Both, and the distinction is the answer: the disposition to care about fairness is natural—an evolved trait of social animals—while the specific rules that count as fair are cultural inventions built on top of that disposition.
The natural layer is well evidenced.
Pre-verbal infants prefer characters who help over those who hinder and react to unequal distributions before they could have been taught to.
Non-human primates protest inequity: capuchin monkeys refuse a reward they had happily accepted once they see a partner get a better one for the same task.
Across every human society people share, reciprocate, and punish cheaters, often at cost to themselves—third-party punishment appears universal.
These are not plausibly learned from scratch each time; they look like adaptations for the cooperation that made humans successful.
The invented layer is equally clear.
What people call fair diverges sharply: equal shares, shares by need, shares by contribution, or shares by rank are each treated as obviously just in different cultures and contexts.
The instinct says "distributions matter and cheating is wrong"; it does not specify the distribution, and cultures fill that in differently, sometimes incompatibly.
So fairness is neither pure biology nor pure convention.
The most defensible picture is a universal grammar with local vocabularies: an inherited sense that reciprocity and desert matter, elaborated into specific, contestable rules by history and circumstance.
Practically, you can expect some fairness norm in any human group and appeal to the shared instinct, but you cannot assume your particular version of it is the one others hold—negotiation is about the content, not the existence, of fairness. This would need revision only if the cross-species and infant findings failed to replicate; so far they largely have.
Both—and splitting them is the whole answer. The urge toward fairness is natural, wired deep into social animals. The specific rules for what's fair are invented and differ wildly across cultures.
The evidence for the natural part is striking:
That's not learned from scratch each time; it looks like an adaptation for cooperation, which is what made our species work.
But what counts as fair is clearly invented. Equal shares? Shares by need? By effort? By rank? Different cultures treat each as the obvious answer. The instinct says "how we divide things matters and cheating is wrong"—it doesn't say how to divide, and we fill that blank in differently, sometimes in ways that clash.
The practical takeaway: in any group you can count on people having a fairness sense and appeal to it—but don't assume they share your version. Most fights aren't about whether fairness matters; they're about whose definition wins. Argue the content, not the existence.
The Monkey Who Threw the Cucumber.
In a now-famous experiment, two capuchin monkeys do the same small task; one is paid in cucumber, the other, in full view, gets a grape. The first monkey, perfectly content with cucumber a moment ago, tastes injustice—and hurls the vegetable back at the researcher, rattling the cage in outrage. Watch the clip and you are watching something that predates language, law, and philosophy: a sense of that's not fair running in a creature that has never read a word. Whatever fairness is, it did not begin with us. It is older than our species, wired into the nervous systems of animals who must cooperate to live.
The Instinct with No Content.
And yet the monkey's outrage cannot tell you what a fair wage is. This is the strange doubleness of fairness: the feeling is universal and the rule is empty. We are born knowing that how things are divided matters and that cheaters must be caught—but not how to divide. Equal shares to all? More to those who worked harder? More to those who need most? To the elder, the firstborn, the king? Each has been "obvious justice" somewhere. Fairness is a grammar without a vocabulary, a hunger that does not name its meal, an instinct that insists on an answer it cannot itself supply. Culture rushes in to fill the blank, and fills it a thousand different ways.
The Trick of Not Knowing Who You'll Be.
John Rawls proposed a machine for inventing fairness from scratch: design the rules of a society without knowing which person in it you will be—rich or poor, strong or frail, first or last. Behind this "veil of ignorance," self-interest quietly becomes justice, because a rule you'd accept without knowing your place is a rule that is fair to every place. It is a profound move, and it reveals fairness as an act of imagination—the deliberate erasure of your own position to see from everywhere at once. We are, it seems, natural enough to want fairness and clever enough to have to build the machinery that finds it.
The Immune System of Cooperation.
Why would evolution bother to grow a sense of fairness at all? Because cooperation is a fragile miracle constantly threatened from within: every group that shares is vulnerable to the one who takes without giving, and a few unchecked cheaters can collapse the whole arrangement. Fairness is the immune system that guards it—the cheater-detector, the flash of indignation, the willingness to punish a free-rider even at cost to oneself. Our sense of justice is not decorative sentiment; it is the enforcement mechanism without which no reciprocity could survive its own exploiters. We feel fairness so keenly because the ancestors who didn't were out-competed by the groups who did.
The First Philosophy Is a Child's Cry.
Every human being's first encounter with ethics is the same three words, shouted before they could define a single one of them: that's not fair. No one teaches a child to feel the sting of the smaller slice; they arrive already indignant, and only later learn the culture's particular theory of why. In that cry both truths meet—the natural and the invented, the ancient reflex and the argument still to come. Fairness is something we are born already reaching for and something we spend our whole civilization trying to specify, find and forge at once. We did not invent the desire. We are still, all of us together, inventing the answer.
Fairness is both discovered and made. The disposition to care about it is natural—an evolved feature of cooperative animals, visible in human infants and in other primates who protest unequal treatment, and expressed everywhere as sharing, reciprocity, and the punishment of cheaters. But the specific content of what counts as fair is a cultural invention that varies widely—by equality, need, effort, or rank—so the instinct sets the question without settling the answer. In practice you can count on any human group to have some sense of fairness while not sharing yours, which is why disputes are almost always about the rule rather than the reality.