In 1872, Charles Darwin published a strange little book called The Expression of the Emotions in Man and Animals, in which he argued that emotions are evolved — universal across the human species, present in other animals, readable on any face anywhere. For roughly eighty years, nobody argued. Then the cultural relativists took over. By the 1950s, the consensus among anthropologists — led by Margaret Mead, who had spent decades documenting how differently cultures communicate — was that emotions are learned. You cry in grief because your culture taught you to. A Balinese person laughs in grief because their culture taught them to. There is no shared emotional language. Every face is a local dialect. And then a young psychologist named Paul Ekman showed up with a suitcase full of photographs and asked: what if you test that?
🧠 The Dictionary That Started It All
Ekman started by showing black-and-white photos of Western actors — people simulating anger, disgust, fear, happiness, sadness, surprise — to participants across the US, Japan, Brazil, Chile, and Argentina. He asked them to match the expression to an emotion label. The agreement was absurdly high. Americans and Japanese identify the same face as “angry”? Brazilians and Chileans too? The data was so clean it almost looked fake. But Ekman knew the skeptics’ counter-argument: these are all literate cultures. Maybe you learn emotional expressions through movies, magazines, shared media. You need a group that has never seen a photograph. So in the late 1960s, Ekman and his collaborator Wallace Friesen packed their bags and flew to the highlands of Papua New Guinea. They found the Fore people — a preliterate tribe living in total isolation, with no exposure to film, television, or any Western media at all. No magazine covers. No Hollywood close-ups. No clue what a “photograph of a face” even was. Ekman showed them the same pictures. And the Fore tribespeople — who had never seen an outsider’s face in their lives — identified anger, happiness, sadness, fear, surprise, and disgust with the same patterns as a college student in Chicago. The cultural relativist argument collapsed on a dirt floor in New Guinea.
🤔 Wait, So Emotions Are Hardwired?
Yes — at least the basic ones. Ekman’s original six (anger, disgust, fear, happiness, sadness, surprise) appear to be universal across every human population ever tested. That does not mean culture does not matter. It means the raw signal is shared. What culture does is install something Ekman called display rules — social norms about who can show which emotion to whom, when, and for how long. A Japanese businessman and an American salesman feel the same frustration at a negotiation. They both have the same face underneath. But one has been trained since childhood to suppress it in front of authority, while the other was raised to “put it on the table.” The emotion is universal. The display is local. This is why you have had the experience of meeting someone from a completely different culture and still being able to read their face — and also having the experience of being completely wrong about what their face means. You read the signal correctly. You misread the display rules.
🔗 Why This Is the Foundation of Every AI That Needs to “Understand” You
If you build an AI companion, Ekman’s work is not academic history — it is the starting line. The entire field of affective computing sits on the premise that emotions have detectable signatures, whether on a face, in a voice, or in a text. A digital confidant that reads your words and infers your emotional state is doing exactly what Ekman’s participants did with photographs: matching a signal to a label. The difference is that your signal is text, not a face. But the core problem is the same — and so is the core risk. Display rules apply here too. A user who types “I’m fine” might actually be fine, or they might be following a lifelong display rule that says you do not burden others with your feelings. The best AI companions are not the ones that read emotions perfectly. They are the ones that know when the display rule is on. And they are the ones that, over time, earn the right to see past it.
🎲 The Duchenne Smile Challenge
Ekman is famous for popularizing the Duchenne smile — the genuine expression of happiness that engages the muscles around the eyes (orbicularis oculi), not just the mouth. A fake smile pulls the corners of your lips up. A real smile crinkles the corners of your eyes. Ekman found that most people cannot voluntarily produce a Duchenne smile. Try it right now in front of a mirror — smile broadly, then try to make your eyes crinkle without genuinely feeling happy. Odds are, it looks like you are lying. Which you are. But here is the real trick: if you hold a Duchenne smile for long enough, your brain starts to generate the corresponding emotion. Ekman demonstrated this — voluntarily making one of the universal expressions actually triggers the physiology and subjective experience of that emotion. Smile hard enough and you get happier. Scowl hard enough and you genuinely feel more irritated. Your face is not a display panel. It is part of the operating system.