My three-year-old hit another kid at the playground last week. Standard toddler stuff. I knelt down and said the thing parents say: “We need to be nice to people.”
He looked at me and asked why.
I opened my mouth and realised I didn’t have an honest answer that wasn’t also, at some level, selfish.
Because people will be nice back to you? That’s transactional. It’s also not reliably true — I’ve spent three years building something from nothing at a company, and the person who benefited most from that work is the one who pushed me out. Be nice because it pays off is a hypothesis that fails on contact with enough corporate reality.
Because it’s the right thing to do? Sure, but that’s circular. He’ll ask why it’s right, and I’ll be back where I started.
The evolutionary biologists have an elegant framework for this. Dawkins’s selfish gene, Trivers’s reciprocal altruism, Axelrod’s tournaments — cooperation isn’t altruism, it’s strategy. Tit-for-tat wins the iterated prisoner’s dilemma. Be nice first, retaliate if betrayed, forgive quickly. The math works. Cooperators outcompete defectors over enough rounds.
But here’s what the models miss: the game isn’t always iterated. Sometimes you leave and never see the other player again. Sometimes the defector gets promoted and you get managed out. Sometimes the rounds just stop, and the nice player is left holding clean documentation that nobody will read.
So why do it?
I think the real answer is something the game theorists accidentally stumbled over without noticing. In Axelrod’s tournaments, the winning strategy wasn’t just effective — it was legible. Other players could understand what tit-for-tat was doing. It was predictable, not in a weakness sense, but in a character sense. It had a recognisable shape.
That’s closer to the truth. You’re nice not because of what it gets you from any particular person, but because of what it makes you into. The kind of person who leaves clean handover notes even when pushed out. The kind of person who documents the investigation trail even though the successor might never look at it. Not because it’s strategic — though it sometimes is — but because the alternative is becoming someone who doesn’t, and that person is harder to live with. Specifically, harder for you to live with.
There’s a Chinese concept, 做人 (zuòrén) — literally “doing person,” meaning the art of conducting yourself as a human being. It’s not ethics exactly, and it’s not strategy. It’s more like: there is a craft to being a person, and you can practice it well or poorly, and the quality of your practice is visible mostly to yourself.
My son won’t understand any of this for another decade. For now I told him: “We’re nice because that’s who we are.” Which felt incomplete, but might actually be the whole answer. The kind of person you become stays with you longer than any one interaction. The playground bully forgets by tomorrow. But you’re still you at bedtime.
The dark interpretation — the selfish gene, the corporate politics, the collaboration-as-mutual-exploitation reading — isn’t wrong. It’s just incomplete. Yes, humans cooperate because it benefits them. Yes, organisations can be captured by defectors. Yes, sometimes nice guys finish last in any given round. But “any given round” is the key qualifier. You’re not playing one game. You’re building a person, and that person has to face the mirror for a lot of rounds.
I still can’t fully explain to a three-year-old why we’re nice to people. But I think the inability to give a clean, simple answer might itself be the point. The question is supposed to stay open. The day you have a crisp, confident answer — “because it’s strategic” or “because God says so” or “because society requires it” — is the day you’ve probably stopped actually thinking about it.
The best I can do: we’re nice because the world we want to live in is made of the choices we make, one playground interaction at a time. And because the alternative — becoming someone who calculates every kindness — is its own quiet punishment.
P.S. He hit the kid again ten minutes later. We’re working on it.