Fire, tools and cooking

Foundations

An organ you cannot afford

Your brain is about two percent of your body mass. At rest it consumes roughly twenty percent of your energy. It is, per gram, about sixteen times as expensive as muscle, and it never switches off.

Now the problem. A primate’s total energy budget is set mostly by body size, and human basal metabolic rate is close to what a mammal of our size should have. We did not solve the cost of a large brain by burning more fuel overall.

So something else had to give.

The answer: your stomach

Compare a human with a chimpanzee of similar body mass and the difference is not only in the head. Our digestive tract is markedly smaller, particularly the colon, where a great deal of the fermentation of tough plant material happens in apes.

Leslie Aiello and Peter Wheeler proposed in 1995 that this is not a coincidence. Both brain and gut are metabolically expensive tissues. If the budget is fixed, expanding one requires shrinking the other. The brain was paid for by the stomach.

That reframes the question. You cannot shrink a gut unless the food arriving in it is already easier to digest. And that is what tools and fire are for. They are external digestion: work done outside the body that a gut would otherwise have to do inside it.

Tools are older than us

Stone tools were long assumed to be a defining human invention, which is why Homo habilis was named “handy man”. That has not survived.

In 2015, tools from Lomekwi in Kenya were dated to 3.3 million years ago. Our genus does not appear until around 2.4 million years ago. The first toolmakers were not members of Homo. They were australopithecines or something close, with brains the size of a chimpanzee’s, and they were deliberately striking stone to make edges.

Tool use itself is not even unusual. Chimpanzees fish for termites with stripped twigs, crack nuts with stone hammers, and pass these techniques on locally. Crows bend wire. Otters use anvils. What is unusual is not making tools. It is making tools that get better across generations, and that took an extraordinarily long time to arrive.

What a sharp flake actually buys you

The Oldowan toolkit, from about 2.6 million years ago, looks unimpressive: a cobble struck to detach flakes, and the flakes are the point. But a sharp edge changes what an animal can eat.

It cuts through hide, which hominins have no teeth or claws to do. It removes meat from bone. It breaks bone for marrow, which is dense fat available to whoever has a hammerstone and no competition, because most carnivores cannot get into a femur.

Cut marks on animal bones at 2.6 million years and older show hominins accessing carcasses. Whether by hunting or by scavenging is argued, and either way it means access to a food far richer in energy and protein than leaves and fruit.

And the same edges cut plant food into pieces, which sounds trivial and is not: reducing food to small pieces before it reaches the mouth removes an enormous amount of chewing. A chimpanzee spends around six hours a day chewing. Humans spend under one.

The handaxe, and a very long silence

Around 1.76 million years ago something new appears: the Acheulean handaxe. Teardrop-shaped, worked on both faces, symmetrical.

That symmetry matters, because it is not what the stone wants to do. An Oldowan flake follows the rock. A handaxe imposes a shape that exists first in the maker’s head and is held there across dozens of strikes, each of which can ruin it. That is a mental template, and it is the first clear archaeological evidence of one.

They are also everywhere. Africa, the Middle East, Europe and Asia, including India: Attirampakkam in Tamil Nadu has Acheulean tools dated to about 1.5 million years ago, which pushed the Indian record back by around a million years and showed that hominins reached South Asia far earlier than had been assumed.

And then here is the fact that should stop you. The handaxe persists, with limited change, for more than one and a half million years. Across three continents. Across at least three hominin species. Tens of thousands of generations, all making broadly the same object.

Compare that with the last five hundred years. Something that could hold a design steady for a million and a half years was doing something very different from what we do. The next lesson but one is about what changed.

Fire

Fire is a bigger prize than any tool, and the evidence for when it was controlled is genuinely messy.

Burnt material at a site is not evidence of control. Wildfires happen; lightning strikes; bone and sediment discolour for many reasons. What is needed is evidence of fire that was in one place repeatedly, inside a shelter, at temperatures and in configurations that natural burning does not produce.

The strongest early case is Wonderwerk Cave in South Africa, where microscopic analysis of sediments found ash from burnt plant material and heated bone fragments about thirty metres inside the cave, in layers around one million years old. Thirty metres in is not where a wildfire reaches.

Even so, evidence for habitual fire use, meaning hearths at most sites rather than a few, does not become common until a few hundred thousand years ago. That gap is the central difficulty for the argument that follows.

The cooking hypothesis

Richard Wrangham’s proposal is that cooking is the thing that made us, and it is worth taking seriously because the chemistry is not in dispute.

Heat gelatinises starch, opening its packed structure so that digestive enzymes can reach it. Heat denatures protein, unfolding it so enzymes can cut it. Heat softens tough tissue so that chewing costs far less. The result is that the same food yields substantially more usable energy cooked than raw, and costs less to process.

Wrangham’s argument is that this is what allowed the changes visible in Homo erectus: smaller teeth, a smaller jaw, a smaller gut, a bigger brain, and a body that could not survive on the raw diet its ancestors ate.

The supporting observation is uncomfortable and direct. Humans cannot thrive on raw food. Studies of people on strict raw diets in modern conditions, with cultivated high-calorie plants, blenders and no predators, still find low body mass and a high rate of disrupted menstrual function. We are obligate cooks. No other animal is.

The objection is the timing. Erectus appears around two million years ago; secure evidence for controlled fire is much later. Either the fire evidence is missing, which is plausible since open-air hearths preserve badly, or the anatomical changes were driven first by non-thermal processing, cutting and pounding, with cooking arriving later and consolidating the gains.

That second version is currently the more cautious reading, and it does not weaken the underlying point. Whether by blade or by flame, humans moved digestion outside the body, and that is what paid for the brain.

What fire does beyond food

Fire also does three things that have nothing to do with digestion, and each matters for later lessons.

It extends the day, adding usable hours after dark. It provides warmth, which widens the range of climates a tropical animal can occupy. And it creates a place where people sit together in the dark, facing each other, doing nothing urgent. Ethnographic work among foragers has found that daytime talk is largely practical while firelit talk turns to stories, other people and the absent.

If collective learning needs a medium, a fire with people around it is a reasonable candidate for the first one.

Deeper

The arithmetic of expensive tissue

Aiello and Wheeler’s argument compares observed organ masses against those expected for a primate of human body size. Human brain mass is roughly double the expectation; the gut is roughly 60 percent of it, with the reduction concentrated in the colon. Because brain, gut, heart, liver and kidneys together account for the great majority of basal metabolism, and heart, liver and kidney masses are close to expectation, the gut is the only candidate for the offset. The hypothesis has been criticised on the grounds that the correlation between brain and gut size does not hold across all primates and other mammals, and later work has emphasised additional factors including reduced locomotor costs, fat storage and higher total energy expenditure in humans than in other apes. The energetic framing has survived even where the strict trade-off has been questioned.

Reading fire from sediment

Distinguishing anthropogenic fire from natural burning requires micromorphology: thin sections of intact sediment examined for ashed plant material, aggregated burnt bone, and mineral alterations indicating specific temperature ranges. At Wonderwerk, Fourier-transform infrared spectroscopy indicated bone heated to above 500 degrees Celsius and ashed plant remains preserved in situ rather than transported, deep inside the cave. The absence of comparable evidence at most early sites is genuinely ambiguous between absence of fire and absence of preservation, since open-air combustion features rarely survive.

Why the Acheulean’s stasis is a real problem

The persistence of a single tool form across more than a million years and three continents is not explained by cognitive limitation alone, since the makers were plainly capable of the design. Proposed explanations include the handaxe being a highly effective general-purpose tool with little pressure to change, low population densities limiting the transmission and accumulation of variants, and the possibility that much variation existed in perishable materials that have not survived. The demographic explanation connects directly to the argument in lesson 6.5, where population size determines whether innovations are retained or lost.

Attirampakkam and the Indian sequence

The Attirampakkam dates were obtained by cosmogenic nuclide burial dating of quartz in the artefact-bearing sediments, giving a minimum age near 1.07 million years and an estimate around 1.5 million years for the earliest Acheulean levels. The result was significant because Indian Acheulean assemblages had previously been assumed to be considerably younger, and it establishes South Asia as part of the early Acheulean world rather than a late recipient of it. The same site’s later sequence has been used to argue for an early Middle Palaeolithic transition in India, which remains debated.

Frontier

When fire became habitual

The gap between the earliest plausible fire evidence and the earliest habitual use is the field’s main open question, and it matters because the cooking hypothesis requires the earlier date. Systematic micromorphological sampling at Lower Palaeolithic sites, and better methods for detecting combustion residues in open-air contexts, are the routes to closing it.

Who made the Lomekwi tools

No hominin fossils are directly associated with the Lomekwi assemblage. Candidates include Kenyanthropus platyops, known from the same region and time, and Australopithecus afarensis. The identification of the maker would considerably change how the origins of tool use are understood, and the artefacts’ status has itself been questioned by some researchers.

Transcript and shot list

Target runtime about 8 minutes. Built as a budget problem: state the energy shortfall, show the three ways it was solved outside the body, then sit with the long silence that follows. Shot types: ANIM is rendered from code with Manim, AI is a generated shot from Higgsfield with the shared style reference, REAL is agency imagery with credit, CAM is the narrator.

[00:00] S01 · ANIM · A human body with the brain highlighted: 2 percent of mass. Then an energy meter: 20 percent of resting expenditure. Then a comparison bar: per gram, sixteen times the cost of muscle.

Your brain is about two percent of your body. At rest it eats roughly twenty percent of your energy. Per gram it costs about sixteen times what muscle does, and it never switches off. Now the problem: human resting metabolic rate is close to what a mammal our size should have. We did not pay for a big brain by burning more fuel overall. So something else had to give.

[00:35] S02 · ANIM · A human and a chimpanzee of similar mass side by side; the human brain enlarged, the human gut and particularly the colon markedly reduced; a balance scale tipping.

Compare a human with a chimpanzee of similar body mass and the difference is not only in the head. Our digestive tract is much smaller, especially the colon, which is where an ape ferments tough plant material. In 1995 Leslie Aiello and Peter Wheeler proposed that this is no coincidence. Brains and guts are both expensive. If the budget is fixed, growing one means shrinking the other. Your brain was paid for by your stomach.

[01:10] S03 · ANIM · The reframed question: you cannot shrink a gut unless the food arriving is already easier to digest; then three external solutions appear: cutting, pounding, cooking.

Which reframes the question entirely. You cannot shrink a gut unless the food arriving in it is already easier to digest. And that is what tools and fire are for. They are external digestion: work done outside the body that a gut would otherwise do inside it.

[01:35] S04 · REAL · A Lomekwi core and flake (research image; confirm credit line). S05 · ANIM · A timeline: Lomekwi tools at 3.3 Ma, the genus Homo appearing at 2.4 Ma, with the gap highlighted.

And the tools are older than we are. Stone tools were long treated as the defining human invention, which is why Homo habilis was named handy man. That has not survived. In 2015, tools from Lomekwi in Kenya were dated to three point three million years ago. Our genus does not appear until around two point four. The first toolmakers were not members of Homo. They had chimpanzee-sized brains, and they were deliberately striking stone to make edges.

[02:15] S06 · ANIM · A chimpanzee termite-fishing, a crow bending wire, an otter using an anvil; then a caption: making tools is not the rare thing.

Tool use is not even unusual. Chimpanzees fish for termites with stripped twigs and crack nuts with stone hammers. Crows bend wire. Otters use anvils. The rare thing is not making tools. It is making tools that get better across generations, and that took an extraordinarily long time to show up.

[02:45] S07 · ANIM · An Oldowan flake cutting hide, stripping meat, and cracking a femur for marrow; a chewing-time comparison: chimpanzee six hours a day, human under one.

The Oldowan toolkit looks unimpressive: a cobble struck to detach flakes, and the flakes are the point. But a sharp edge changes what an animal can eat. It cuts hide, which we have no teeth or claws to do. It strips meat from bone. It breaks bone for marrow, which is dense fat that almost no carnivore can reach. And the same edges cut plant food into pieces before it reaches the mouth, which sounds trivial and is not. A chimpanzee spends about six hours a day chewing. You spend under one.

[03:25] S08 · ANIM · An Oldowan flake following the rock’s natural fracture; then an Acheulean handaxe imposing a symmetrical shape held across dozens of strikes; a mental template shown as a ghost outline before the first blow.

Then, around one point seven six million years ago, something new. The Acheulean handaxe. Teardrop-shaped, worked on both faces, symmetrical. And that symmetry matters, because it is not what the stone wants to do. An Oldowan flake follows the rock. A handaxe imposes a shape that exists first in the maker’s head and has to be held there through dozens of strikes, any of which can ruin it. That is a mental template, and it is the first clear archaeological evidence of one.

[04:05] S09 · REAL · Acheulean handaxes from Attirampakkam (Sharma Centre for Heritage Education; confirm credit line). S10 · ANIM · A world map of Acheulean sites including Tamil Nadu; a duration bar of 1.5 million years across three continents and three species.

And they are everywhere. Africa, the Middle East, Europe, Asia. Including India: Attirampakkam in Tamil Nadu has Acheulean tools dated to about one and a half million years ago, which pushed the Indian record back by roughly a million years and put South Asia in the early Acheulean world rather than at the end of it. And now the fact that should stop you. This design persists, with limited change, for more than one and a half million years. Across three continents. Across at least three species of human. Tens of thousands of generations, all making broadly the same object. Compare that with the last five hundred years. Whatever they were doing, it was not what we do.

[04:55] S11 · ANIM · Burnt material at a site with three innocent explanations; then the criteria for control: repeated, in one place, inside shelter, at temperatures nature does not produce.

Fire is a bigger prize than any tool, and the evidence is genuinely messy. Burnt material at a site is not evidence of control. Wildfires happen. Lightning strikes. Bone and sediment discolour for lots of reasons. What you need is fire in one place, repeatedly, inside a shelter, at temperatures and in configurations natural burning does not produce.

[05:25] S12 · REAL · The interior of Wonderwerk Cave (research image; confirm credit line). S13 · ANIM · A cave cross-section, the sampling point 30 metres in; ashed plant material and bone heated above 500 °C identified in thin section.

The strongest early case is Wonderwerk Cave in South Africa. Microscopic analysis of intact sediment found ashed plant material and burnt bone about thirty metres inside the cave, in layers around a million years old. Thirty metres in is not where a wildfire reaches. Even so, evidence for habitual fire, meaning hearths at most sites rather than a handful, does not become common until a few hundred thousand years ago. And that gap is the main problem for what comes next.

[06:05] S14 · ANIM · Starch granules gelatinising under heat; protein unfolding; an energy-yield comparison raw versus cooked; then Homo erectus with smaller teeth, smaller jaw, smaller gut, bigger brain.

Richard Wrangham’s proposal is that cooking is what made us, and the chemistry is not in dispute. Heat gelatinises starch, opening its packed structure so enzymes can get in. Heat unfolds protein so enzymes can cut it. Heat softens tough tissue so chewing costs less. The same food yields substantially more usable energy cooked than raw. And his argument is that this is what allowed the changes you see in Homo erectus: smaller teeth, smaller jaw, smaller gut, bigger brain.

[06:45] S15 · ANIM · A modern raw-food study: cultivated high-calorie plants, blenders, no predators; outcomes showing low body mass and disrupted reproductive function; a caption: obligate cooks.

And the supporting evidence is uncomfortable and direct. Humans cannot thrive on raw food. Studies of people on strict raw diets in modern conditions, with cultivated high-calorie plants, blenders and no predators, still find low body mass and a high rate of disrupted menstrual function. We are obligate cooks. No other animal is. The objection is timing: erectus appears around two million years ago and secure fire evidence is much later. Either the evidence is missing, which is plausible because open-air hearths preserve badly, or cutting and pounding drove the anatomy first and cooking consolidated it later. That second version is the more cautious reading, and it does not change the point. Whether by blade or by flame, digestion moved outside the body, and that is what paid for the brain.

[07:30] S16 · AI into ANIM · People sitting around a fire after dark, faces lit, nothing urgent happening; then a comparison of daytime talk and firelit talk.

And fire does three things that have nothing to do with digestion. It extends the day. It provides warmth, which lets a tropical animal live in cold places. And it makes somewhere people sit together in the dark, facing each other, with nothing urgent to do. Studies of foragers have found that daytime talk is largely practical, and firelit talk turns to stories, to other people, to the absent. If collective learning needed a medium, that is a reasonable candidate for the first one. Next lesson: the evidence that people had started to think in symbols.

Shot list

One AI shot, twelve code-rendered shots, three research images. Estimated Higgsfield use: 2 to 4 generations including retries.