Several months ago, my partner and I bought an apartment in South London. Our previous home was a rental in which, for reasons best known to the landlord, there were mirrors everywhere. The bathroom had two; there was one outside on the terrace; in the bedroom, mirrored panels stretched across a twenty-foot-long wall. On moving day, we realized that we had a problem: the new apartment was mirror-free, and because we’d been so spoiled we weren’t bringing one of our own. We spent a few days filling our drafty rooms, decanting books, building furniture, and dressing every morning without seeing ourselves in profile. It was a couple of weeks before we bought a simple mirror, wooden and round, to hang above the bathroom sink. By then, I joked, we didn’t recognize ourselves.
As my feeble quip indicates, we take for granted that we possess a steady mental image of our own faces, and that we can reconfirm these images frequently. If I asked you to picture yourself, you could almost certainly do this with ease. You might refer to the memory of how you looked in the mirror this morning, or in the selfies you took last night, or in that passport photo you hate. And if I suggested that you double-check your mental image, you could probably do so by looking in the mirror again, or at your reflection in a nearby window, or, via a camera app, using the very screen on which you may be reading this paragraph. Most of the time, we think we know how we look.
Discover notable new fiction, nonfiction, and poetry.
Yet that’s possible only because we live in the aftermath of a revolution. In the past few centuries, the mental self-portrait has become commonplace. To be sure, the world has held pockets of reflection as long as humans have existed to see them. As a piece of technology, the mirror seems to date back several millennia. The ancient Egyptians had polished bronze, copper, and obsidian at their disposal, as did the Greeks and the Romans. But such objects were hardly ordinary, and people had to improvise: in Ovid’s Metamorphoses, Narcissus, the ultimate lover of reflections, sees himself for the first time in a crystalline pool. Pausanias, writing a century later, called this scene “totally absurd,” because any young man would have seen a reflection before. Still, for thousands of years, there was an obvious gap between image and life. Now in a kitchen pot, darkly; now face to face.
The revolution came in the sixteenth century, when Venice’s glassmakers applied a tin-and-mercury alloy to their panes, producing a far clearer reflection than had previously been achieved. These artisans, with their trade secrets, were kept on Murano, a cluster of islands just across the lagoon from the city, ostensibly as a precaution against fire, though the state would also send assassins after anyone who tried to leave. Still, a few hardy glassmakers escaped to the Saint-Gobain glass workshop in Paris, which they lifted to pan-European dominance. The workshop gave us, among other monuments to subtlety, the great Galerie des Glaces in the Palace of Versailles, which contains three hundred and sixty-seven mirrors, stretching eighty yards. The technology could be replicated, spread, sold. Mirrors quickly infiltrated the private homes of the Continent, then its colonies; they became cheap and unremarkable, as they remain today, if you stop shelving your books long enough to buy one.
In the centuries that followed, the private face became increasingly public. More people could look at their own countenances more accurately, and more often. Candlelight and gaslight gave way to electricity, an unwobbling source of interior light. Photography arrived, kindling an enthusiasm that Baudelaire, ever the cynic, compared to a lust for pornography. Cosmetics, as venerable as mirrors, boomed to an unprecedented degree. Think of Esther marinating her skin in myrrh, then of Eugène Rimmel, who introduced the first mass-produced mascara via an early mail-order catalogue. Pierre-François-Pascal Guerlain, a trained chemist, sold skin creams with such aplomb that he was hired by Napoleon III’s empress, Eugénie. Fast-forward to the twenty-first century: to fix your flesh, there’s Botox, fillers, and plastic surgery; to augment the image, there are camera-app filters and Photoshop. In 2024, according to Apple, the world’s iPhone users took five hundred billion selfies. Facial-recognition technology is used across the globe. In the span of a few centuries, the face has become inescapable. We seem to be condemned to look our best.
Much of our social life, our ways of making and sharing a world, orbits the face. Emmanuel Levinas, the supreme theorist of facial affairs, described our fleshy ovals as the site where ethics—his “first philosophy”—begins. This rings true when it comes to everyday behavior. The phenomenon may be hardwired: babies prefer faces to other complex objects. And, for the rest of our lives, faces seem charged by something that arms or legs, however attractive, don’t have. When you talk to me, I look at your eyes, regardless of whether you’re waving your hands. If the gaze is held too long, one of us has to look away. If I can’t see your face—say, you turn aside, or you’ve called me on the phone—I can get by, can make the conversation work, but I might feel that our connection is lacking. The face’s centrality to social life is inscribed in our language, in the metaphors that betray our world view. In English, we face up to things, or face them down; we show our faces where required; we try to save face, and should that fail we make an about-face to save ourselves.
Yet there’s another side to this coin. Our vocabulary also speaks of invention, even artifice. The English word “face” derives from the Latin facies, implying a created form; so does the French visage, from videre, suggesting something seen from without. “Mask,” “masque,” “mascara,” “maquillage,” and their European relations seem to be etymologically linked, and carry long-standing associations with concealment, distortion, pretense. For Socrates, the art worth prizing was the cultivation and preservation of natural beauty; we’ve happily overwritten it with what he treated suspiciously as kommōtikē, the art of changing how one looks. Even “person,” along with “impersonation” and “personae,” derives from persona, a theatrical mask through which classical actors spoke. The cultural historian Hans Belting suggested that faces and masks were conceptually inseparable: we shouldn’t think of one as “real” and the other as “fake”—one as the thing we have and the other as the thing we temporarily don. Life, he wrote, was fully “a perpetuum mobile,” an “expressive drama,” in which our faces resolve into one legible position, one legible role, then reassemble themselves into the next. We make them up, in every sense.
The face, then, lives by its pliability. We might encounter a completed sentence or picture, but never a completed face. It can always shift into something else, pushed by personal whim or social pressure, and keep doing so until frozen into one last mask by sickness or death. This means that we live together in an interpretative flux. We constantly read other faces, and understand that ours, at the same time, will be interpreted. As T. S. Eliot put it, we prepare a face to meet the faces that we meet, and we do so knowing that they’ve been prepared as well, and that our interpretations will almost immediately become obsolete.
Levinas tends to be the philosophical lodestar here, with his near-mystical love of the face-to-face encounter. He wrote that the face was “a living presence,” that “the Other faces me and puts me in question and obliges me.” Obliged him to do what? This may not be clear. Encountering a face means reëncountering the world: it is a silent epiphany about the fact that other people are radically outside of you, maybe unknowable in a final sense, and yet you must engage with them anyway.
At the same time, I prefer Ludwig Wittgenstein. He was more impish than people remember, and he liked to draw little faces when he was teaching or writing, sketching each one in a few strokes and usually making it smile. Such doodles recur in his manuscripts from the early nineteen-thirties until shortly before his death, in 1951; he used them to illustrate a battery of subtle points, one of which was how misleading our language could be. Here’s one he drew for his students:
After which he said:
One of the things Wittgenstein implies is that we don’t theorize the faces we see. I might talk of “reading” your face, but it isn’t like reading a code or a map: I don’t look at you, add together your liquid eyes and droopy mouth, and pronounce my final judgment that you’re sad. I simply look at your face, apprehend its gestalt as a unity, and go from there. Recent neuroscientific research has shown that we process whole faces more readily than mere parts; we aren’t adding the bits up. We look, and we speak, and we look again: that’s life.
One oddity, given all this, is that, although humans have made pictures of faces for as long as we’ve represented anything, realism has only recently become the norm. As the scholar Fay Bound-Alberti points out in “The Face: A Cultural History” (Grand Central), between the thirty-thousand-year-old figures on the walls of the Chauvet-Pont d’Arc cave and ancient Cycladic statuary, from the fourth to second millennium B.C., human figures rarely showed recognizable facial features. There were exceptions. In Egypt, certain pharaohs seemed, for reasons still debated, to tilt their own representations from the ideal to the natural. Senusret III, the most outlandish example, had temples filled with statues of himself, his gigantic faces lined and grim and aged, though set atop a young and muscular body. Centuries later came an interlude in Greece and Rome, beginning with the so-called Severe style of Greek sculpture, in which faces gained clearer features and bodies sharper definition. Socrates’ ugliness can be reliably inferred from his many unsexy sculptural representations; the statues and busts of individual Caesars, though they blend truth and propaganda, remain easy to distinguish. But this facial gallery closed with the Western Roman Empire and the rise of Christianity, which inherited a suspicion of idolatry from the Old Testament and Judaism. David Le Breton points out in “Faces” (Polity) that the Apostolic Constitutions, a fourth-century collection of Christian ecclesiastical laws, recommended that painters be exiled altogether from Christendom, just as poets would have been from Plato’s ideal state. They stayed, but, rather than memorializing individual heroes, they largely devoted themselves to religious imagery, in which realism was again forgone in favor of iconic value. The old order had resumed.
Only with the Renaissance would it break again, with the fifteenth-century blossoming of the naturalistic portrait: van Eyck, Campin, Piero della Francesca. The archetype of painted interiority may still be the good old “Mona Lisa,” known in Italian as “La Gioconda”—the face that launched a thousand conjectures. Bound-Alberti brings her in, dutifully, to say how “enigmatic” her smile is, though Wittgenstein, again, made more of the cliché: “When we speak of the enigmatic smile of the Mona Lisa, that may well mean that we ask ourselves: In what situation, in what story, might one smile like that?” To talk of stories, he means, is to talk of time; you need the before and the after, the rounded sense of a life. After all, some people laugh in anger and weep with joy. If I read your smile as happiness, I’m drawing on a fluid sense of the character I know, and on our place in an emotional culture. One reason Leonardo’s painting has endured in the popular imagination is straightforward: we want to ask questions of her—how she feels, what she’s thinking—and though we don’t believe she’s alive any more than we believe Hamlet or James Bond to be alive, we can’t help responding to her face as if she were.
One unexpected consequence of the return of realism was a renewed attention to anatomy. Take dissection, which was mostly in abeyance through the medieval era, though by the thirteenth century in Europe it was beginning to appear in university curricula. During the Renaissance, it flourished again, and became a routine part of an artist’s studies, as Leonardo’s sketches attest. At the same time, another art returned, equally venerable and equally obsessed with the human form: physiognomy. The thinking, from the Greeks onward, was that you could, if sufficiently skilled, read people’s inward rottenness from their outward appearance. They might alter their behavior or speech, but a nose could never lie. Physiognomic treatises reappeared in the sixteenth and seventeenth centuries, then took off in the eighteenth and nineteenth. The star was Johann Kaspar Lavater, a Swiss poet and philosopher, whose “Physiognomische Fragmente,” a complete manual of flesh-reading, was published in four volumes between 1775 and 1778, featuring illustrations by Henry Fuseli and William Blake of faces and other body parts. It was successful enough to be abridged in multiple forms—“The Portable Lavater,” “Ladies’ Lavater”—and by the end of the century, Le Breton writes, “a visit to Lavater, complete with a physiognomy consultation with the master, was a must for anyone travelling to Switzerland.” The fashion spread. George Sand was a Lavater fan. So was Honoré de Balzac, who called physiognomy “prophetic.” Balzac used Lavater’s work in devising his own characters, who are practically lashed to destiny; Samuel Beckett would deride them as “clockwork cabbages.” One eighteenth-century chief justice of Naples took aside convicted prisoners who still refused to confess and personally examined their heads. If he found proof of inbuilt depravity, he would approve their execution; if he didn’t, he’d set them free.
Such thinking had serious critics. If you judged someone’s soul from his appearance, Montaigne noted, it would be curtains for Socrates. The entry for “physionomie” in Diderot and d’Alembert’s Encyclopédie described it as “imaginary” and “ridiculous.” Hegel compared physiognomists to a merchant who complains that it rains whenever the fair comes to town. But the belief died hard, and photography spurred its adherents on. Once faces had been “captured,” as we still say, they could be put into giant comparative systems, proto-databases of facial types. Francis Galton, the father of eugenics, used superimposed portrait photos in his hunt for “deviant” traits that criminals supposedly shared. His friend Cesare Lombroso, an Italian criminologist, agreed, suggesting that the signs of a “born criminal” included a receding brow and a heavy jaw.
As genetics, psychology, and neuroscience ascended, the twentieth century sent physiognomy back into disrepute, and today, from Lavater to Lombroso, its promoters may seem a racist shade of quaint. Yet the conviction that your face will, if read skillfully, betray your hidden qualities has never gone away. The most influential facial theorist of the late twentieth century was the psychologist Paul Ekman, who argued that seven “universal emotions”—happiness, sadness, anger, surprise, disgust, fear, and contempt—recur across all cultures. Their display, he argued, can be too fleeting to control. Spot these “microexpressions” and map them, perhaps through Ekman’s proprietary Facial Action Coding System, and you’ll have a key to the human soul. Ekman’s ideas travelled unusually well: he wrote popular books, worked on the Pixar film “Inside Out” and the Fox series “Lie to Me,” and provided training to Scotland Yard, the T.S.A., the F.B.I., and the C.I.A. The appeal is plain enough. It promises a portable science of human disclosure. The trouble is that there is no context-free understanding of a social phenomenon; every use of a word like “anger” is imbricated in its past and helps shape its future. Ekman’s seven-point system, however useful as a package, leans toward rigidity, turning the face once again into a surface to be decoded.
The drive to define the face as a data point, something to be grasped and controlled, underpins the bureaucracy of the modern nation-state, in which faces are surveyed, categorized, and stored in digital banks. We show our faces when crossing the border, whether by holding out our passports or by staring into a digital camera. We carry papers that show our faces. Not coincidentally, plastic surgery has been advanced in part by governments: its pioneer, Sir Harold Gillies, fixed the faces of British soldiers who had been disfigured in the First World War, and today the U.S. Department of Defense funds face-transplant research in the United States.
Facial-recognition systems are now operative in moments big and small, whether you’re activating your Apple device or being watched as you march against the regime. Some defend the practice, and others shrug. But the biases of these systems suggest that physiognomy has taken a new, less explicit form. Police facial-recognition systems in the U.S. have had a disproportionately high error rate among Black people. Black Americans are also disproportionately likely to be subjected to such systems; they are overrepresented in the prison population, and in recorded recidivism rates. Their faces, in other words, can be drawn into a technologically enhanced ouroboros. And my own home, the U.K., always seems to follow its cousin’s lead. Between 2024 and 2025, our police forces almost doubled their use of the technology on pedestrians; on the day I write this sentence, there are two huge marches in London, one pro-Palestine and one anti-immigration, and both are being actively surveilled by facial-recognition technology. Since 2013, Britain has had an independent Biometrics Commissioner, and every holder of the post has called for greater oversight of these systems, but privacy advocates have argued in court that there are “no meaningful limits” on what the police can do.
Many of the tech behemoths, too, have invested in facial recognition, and, for a while, they augmented it with emotion-detecting tools. Amazon, for instance, announced in 2019 that its system Rekognition could more accurately identify “all 7 emotions”—namely “Happy,” “Sad,” “Angry,” “Surprised,” “Disgusted,” “Calm,” and “Confused”—and had added a new one, “Fear.” This diverges a little from Ekman’s schema; perhaps the publicity chiefs found “Contemptuous” too unpleasant to sell. Rekognition has a broad spectrum of clients, from ticketing platforms to credit bureaus. It once sought to sell its services to Immigration and Customs Enforcement, though ICE plumped for a rival firm, Clearview AI, which is less interested in emotion than in flat identification. Clearview scrapes facial data from the web on a mass scale—at one point, it absorbed ten billion social-media photos—and has faced dozens of lawsuits and regulatory actions from American states and European countries. The future of emotion recognition, meanwhile, seems uncertain: Microsoft and Google both developed such tools but have since limited or ditched them. Maybe, outside the F.B.I.’s interrogation rooms, knowing how someone feels isn’t proving that valuable.
The trend toward standardization and control has become coldly clear. And as with bureaucrats, so with beauticians: we can capture faces, sort them, approve them, reject them, modify them, and rejuvenate them. What you can measure you can try to perfect. The ancient approach to cosmetics seems like frippery next to the “looksmaxxers” of today, led by Clavicular, a razor-jawed American influencer whose commitment to surgery extends to smashing his own bones, and who has said that beauty isn’t “about sexual attraction, it’s about mathematics.”
Clavicular, though, represents merely the crunchy end point of a logic that has taken hold of affluent life. The same mathematical approach underpins platforms such as Instagram, a wet market for faces, on which users self-brand and sell themselves—and on which the most successful products are those which top the algorithmic charts. Hence the rise, in the twenty-tens, of the so-called Instagram face, on which so many visages slowly converged, altered by fillers and the like to become ethnically nebulous, big-lipped, and often, as videos indicated, hard to move. Last year, an ITV survey suggested that among Brits between the ages of eighteen and twenty-five—a group with near-total smartphone and social-media penetration—one in five had some form of tweakment. Le Breton, with some justification, compares the plastic surgeon to “a psychiatrist advising a paranoid patient to be more wary of his enemies.”
Social media needs the face because it burns attention for fuel, and the face is where I go when I look for you. But I can’t look to these platforms. I find it all so unsatisfying: the resculpted faces, the curated lives, the naked grabbing of my gaze. Something uncertain, something human, has been drained like blood from this world. There’s evidence that Botoxed people become worse at reading other people’s emotions, because when you engage with others you simulate their expressions in your own facial musculature. It’s as if we were commodifying the face, putting it under glass, turning it into a transactional token rather than a pliable living thing.
That little half smile you see could be smug or coy, or even lady-killing; you need to ask which, and why, and in doing so you’re doing moral work, because you’re treating the smiler as an autonomous being whose form of life you need to grasp, however provisionally. But think of the faces we now trade most readily, in the billions every day: emoji, complete and prepackaged, limited in number, coined and issued by a consortium of leading tech companies. I can never shake the sense that they’re fatally cheap. If I saw a friend post about this piece, or send me a message about it, and use an emoji in doing so, I might first envision the real face of that friend, or try to, but then I would feel unmoored. No true face is as standardizable as the image the person has sent in its stead, and another friend, my partner, even my boss might have sent me the exact same little face earlier. It can’t represent them all. It’s like canned laughter, like a cliché. Yes, we know that emoji are useful lies, saving us time and dodging the difficulty of putting emotions into words; still, in so dodging, we falsify something essential about what our real faces can do. The emoji could be the culmination of everything we’ve done to the face over the centuries, all the multiplication and commodification and repacking. We’ve made it trivial.
This isn’t what I want for faces. I want them to shift, to resist, to allure. I find myself thinking of a photograph taken by Peter Hujar in 1975: a black-and-white portrait of the poet and dance critic Edwin Denby, at the age of seventy-two. His eyes are shut, his mouth neither smiling nor frowning.
The picture tempts you to suppose that you can read Denby’s face, that he’s experiencing fatigue, or stress, or agitation. Or maybe he’s just world-weary. But if you learn that Denby suffered, in his later years, from depression, and would kill himself some years later, you might see something like pain. Then you might learn a little about Hujar, too, and about the way he liked to concentrate on the emotions of his subjects. You might be less inclined to assume that a feeling has been casually captured, as it might be in a family snap. By this point, the question the image first posed, “What is this man feeling?,” seems as urgent as ever, perhaps even more charged. Yet every attempt to answer it has led nowhere. Denby sits there forever; his troubles are radiant.
Here’s another way of approaching that image. Denby, as a critic, was fascinated by the great dancer Vaslav Nijinsky, as much for his expressions as for his torso, arms, or legs. “The manner in which Nijinsky’s face changes from role to role is immediately striking,” Denby wrote, in 1943. “It is enhanced by makeup, but not created by it. In fact, a friend pointed out that the only role in which one recognises Nijinsky’s civilian face is that of Petrouchka, where he is most heavily made up.” It was as if the images held several traits in suspense. Many people who met Nijinsky were enthralled by that curious face. Stravinsky recalled that it “could become the most powerful actor’s mask I have ever seen.” It was part of Nijinsky’s weird magic, whether he was embodying a sex-hungry faun or a wooden puppet come to life: he drew other personae out of himself.
After Nijinsky began to show signs of schizophrenia, this mask, or persona, grew disturbed. He seemed to recede into his eyes. His diary, written during two feverish months in 1919, teems with scribbled drawings of eyes, giant and black. He believed that he was being watched, which for most of his life had been true. One winter night, in the late nineteen-twenties, he was taken out of an asylum in Paris by his former boss, the Ballets Russes’ Serge Diaghilev, to attend a production of “Petrouchka,” in which the title role was now performed by a younger man, Serge Lifar. Throughout the evening, Nijinsky barely spoke. He just stared. The diplomat Harry Kessler, who was there, said that his “big eyes” were like those of “a sick animal”; the dancer Tamara Karsavina caught him straining to look at her face, then, when she met his gaze, he “turned his head like a child that wants to hide tears.”
A photograph of the group survives from that night. Although Lifar was the star of the show, all the faces are turned to Nijinsky, whose gaze is directed offstage. His eyes and smile are too wide, and his right hand grasps the air. This face conveys a need, but a need that eluded everyone; it eludes me now, as I look at the picture, though I know it’s there. And, looking again at Hujar’s portrait of Denby, I see something similar. Another man, engulfed in feelings, shows his face to the world. His eyes are closed; his expression seems charged. I wish I knew how he felt. It seems important that I do. All I know is that I cannot look away. ♦