The manifesto uses about a dozen facts that sound false the first time you hear them: that a hunting people's favourite food is sugar, that walking fifteen kilometres a day doesn't make you burn more than an office worker, that forty thousand years ago there was a ritual designed to stop an athlete from feeling important.
They're all true, and this is where you check them.
But the opposite is needed too. A manifesto built on studies stretched further than they hold is a manifesto that collapses in front of the first competent reader. So every entry says what it proves and, where it matters, what it doesn't.
I — chapterThe body
Bramble, D. & Lieberman, D. — "Endurance running and the evolution of Homo", Nature, 2004
The paper that holds up the entire chapter, and the source of every anatomical trait cited in the opening. It lists over twenty features that have no function in walking and do have one in running long.
- Nuchal ligament. Stabilises the head against the pitching of running. Present in dogs and horses, absent in other living primates. In the fossil record it appears with Homo erectus and is missing in Australopithecus: it's the point where the tool changes trade.
- Achilles tendon and plantar arch. Store and return elastic energy at every footfall; the contribution is negligible in walking and becomes substantial above a certain speed. The chimpanzee's Achilles is short.
- Gluteus maximus. Far larger than in any other primate. Electromyographic activity is modest in walking and high in running: it controls forward trunk flexion at each impact.
- Sweat glands. Standard estimates run from two to four million, with wide individual variation, and not all of them are functionally active. The manifesto deliberately uses the lower figure. What matters isn't the exact number but the comparison: no other primate cools by evaporation across the whole body surface this way.
What it doesn't prove: that humans evolved in order to hunt by running. It proves we evolved to cover distance in heat. What that distance was for is a reasonable inference, not an excavated fact.
Bramble, D. & Carrier, D. — "Running and breathing in mammals", Science, 1983
The respiratory constraint in quadrupeds. In galloping mammals the breathing cycle locks onto the stride at 1:1: the viscera swing against the diaphragm like a piston and set the rhythm. A running human has variable coupling — 2:1, 1:1, 4:1 — or none. It's the mechanical reason we can regulate cooling while moving, and they can't.
Liebenberg, L. — observations of persistence hunting among the San of the Kalahari
The most cited and the most fragile source in the genre. Liebenberg genuinely ran those chases and described them, but they are few, in a single group, and some of the best-known episodes — including a famous BBC documentary sequence — are reconstructions arranged with the hunters, not footage of a spontaneous hunt.
The manifesto uses it as an image and says so in the text. If someone tells you persistence hunting is a myth, they aren't entirely wrong — and the argument doesn't depend on it.
Lieberman, D. — Exercised, 2020
The observation that carries the whole manifesto comes from here: in hunter-gatherer societies nobody trains, and the idea of producing effort that serves no purpose would be incomprehensible. It isn't an anti-modern pose — it's an ethnographic observation, and Lieberman is a fierce defender of physical exercise. The book is also the best antidote to fanatical versions of all this, including ours.
II — chapterThe food
Marlowe, F. et al. — Journal of Human Evolution, 2014
Hadza food preference rankings, collected over years. Honey comes first, and not narrowly. It's the single data point that dismantles the commercial "paleo diet" fastest.
Isack, H. & Reyer, H. — Science, 1989
The greater honeyguide leads people to hives, and the study documents that the human call significantly increases the chance of being guided. Cooperation between two wild species, with no training involved. It also does narrative work in the manifesto: the best food we had required hours of walking behind a bird.
Perry, G. et al., 2007
Salivary amylase gene copy number: populations with historically starch-rich diets carry more, and we carry more than other primates regardless. Starch isn't a recent accident.
What it doesn't prove: anything about low-carb diets as an individual strategy. It only shows that "carbohydrates are foreign to our species" is false.
Hall, K. et al. — Cell Metabolism, 2019
Controlled inpatient study: identical macronutrients on paper, ultra-processed versus minimally processed food, eating ad libitum. The ultra-processed group took in roughly five hundred extra calories a day and gained weight. Small sample, short duration — but one of the few true experiments in a field full of observational studies.
Pontzer, H. et al., 2012 and 2015
Hadza total energy expenditure measured with doubly labelled water: comparable to sedentary Western adults once body mass is accounted for. Hence the constrained expenditure model — the body compensates for extra work by cutting back elsewhere.
This is the most misreadable data point in the manifesto, and the reason there's a callout box in the text. It doesn't say movement is useless: it says exercise isn't the main lever on body weight. The model is debated and not universally accepted.
Goren-Inbar, N. et al. — PNAS, 2002
Gesher Benot Ya'aqov, Lake Hula, roughly 780,000 years ago: stone anvils and hammers used to crack nuts, seven different species at the same site. Nuts are not a 2010 superfood.