What's In a Face?

Image: Glamorous woman looking in mirror
Jeff Riedel
Several years ago, a woman named Brook White appeared on the reality TV competition show American Idol. White was 24 years old, blond, and strikingly pretty. When she sang her song, "Like a Star," she struck a familiar chord among some viewers. White said nothing about her religion, but Mormons, members of the Church of Jesus Christ of Latter Day Saints, were certain that she was one of their own.

"She has the Mormon Glow," one blogger wrote, referring to the belief that the faithful radiate the Holy Spirit. White mentioned that she never drank a cup of coffee or watched an R-rated movie—signs of a Mormon-like squeaky-clean lifestyle. But the "glow" clinched it, and it turned out that her fans were right. "I didn't know I was setting off the Mormon radar," White remarked later in an interview with The Arizona Republic.

Soon after, psychologists Nalini Ambady, then at Tufts University, and Nicholas Rule, at the University of Toronto, set out to test the Mormon glow. One way to do this is to see if even non-Mormons can detect it. The psychologists began their experiment by cropping head shots of Mormons and non-Mormons and asking undergraduate volunteers whether they could pick out the Mormons.

Find a Therapist

Search for a mental health professional near you.

They certainly could—and in just a glance. While random guessing would yield 50 percent accuracy, as in a coin toss, the volunteers accurately identified Mormon men and women 60 percent of the time. (Mormons themselves were only slightly more accurate.) This means that "Mordar" isn't foolproof, but it's statistically significant—about as accurate as the ability to tell if a face looks envious or anxious.

"Thin-slicing" is the term that Ambady and her colleague, Richard Rosenthal, coined in 1992 to describe the ability to infer something about a person's personality, character, or other traits after a very brief exposure. Thin-slicing relies on a brain network that includes the fusiform gyrus, which perceives faces, and the amygdala, which filters that information for anything that might be useful or threatening to survival.

To determine what exactly triggers Mordar, Ambady and Rule cropped photos beyond recognition. Some faces had only eyes or hair. Could judges identify Mormons from these features alone? Fail. Others had only noses or mouths. Nothing. Other faces had no features or even an outer shape. Just a patch of flesh, basically. Success.

"What the judges were primarily picking up," Rule explains, "are cues of health in the skin." The tone and texture of facial skin reflects immune function. "We have a system set up to assess others' health for mate selection and disease avoidance," Rule says. "This can be co-opted for social purposes as well —such as detecting religiosity."

Mormons don't drink or smoke. They enjoy community support, which relieves stress. They live 10 years longer than the average American. Holy Spirit aside, their skin may glow because it's healthier. While the judges likely knew that Mormons are clean-living, they weren't consciously aware when categorizing faces that they were associating religious purity with good skin. It was a gut feeling.

Over evolutionary time, the ability to quickly extract information from faces has given us an edge in predicting character and behavior. It helps us to discern who's sick and whom to trust, who's flirtworthy, and who might blow up at a moment's notice. To get a sense of others' religiosity, sexual orientation, promiscuity, aggressiveness, competence, intelligence, and even trustworthiness, you might think that you should focus on how they act, not how they look. But then you'd neglect your swiftest insights.

Snap judgments about faces are not fail-safe, but they are far too accurate to ignore, especially in moments when danger may loom. And the consistent above-chance rate (often 60 percent or higher) at which we correctly extrapolate character traits from facial features taps into core questions of identity: Are appearance and behavior linked by underlying biological processes such as hormones—or do beautiful visages and aggressive miens elicit lifelong treatment that reinforces an individual's behavior? It is the ultimate chicken-and-egg question, and it plays out on every face you meet.

I.

BEYOND THE HALO EFFECT: Attractiveness and Personality

Our assessment of attractiveness is automatic, and strongly influences how we judge the person on a range of other traits, including personality. In a surprising experiment at the University of Pennsylvania, neuroscientists Ingrid Olson and Christy Marshuetz asked volunteers to judge pre-rated faces, some beautiful and others homely. The judges claimed that they didn't see anything; the faces flashed by too quickly. Yet when coaxed to rate the attractiveness of the faces that they thought they hadn't seen, they were astonishingly accurate. Each face was exposed for 13 milliseconds, well below the threshold of conscious awareness. That's how quickly we judge looks.

Beauty and health are tightly linked. The closer a face is to the symmetrical proportions of Gwyneth Paltrow or Zac Efron, and to the average face in a population, the more it advertises developmental stability, meaning that pathogens or genetic mutations have not adversely affected its owner.

Good looks also confer a well-documented "halo effect": a beautiful man or woman is consistently evaluated in a positive light. Good-looking people are assumed to be smarter than their homelier peers, although there is no correlation between intelligence and appearance above a median level of attractiveness.

Appearance interacts with personality in complicated ways—good-looking people are consistently rated higher on positive traits. When volunteers were asked to evaluate faces in a UK study, the most attractive individuals received the highest ratings for extraversion and agreeableness. Yet more than the halo effect is at work, because the owners of those good-looking faces also rated themselves to be higher on these traits. More impressively, when judges looked at digital composites (averages of faces) made from people who scored at the extremes for extraversion and agreeableness (and, for women, openness), they gave those faces the highest attractiveness ratings. The judges didn't know that the composites were made from the faces of exceptionally outgoing and easygoing people. They just thought those faces were better-looking than average. (For men only, facial composites generated from the most conscientious and emotionally stable subjects were also rated as more attractive than those made from subjects with the lowest scores for these attributes.) Clearly, the stereotype "what is beautiful is good" contains at least a kernel of truth.

Here, then, is the big chicken-or-egg puzzle that runs throughout face perception research. Do the biological blessings behind good looks give rise to a sparkling personality; or do attractive people exhibit the socially desirable traits of extraversion and agreeableness because society treats swans better than ugly ducklings? Or do individuals with attractive personalities develop more attractive faces over time? Whether nature or nurture, the relationship between beauty and "positive" personality traits is real—and readily discernible.

Sex hormones are one clear link between appearance and personality. Testosterone and estrogen influence facial development as well as behavior. High testosterone shows itself in strong jawbones, darker coloring, and hollower cheekbones. High estrogen reveals itself in smooth skin, a small chin, sparse facial hair, arched eyebrows, and plump lips.

We make numerous assumptions about people with high-hormone profiles that conform to gender norms: First, that they're hot. In a lineup, the high-estrogen Jessica Alba and Beyoncé types receive the highest attractiveness ratings by both genders. Their pretty faces predictably get top ratings for social dominance (high status). As for men, high-testosterone faces are especially desired by women who are ovulating, although women may have a default preference for men with a mix of masculine and feminine features; dominant and cooperative. Think Brad Pitt's manly jawline and sensuous lips.

At the University of St. Andrews, volunteers of both genders could tell, with above-chance accuracy, whether people were promiscuous (open to one-night stands and sex without love) just by looking at photos of their faces. Among women, high-estrogen feminine faces were accurately rated as the most promiscuous—and the most beautiful. Among men, the Lothario face (a composite of the most promiscuous males) had high-testosterone features: slightly smaller eyes, larger noses, and broader cheekbones. Women accurately judged this face as belonging to a playboy and downgraded it in favor of men who looked—and actually were—more committed and monogamous.

Do highly feminine-looking women and masculine-looking men have hormone profiles that give rise to stronger sex drives, or do their looks simply lead to more sexual opportunities? The likely answer is both: nature and nurture are inseparable. And yet, there's a clear message. The next time you're perusing photos on an online dating site and get a suspicious feeling about a person's romantic trustworthiness, you might listen to that instinct.

Image: Business Woman

Jeff Riedel
II.

GAYDAR: The Dorian Gray Effect

No one was shocked when Suze Orman jokingly outed herself as a 55-year-old "virgin"—a lesbian who had never slept with a man. When the singer Adam Lambert came out, nobody blinked. Even without all the circumstantial evidence, we might have a feeling about the sexual orientation of these two celebrities just by looking at their faces. "Gaydar"—the ability to determine in a glance whether someone is gay or lesbian—depends in part on gender norms. Some faces appear "gender inverted": males with some feminine features, females with some masculine ones.

Curious about gaydar's reliability, Ambady and Rule devised experiments in which they asked volunteers to take a look at close-cropped head shots and guess whether each face belonged to someone who is gay, lesbian, or straight. Impressively, most people could identify sexual orientation in just a split-second thin-slice. "People can even identify orientation by mouth and eyes alone," says Rule.

The more motivated you are to know someone's orientation, the better your intuition. People with the sharpest gaydar are gay men and lesbians, naturally, and ovulating women. Subjects identify lesbians accurately between 64 and 70 percent of the time; gay men are correctly identified with slightly less reliability, in the 60 to 65 percent range. "The more we ask people to think about their choices before they make them, the worse they are," says Rule. "When they don't go with their gut, they fail."

It's easy to conclude that the ability to detect sexual orientation depends on biological cues alone —the gay man with doe eyes or delicate cheekbones, the lesbian with a strong jawline. While there's a kernel of truth here, it doesn't yield anything like a full crop. Metrosexuals trigger false alarms; lesbian femmes and gay Marlboro men often ride under the gaydar.

What's especially interesting is when gaydar goes off in the absence of gender-atypical cues. It's just a feeling, like those that fueled rumors about actor Neil Patrick Harris (before he came out). Observers may be picking up on cues that have more to do with "nurture" (experience) than "nature" (biology). Rule has a theory that gay-related expressions may create "repetitive patterns of musculature that result in a certain look." Stereotypically, gay men are more emotionally expressive than straight guys, adopting more female-typical facial movements, and some lesbians may express themselves more like straight men. It's the "Dorian Gray Effect": Appearance reflects behavior in telltale ways, just as Gray's portrait did in Oscar Wilde's novel.

The Dorian Gray Effect also explains why we can single out people who are chronically cantankerous or ever-exuberant: Stress and laugh lines linger even when the face is still. It also solves the mystery of why old married couples grow to look like each other; for decades, they've been mirroring each other's facial expressions.

Does the face we're born with reflect the way we behave, which influences how others see us, which affects the way we behave? This question cuts to the heart of identity and is especially pressing for men with certain facial characteristics.

III.

AGGRESSION AND CRIMINALITY: A Telltale Ratio

Several years ago, Cheryl McCormick, a neuroscientist at Brock University in Ontario, was listening to a radio program about the effects of testosterone on male skull size during puberty. Under the influence of the hormone, men's facial width increases in relation to height (width-height ratio or WHR), independent of body size. WHR is the distance measured from cheekbone to cheekbone versus the distance between the top of the lip and midbrow. A high WHR is 1.9 or above. Bill Clinton's is 2.07; John Edwards's, 2.38; and Richard Nixon's, 2.02; compared to John Lennon's, 1.63, and George Washington's, 1.65. (Now try to resist eyeballing every man's WHR.)

Intrigued, McCormick and her colleagues, psychologist Cathy Mondlach and graduate student John Carre, wanted to know if a high WHR might be a biomarker for aggression, which is also related to high testosterone levels. They asked volunteers to guess from photos of male Caucasian faces which ones were the fighting type.

A surprise: Not only could the judges in her study accurately detect which men are aggressive based on a headshot alone, they could do so after seeing the face for only 39 milliseconds. These men had the highest WHRs. Unknown to the judges, they also scored highest on an aggression test that measures how often a player steals points from an opponent, to no benefit for the player himself. (Women with high WHRs are perceived as more aggressive than average but aren't, presumably because the skull-shaping pubescent testosterone surge affects males only.)

Male face width is also associated with a propensity to deceive. A business school professor now at UC Riverside, Michael Haselhuhn, and his colleague Elaine Wong, discovered that men with high WHRs were three times likelier than their narrower-faced peers to lie to increase their financial gain in hypothetical scenarios: a lottery drawing and a buy-sell negotiation that occurred over email. While most men with high WHRs—60 percent—did not break the rules, this finding is still startling.

Don't dismiss your instincts, especially if your safety and well-being are at risk. Guesses about trustworthiness based on headshots tend to correlate with those individuals' self-reports and judgments by their acquaintances. In one experiment, people who were perceived as dishonest were likelier to mislead their peers than were those whose faces were thought to look honest. In studies involving the prisoner's dilemma game, participants, going by facial photos alone, could accurately identify people who were likely to deceive, and also remember the faces of prospective cheaters more than cooperators. Consciously or not, you may be more cautious and attentive around potential cheaters and other offenders.

Image: Psychotic woman

Jeff Riedel
At Cornell University, psychologist Jeffrey Valla and his colleagues set out to test just how readily people can spot criminals based on facial appearance alone. They prepared close-cropped, expressionless, facial photos of clean-shaven Caucasian men in their twenties and asked volunteers to identify the murderers, rapists, thieves, forgers, drug dealers, and so on. Men and women alike could distinguish convicts from noncriminals with above-chance accuracy, but, interestingly, not violent offenders from nonviolent ones. While one might expect violent criminals to be more readily discernible, the fact is that people who commit one type of crime are more likely to commit others. The boundary between violent and nonviolent offenders can be delineated in the lab, but not as readily in the real world. It is adaptive to be sensitive to any indices of potential criminality.

Criminal identification ("facial profiling") isn't hugely reliable, in part because there are too many false alarms. "Our participants exhibited a 'better safe than sorry' approach to their judgments, erring on the side of accusation," says Valla. And yet, one criminal type consistently crept under the radar: Rapists. Shockingly, women (but not men) thought that the convicted rapists would be less likely to commit a crime than anyone else. Nonthreatening, even attractive, their faces didn't fit the stereotype. There may be a slyly adaptive reason for this disparity. Men who look aggressive or dangerous will have a harder time luring potential victims into situations where a rape can take place. A lesson: While our instincts can protect us from harm in a general sense, we can't depend on them to protect us from sexual assault.

What exactly are judges picking up on in studies of deceit and dangerous individuals? The most obvious features are those associated with high testosterone—a high facial width-to-height ratio along with a superhero jawline and brow ridge —because of their association with a strong sense of power. It's not hard to imagine how high testosterone may lead to heightened aggressiveness, which could lead to a willingness to break rules. For some people, this behavior seems to be in their nature.

But nurture (experience) may also change a face in telltale—albeit subtle—ways. Over time, people who have acted disingenuously may develop a slightly crooked mien because they make more asymmetrical facial expressions. Or, we may be picking up on slightly angry expressions; they're associated with untrustworthiness, as found in a study by Alexander Todorov at Princeton. Not coincidentally, anger increases the face's WHR (raising the upper lip and lowering the brow), and chronic anger may linger on a face. Committing repeated aggressive or violent acts might also change the perpetrator's face in telling ways. Valla speculates that this behavior exposes a person to chronic high levels of testosterone and the stress hormone cortisol, and that we may detect the influence of these hormones on facial skin tone and shape. The Dorian Gray Effect strikes again.

"Even if there is a kernel of truth to the WHR-behavior link," Valla warns, "it could be overextended in false assumptions about individuals who might not otherwise be aggressive or deceptive, leading to self-fulfilling prophecy effects that appear to confirm appearance-behavior links." Accustomed to being treated with fear or deference, masculine-faced men may learn that they can get away with acting deceitfully. Or they may be more likely to lie or strike out because others expect it of them.

IV.

DOMINANCE: When Features Cut Both Ways

High testosterone has two faces. One is the devious, aggressive cheater. The other is a strong and capable leader. Both of these stereotypes play out in face perception research.

Among business leaders, powerful-looking features are predominant. When Haselhuhn and Wong analyzed Fortune 500 company executives, they found that firms led by men with high WHRs had superior financial results compared with firms led by men with lower WHRs. Similarly, the undergrads in Ambady and Rule's experiments guessed with above-chance accuracy which faces belong to Fortune 1000 CEOs and managing partners of the most profitable law firms based solely on how dominant those faces looked.

While masculine-faced leaders may appear more competent, it's unclear if they really are so. The most profitable firms may simply hire dominant-looking people to be their public face. Equally possible is the self-fulfilling prophecy: Parents and teachers may groom these men from an early age to be leaders, so they see themselves as such, as do others.

Nevertheless, it's a statement about corporate dynamics that an appearance of dominance, not warmth, also predicted which faces belonged to the most successful female CEOs. Will Yahoo!'s sweet-looking Marissa Mayer, dubbed "The Hottest CEO Ever," crack the high-WHR ceiling?

For women, competence can also be conflated with comeliness. Shawn Rosenberg, a political scientist at UC Irvine, presented photos of the same woman appearing in two faux campaign photos. In one, she's professionally made up, and in the other she looks dowdy. Regardless of whether she ran as a Democrat or Republican, she won about 56 percent of the vote—a serious margin—when portrayed by the flattering photo. In a separate mock election study, Rosenberg asked judges to rate headshots of candidates in terms of how competent they appeared. The traits associated with an edge: a strong curve of the upper eyelid, thinner lips, thinner eyebrows, a broad face, and, oddly, a widow's peak.

"This is politics. Perception is reality," is an old adage, and so it goes: A recent Princeton University study found that inferences of competence based solely on facial appearance predicted about 70 percent—a staggering number—of the outcomes of U.S. Senate races in 2004. Asking the judges to first think about their decision reduced their accuracy. We vote with our gut.

That said, the faces of elected leaders may depend on the situation. In a study at the University of Aberdeen, people who were primed to choose a president in a wartime context preferred a morphed George W. Bush face, judging it to be more masculine; in a peacetime context they preferred a morphed John Kerry face, judging it to be more intelligent. In another experiment, volunteers guessed—with about 60 percent accuracy—which faces belonged to Republicans, perceived as more dominant, and which were Democrats, seen as warmer.

Generally speaking, our biases about what competent leaders should look like appear to be hard-wired, or at least learned early. In a Swiss experiment, schoolchildren were presented with pairs of headshots of actual political candidates and asked which of the two they would choose to be the "captain of their boat." Over 70 percent of the kids chose the candidates who would later win the election. They looked like leaders.

V.

FACING THE LIMITS: The Dangers of Stereotyping

What does all this mean for those who don't look authoritative, such as baby-faced men? Their pudgy, youthful faces are the yin to the masculine-faced yang. They're perceived as soft, suggestible, incompetent, and honest to a fault.

Leslie Zebrowitz, a social psychologist at Brandeis, and her colleagues had a theory that baby-faced boys would strive to confound these expectations. Drawing on archival data that included photos of men over time and information about their socioeconomic status, race, grades, and IQ scores, she discovered that the hunch is true—yet manifests in complex ways.

A baby-faced boy who was white, middle-class, or had a high IQ was apt to compensate for the perception of a childlike intellect by getting good grades. But if he came from a poor family and scored poorly, he was likelier to compensate with assertive, hostile—even criminal—behavior, especially if he was also short. Would history be the same if George "Baby Face" Nelson, the toughest of Depression-era mobsters, hadn't been born in the slums of Southside Chicago? We'll never know for sure. Nature and nurture—the chicken and the egg—cannot be unscrambled. The face we're born with reflects personality and is molded by experience, which in turn reflects how we are perceived.

Another pitfall of face perception presents itself in Zebrowitz's research on deception. She and her colleagues asked volunteers to rate people's trustworthiness based on headshots taken throughout their lives and compared the ratings of each face with its owner's scores on personality tests. While they found that men's trustworthiness could be predicted from an early age, women's could not. Women who were less honest in their youth were judged as more honest-looking in adulthood, even if they weren't actually more trustworthy. These ladies could improve their appearance with cosmetics and hairstyle, which—thanks to the halo effect—made them appear more honest. "Dishonest women may be more likely to look honest than dishonest men because [women] have less power to achieve their goals through other means," the researchers suggest.

Image: Headshot of plain woman

Jeff Reidel
Our biases can come back to bite us. By overvaluing attractiveness, we may elect bad leaders and hire the wrong employees, believing them more honest, capable, and intelligent than they are. By conflating competence with dominance, we may discourage nurturing or egalitarian leadership styles. By pigeonholing all gender-atypical appearances as gay, we limit diversity of expression. To categorize all wide-faced men as aggressive is not only inaccurate, it distorts our perception of other facets of their personalities. We slip into superficiality.

Snap judgments about faces arise in an ancient region of the brain that specializes in self-protection. Apply them too broadly, and we risk turning a survival mechanism into knee-jerk prejudice. Compassion, fairness, and rational decision-making happen only when the slower, more recently evolved prefrontal cortex weighs in.

A lesson comes from one of our first American idols, Abraham Lincoln. The Great Emancipator's face is a battleground for our biases. With a WHR of 1.9, he had the biomarker of dominance and authority yet still had a reputation for integrity. But Honest Abe enjoyed no halo effect. From an early age, his facial skin was yellowish and creased, revealing his poor health. He was severely asymmetrical; the left half of his face sagged sadly. Evidence that he was chronically depressed rests alongside gossip that he carried on a lifelong love affair with another man. Today, cameras would be in his face and we'd all be judges. The question remains: Would we give him a chance?

Jena Pincott is a PT Blogger and science writer. Her latest book is Do Chocolate Lovers Have Sweeter Babies?: The Surprising Science of Pregnancy.

The Power of Stereotypes

Use them, but do so with care. How the experts manage their own gut instincts and biases:

"Snap judgments are most useful for dealing with strangers and quick encounters," advises Cheryl McCormick. "We've evolved to err on the side of caution," she says, because it was safer in social exchanges with outsiders—to avoid disease, rejection, violence, unsuitable mates, and so on. "I remind myself that these judgments are good for groups, but they don't have a lot of predictive power at the level of the individual. While we shouldn't judge a book by its cover, we can judge the library by its books."

Barring dangerous situations in which snap judgments can be useful, Leslie Zebrowitz hopes that her research will lead us to withhold hasty judgments. "Knowing that first impressions may well reflect over-generalizations will, ideally, lead people to wait for more diagnostic data before reaching any conclusions about someone." Yet Nicholas Rule reminds us that too much information can make us less accurate. "These feelings are coming for a reason, from the unconscious," he argues. "I let the feelings come up and then assess them. If I feel a little suspicious, maybe I'll realize it has to do with [facial characteristics]. It's all anecdotal, but I've had incredible luck with intuition steering me in the right direction."

Why Eyes Are So Revealing

"The countenance is the portrait of the soul," Cicero said, "and the eyes mark its intentions." The pupils advertise desire. Women's dilate more widely during the hormone surge before ovulation and when looking at attractive men, and men find large-pupiled women more attractive without knowing why. Women prefer men with medium-sized pupils. There are two exceptions: women who are ovulating and those who prefer a "bad boy." They both like big pupils.

An equally subtle cue is the eye's limbal ring, the black circle that separates the colored iris from the white sclera. When people judge a pair of faces, one with limbal rings and the other without, they strongly prefer the former even though they can't consciously detect a difference. The limbal ring, which shrinks with age, is a signal of youth and health—desirable qualities, reproductively speaking.

 

Current Issue

Let It Go!

It can take a radical reboot to get past old hurts and injustices.