On a January morning in 2007, Wesley Autrey saw a stranger collapse on a New York City subway platform, then tumble onto the tracks. Without hesitation, Autrey jumped down, pressed the man into the drainage trench between the rails, and lay on top of him as a train passed over them both. They survived with millimetres to spare.
Autrey didn't know the man. He had no training. He had his two young daughters standing on the platform watching. When asked why he did it, he said: "I just saw someone who needed help."
The Evolutionary Puzzle
From a strict evolutionary perspective, altruism toward strangers shouldn't exist. Natural selection favours behaviours that help your genes survive. Helping family makes evolutionary sense — they share your DNA. Helping your community makes social sense — reciprocity and reputation matter.
But jumping in front of a train for a stranger? Donating a kidney to someone you found on the internet? These acts seem to defy the logic of self-interest entirely.
Kin Selection and Reciprocal Altruism Fall Short
Biologist W.D. Hamilton's theory of kin selection explains why we help relatives. Robert Trivers' theory of reciprocal altruism explains why we help people who might help us back. But neither theory adequately explains why humans regularly help strangers they'll never see again, at significant personal cost.
Something else is going on.
The Empathy-Altruism Hypothesis
Psychologist C. Daniel Batson spent decades testing what he called the empathy-altruism hypothesis: when we feel empathic concern for another person — when we genuinely share their suffering — we're motivated to help them for their sake, not ours.
His experiments systematically ruled out alternative explanations: it's not about reducing our own distress, it's not about looking good, and it's not about expected reciprocity. Pure empathic concern produces genuinely altruistic motivation.
The Helper's High
Neuroscience adds another dimension. Helping others activates the brain's mesolimbic reward system — the same circuitry involved in food, social bonding, and other pleasurable experiences. Giving produces a measurable "helper's high."
But here's the crucial detail: the reward follows the impulse, it doesn't precede it. We don't help in order to feel good. We feel good because we helped. The reward system reinforces altruistic behaviour after the fact.
What This Means for Sonder
The capacity for altruism toward strangers may be evolution's most remarkable gift. It suggests that beneath the surface differences — language, culture, appearance, belief — we recognise something shared in every human face.
That recognition is sonder. And when we act on it, something extraordinary happens: two strangers, for a moment, become something more.