and Exercise Metabolism, 16(2), 129–152. https://doi.org/10.1123/ijsnem.16.2.129
105. Baker, T. G. (1963). A quantitative and cytological study of germ cells in human ovaries. Proceedings of the Royal Society B, 158(972), 417–433. https://doi.org/10.1098/rspb.1963.0055
106. Heijmans, B. T., Tobi, E. W., Stein, A. D., et al. (2008). Persistent epigenetic differences associated with prenatal exposure to famine in humans. PNAS, 105(44), 17046–17049. https://doi.org/10.1073/pnas.0806560105
107. Ng, S.-F., Lin, R. C. Y., Laybutt, D. R., Barres, R., Owens, J. A., & Morris, M. J. (2010). Chronic high-fat diet in fathers programs β-cell dysfunction in female rat offspring. Nature, 467, 963–966. https://doi.org/10.1038/nature09491
108. Donkin, I., Versteyhe, S., Ingerslev, L. R., et al. (2016). Obesity and bariatric surgery drive epigenetic variation of spermatozoa in humans. Cell Metabolism, 23(2), 369–378. https://doi.org/10.1016/j.cmet.2015.11.004
109. Kaati, G., Bygren, L. O., & Edvinsson, S. (2002). Cardiovascular and diabetes mortality determined by nutrition during parents’ and grandparents’ slow growth period. European Journal of Human Genetics, 10, 682–688. https://doi.org/10.1038/sj.ejhg.5200859
110. Pembrey, M. E., Bygren, L. O., Kaati, G., et al. (2006). Sex-specific, male-line transgenerational responses in humans. European Journal of Human Genetics, 14, 159–166. https://doi.org/10.1038/sj.ejhg.5201538
111. Vågerö, D., Pinger, P. R., Aronsson, V., & van den Berg, G. J. (2018). Paternal grandfather’s access to food predicts all-cause and cancer mortality in grandsons. Nature Communications, 9, 5124. https://doi.org/10.1038/s41467-018-07617-9
112. Horsthemke, B. (2018). A critical view on transgenerational epigenetic inheritance in humans. Nature Communications, 9, 2973. https://doi.org/10.1038/s41467-018-05445-5
113. Milot, E., Moreau, C., Gagnon, A., et al. (2017). Mother’s curse neutralizes natural selection against a human genetic disease over three centuries. Nature Ecology & Evolution, 1, 1400–1406. https://doi.org/10.1038/s41559-017-0276-6
114. Landry, M. J., Ward, C. P., Cunanan, K. M., et al. (2023). Cardiometabolic effects of omnivorous vs vegan diets in identical twins: A randomized clinical trial. JAMA Network Open, 6(11), e2344457. https://doi.org/10.1001/jamanetworkopen.2023.44457
115. Dwaraka, V. B., Aronica, L., Carreras-Gallo, N., et al. (2024). Unveiling the epigenetic impact of vegan vs. omnivorous diets on aging: Insights from the Twins Nutrition Study (TwiNS). BMC Medicine, 22, 301. https://doi.org/10.1186/s12916-024-03513-w
116. Poore, J., & Nemecek, T. (2018). Reducing food’s environmental impacts through producers and consumers. Science, 360(6392), 987–992. https://doi.org/10.1126/science.aaq0216
117. Scarborough, P., Clark, M., Cobiac, L., et al. (2023). Vegans, vegetarians, fish-eaters and meat-eaters in the UK show discrepant environmental impacts. Nature Food, 4, 565–574. https://doi.org/10.1038/s43016-023-00795-w
118. Xu, X., Sharma, P., Shu, S., et al. (2021). Global greenhouse gas emissions from animal-based foods are twice those of plant-based foods. Nature Food, 2, 724–732. https://doi.org/10.1038/s43016-021-00358-x
119. Willett, W., Rockström, J., Loken, B., et al. (2019). Food in the Anthropocene: The EAT–Lancet Commission on healthy diets from sustainable food systems. The Lancet, 393(10170), 447–492. https://doi.org/10.1016/S0140-6736(18)31788-4
120. Rockström, J., Thilsted, S. H., Willett, W. C., et al. (2025). The EAT–Lancet Commission on healthy, sustainable, and just food systems. The Lancet, 406, 1625–1700. https://doi.org/10.1016/S0140-6736(25)01201-2
121. Gu, X., Bui, L. P., Wang, F., Wang, D. D., Springmann, M., & Willett, W. C. (2024). Global adherence to a healthy and sustainable diet and potential reduction in premature death. PNAS, 121(50), e2319008121. https://doi.org/10.1073/pnas.2319008121
122. Van Boeckel, T. P., Pires, J., Silvester, R., et al. (2019). Global trends in antimicrobial resistance in animals in low- and middle-income countries. Science, 365(6459), eaaw1944. https://doi.org/10.1126/science.aaw1944; see also Van Boeckel, T. P., et al. (2015). Global trends in antimicrobial use in food animals. PNAS, 112(18), 5649–5654. https://doi.org/10.1073/pnas.1503141112
123. Our World in Data (2024). More than 80 billion land animals are slaughtered for meat every year (FAO data). ourworldindata.org; Faunalytics (2026). Global Animal Slaughter Statistics & Charts (FAOSTAT 2024 data). faunalytics.org
124. Mood, A., Lara, E., Boyland, N. K., & Brooke, P. (2023). Estimating global numbers of farmed fishes killed for food annually from 1990 to 2019. Animal Welfare, 32, e12. https://doi.org/10.1017/awf.2023.4
125. Low, P., et al. (2012). The Cambridge Declaration on Consciousness. Francis Crick Memorial Conference, Churchill College, Cambridge, 7 July 2012. fcmconference.org
126. Andrews, K., Birch, J., Sebo, J., et al. (2024). The New York Declaration on Animal Consciousness. New York University, 19 April 2024. sites.google.com/nyu.edu/nydeclaration
127. UK Public General Acts (2022). Animal Welfare (Sentience) Act 2022, c. 22. legislation.gov.uk
128. Gibert, M., Giroux, V., & Jaquet, F., et al. (2022). Montreal Declaration on Animal Exploitation, 4 October 2022. lamorce.co
129. The Vegan Society. Definition of veganism. vegansociety.com

Further reading: Pontzer, H. (2021). Burn. Avery. | Lieberman, D. E. (2013). The Story of the Human Body. Pantheon. | Ungar, P. S. (2017). Evolution’s Bite. Princeton. | Wrangham, R. (2009). Catching Fire. Basic Books. | Smithsonian National Museum of Natural History, Human Origins Program: humanorigins.si.edu

Last updated: October 6, 2026 at 12:40 PM ET