The Pointer on Your Face

Try a small experiment. Find a mirror, and hold your head perfectly still. Now, without moving anything else, slide your eyes to the left. Then to the right. Watch what happens: a dark disc travels across a field of white, and the direction of its travel is unmistakable. You could sit across a room from someone doing this and read every move.

That is the whole signal. A dark iris against a bright background, tracked by everyone around you, constantly, without anyone giving it a moment’s conscious thought. We tend to assume we follow another person’s attention from the tilt of their head, the turn of their shoulders, the orientation of their whole body. We do read those things. But the finest, fastest, most precise information about where a person is looking comes from something much smaller and much stranger: the exposed white of the human eye.

No other animal on Earth advertises its gaze quite the way we do. The white sclera, framing the colored iris, is a message you cannot switch off. Scientists have a name for the trait and the idea behind it. They call it the cooperative eye. It is one of those features so ordinary that it hides in plain sight, and yet it may be among the more quietly consequential adaptations in the human story.

What the chimpanzee eye conceals

Look closely at a photograph of a chimpanzee’s eye and something is missing. The whites are barely there. Where a human eye shows a bright, clearly bounded sclera on either side of the iris, the chimpanzee’s is dark, pigmented, blended into the surrounding tissue. The result is an eye that gives away very little. You can tell, roughly, which way the animal has turned its face. You cannot easily tell, from across a distance, exactly where the eyes themselves are pointing.

This turns out to be the rule rather than the exception. Most primates have dark sclera. The white, when it exists at all, is muted and camouflaged. In evolutionary terms this is not an accident but a feature. An animal that can conceal the target of its gaze holds an advantage. A predator does not want to telegraph the exact prey it has fixed on. A subordinate in a tense social hierarchy does not want a dominant to know precisely what it is watching. In much of the animal world, the safest eye is an inscrutable one.

The human eye does the opposite. It is built for exposure. The sclera is bright and unpigmented, the border between white and iris is sharp, and the horizontal proportions of the human eye are unusually wide, giving the iris a long, high-contrast runway to travel along. Every one of these features makes the direction of a person’s gaze easier to read. We did not evolve to hide where we are looking. We evolved to show it.

In 2001, two researchers at the Tokyo Institute of Technology set out to measure exactly how unusual this was. Hiromi Kobayashi and Shiro Kohshima photographed and measured the eyes of eighty-eight primate species, comparing the size of the exposed sclera, its coloration, and the shape of the eye outline ¹. Across the full sweep of primate diversity, humans stood out. The exposed white was proportionally the widest, the outline the most horizontally elongated, the contrast between sclera and iris the starkest. We were, in the plainest visual terms, the odd primate out.

From this comparison Kobayashi and Kohshima drew a bold proposal. The white sclera, they argued, evolved precisely because it makes gaze direction legible to others. A dark-eyed animal keeps its attention private. A white-eyed animal broadcasts it. And a species that broadcasts its attention is a species that has come to depend on others reading it. This became known as the cooperative eye hypothesis: the idea that the visible human eye is an adaptation for shared attention, for the silent coordination of two minds looking at the same thing.

The head or the eyes

The hypothesis was elegant, but it left a puzzle unresolved. When we follow another person’s gaze, what are we actually reading? Is it the eyes themselves, the dark disc against the white? Or is it the larger, cruder cue of the head, the way a face turns toward its object of interest? In ordinary life the two move together. Someone who looks at the window usually turns their head toward it as well. To find out which cue does the real work, you have to pull the two apart and see which one people follow when they are forced to choose.

That is precisely what Michael Tomasello and his colleagues at the Max Planck Institute for Evolutionary Anthropology in Leipzig set out to do. Tomasello, a developmental psychologist who has spent his career comparing the minds of human children and great apes, wanted to know whether our unusual eyes had rewired the way we read one another. In 2007 his team designed a test of disarming simplicity ².

An experimenter would look up toward the ceiling in one of four ways. In the first condition, the experimenter moved only the head, keeping the eyes closed or fixed. In the second, the head stayed still and only the eyes moved upward. In the third, head and eyes moved together, the natural combination. In the fourth, nothing moved at all, a control. The researchers then watched whether their subjects would follow the experimenter’s attention upward. They ran the study with great apes, including chimpanzees, bonobos, and gorillas, and separately with human infants.

The results split cleanly along species lines. The great apes tracked the head. When the experimenter turned his head toward the ceiling, the apes looked up. But when the head stayed still and only the eyes moved, the apes largely missed it. The eyes alone did not carry enough information for them. Human infants did the reverse. They followed the eyes, looking up even when the head remained motionless and only the gaze had shifted. Where the apes relied on the gross movement of the head, the infants keyed in on the fine movement of the eyes.

The pattern was striking enough to state simply: great apes rely on the head, human infants rely on the eyes. It fit the cooperative eye hypothesis almost too neatly. Our species had evolved the visible white sclera, and along with it a perceptual sensitivity that leans on exactly the information that white provides. The eyes had not merely become more legible. We had become tuned to read them.

Still, a fit is not a proof. That human infants attend to eyes while apes attend to heads is consistent with the idea that the white sclera matters, but it does not demonstrate that the white sclera itself is doing the work. Correlation is not cause. To show that the bright field behind the iris genuinely carries the signal, you would need to do something more direct. You would need to take the white away and watch what breaks.

Turning the eye inside out

Fifteen years after Tomasello’s study, a team led by Fumihiro Kano, working at the Kumamoto Sanctuary of Kyoto University, found a clever way to do exactly that. In a 2022 paper published in the journal eLife, Kano and colleagues manipulated not the eyes of a living face but the photographs of them ³. They took images of eyes and reversed the contrast, the way a photographic negative inverts light and dark. The bright white sclera became dark. The dark iris became pale. Every other feature of the eye stayed the same. Only the contrast polarity flipped.

Then they asked people to judge where each eye was looking. With normal contrast, the arrangement every human sees from birth, people read gaze quickly and accurately. They could pinpoint the direction of a glance with ease. With reversed contrast, their performance dropped. Judgments became slower and less reliable. The same eyes, carrying the same geometric information about direction, suddenly became harder to read once the light and dark were swapped.

The effect was not uniform. It grew sharpest under exactly the conditions where the white sclera should matter most: in dim light, and at a distance. Up close and in bright illumination, people could compensate, drawing on the shape and position of the iris regardless of the surrounding tone. But push the viewing conditions toward the edges of what the eye can resolve, and the bright-against-dark contrast turned out to be doing real perceptual work. It was the high contrast between white sclera and dark iris that let people read a gaze across a room or in low light. Strip that contrast, and the signal degraded.

Here the study did something the earlier work could not. Kano and colleagues ran the same reversed-contrast task with chimpanzees, testing whether they, too, depended on the ordinary polarity of light and dark. The chimpanzees showed the same drop in performance. When the contrast was normal, they read the direction of a gaze; when it was reversed, they struggled, and again the deficit was worst in the harder viewing conditions. Both species, human and chimpanzee, leaned on the same dark-against-light contrast to extract gaze from an eye.

That last finding matters, because it complicates the tidy story. It shows that chimpanzees are not blind to the mechanics of gaze reading. Given the right contrast, they use it much as we do. The difference between us and them was never that they cannot read eyes. The difference lies in what our eyes give them, and us, to work with.

Where the famous story goes wrong

A popular version of this science has hardened into something close to a slogan: no other primate can follow a human gaze; only we possess the trick of reading eyes; the cooperative eye is what makes us unique. It is a satisfying story, and it is wrong in an important way.

Great apes follow gaze perfectly well. Decades of research have shown that chimpanzees, bonobos, and other apes track where others are looking, will follow a companion’s line of sight around a barrier, and use gaze as a source of information about the world ². What Tomasello’s experiment revealed was not that apes are gaze-blind but that they weight the cues differently. They lead with the head. When the head turns, they follow. The eyes, in an ape, carry less of the signal, in part because an ape’s dark sclera offers less of a signal to carry.

So the real distinction is subtler and, in a way, more interesting than the slogan. It is not that apes cannot read gaze and humans can. It is that we have built a visible pointer and wear it on our faces, a pointer so clear that even another species can read it when we hand it to them at the right contrast. The uniqueness is not in the perceiving mind. It is in the broadcasting eye.

This reframing dissolves a false hierarchy. We are not the only primates capable of joint attention, nor the only ones sensitive to where a companion looks. We are the primates who chose exposure over concealment, who traded the tactical advantage of a hidden gaze for the cooperative advantage of a legible one. Somewhere in our evolutionary past, the benefit of being read by our own kind came to outweigh the cost of being read by our rivals and prey.

The advantage of being read

Why would a species give up the ability to hide where it is looking? The answer the cooperative eye hypothesis offers is that we became a species for whom being understood was worth more than being inscrutable. A visible gaze is a low-cost, high-bandwidth channel of communication. A glance can direct attention to a threat, to a tool, to a piece of food, to a third person, without a single word or gesture. In a species that hunts together, raises children together, and coordinates in groups, an eye that silently says look there is a genuine asset.

The dependence runs deep and begins early. Newborn infants, only days old, will lock onto a human face and, in particular, onto the eyes. They prefer faces that look directly at them over faces that look away, and they orient toward the eyes before they can hold up their own heads or reach for an object. Long before an infant can speak, walk, or grasp, it is already reading the whites of the eyes turned toward it. The pointer is legible from the very start of life, and the sensitivity to it seems to come as standard equipment.

Consider how much of ordinary human life runs through this narrow channel. A teacher’s eyes flick to the student who should answer next. Lovers hold a gaze a beat too long across a crowded room, and something passes between them that no words carry. A parent’s glance can stop a child mid-reach. A speaker checks whether a listener is still with them by watching where their eyes have gone. Deception, flirtation, warning, invitation, correction, all of it can travel on a dark disc moving across a white field. We are so fluent in this language that we forget we are speaking it.

The white of your eye, then, is not a neutral piece of anatomy. It is a signal, evolved and maintained because others use it. You cannot turn it off. Whenever your eyes are open, you are transmitting, and everyone around you is quietly receiving. The next time someone catches your eye across a room, it is worth remembering what actually happened. You did not read the tilt of their head or the set of their shoulders. You read the white of their eye, the one message the human face has never learned to keep to itself.

Sources

1. Kobayashi, H. & Kohshima, S., Unique morphology of the human eye and its adaptive meaning, Journal of Human Evolution, 2001. — https://www.sciencedirect.com/science/article/abs/pii/S0047248401904686
2. Tomasello, M., Hare, B., Lehmann, H. & Call, J., Reliance on head versus eyes in the gaze following of great apes and human infants: the cooperative eye hypothesis, Journal of Human Evolution, 2007. — https://www.sciencedirect.com/science/article/abs/pii/S0047248406001680
3. Kano, F., Kawaguchi, Y. & Yeow, H., Experimental evidence that uniformly white sclera enhances the visibility of eye-gaze direction in humans and chimpanzees, eLife, 2022. — https://elifesciences.org/articles/74086
4. Farroni, T., Csibra, G., Simion, F. & Johnson, M. H., Eye contact detection in humans from birth, Proceedings of the National Academy of Sciences, 2002. — https://www.pnas.org/doi/10.1073/pnas.152159999
5. Emery, N. J., The eyes have it: the neuroethology, function and evolution of social gaze, Neuroscience & Biobehavioral Reviews, 2000. — https://www.sciencedirect.com/science/article/abs/pii/S0149763400000257
6. Tomasello, M., Origins of Human Communication, MIT Press, 2008. — https://mitpress.mit.edu/9780262515207/origins-of-human-communication/

Related reading

• The Door at the Top of Your Face
• The Map Printed Inside Your Hands

More from the Body edition →