From: Zhang, Yangkun (Yangkun.Zhang@FMR.COM)
Date: Mon Oct 16 2000 - 14:53:49 PDT
> I believe the thing in question was called "FaceMaker." (Too lazy to
> hunt for the link for reasons described below.) After an initial write
> up, all the data (i.e. raw and evolved pictures, and the mechanisms for
DISCOVER Vol. 21 No. 2 (February 2000)
Table of Contents
Isn't She Lovely?
If you think that physical appeal is strictly a matter of personal taste
and cultural bias, think again. Who you find attractive, say
psychobiologists, is largely dictated by evolutionary needs and hardwired
into your brain
By Brad Lemley
She's cute, no question. Symmetrical features, flawless skin, looks to be
22 years old-entering any meat-market bar, a woman lucky enough to have
this face would turn enough heads to stir a breeze. But when Victor
Johnston points and clicks, the face on his computer screen morphs into
what a mesmerized physicist might call a discontinuous state of
superheated, crystallized beauty. "You can see it. It's just so
extraordinary," says Johnston, a professor of biopsychology at New Mexico
State University who sounds a little in love with his creation.
The transformation from pretty woman to knee-weakening babe is all the
more amazing because the changes wrought by Johnston's software are,
objectively speaking, quite subtle. He created the original face by
digitally averaging 16 randomly selected female Caucasian faces. The
morphing program then exaggerated the ways in which female faces differ
from male faces, creating, in human-beauty-science parlance, a
"hyperfemale." The eyes grew a bit larger, the nose narrowed slightly,
the lips plumped, and the jaw contracted. These are shifts of just a few
millimeters, but experiments in this country and Scotland are suggesting
that both males and females find "feminized" versions of averaged faces
Johnston hatched this little movie as part of his ongoing study into why
human beings find some people attractive and others homely. He may not
have any rock-solid answers yet, but he is far from alone in attempting
to apply scientific inquiry to so ambiguous a subject. Around the world,
researchers are marching into territory formerly staked out by poets,
painters, fashion mavens, and casting directors, aiming to uncover the
underpinnings of human attractiveness.
The research results so far are surprising-and humbling. Numerous studies
indicate that human beauty may not be simply in the eye of the beholder
or an arbitrary cultural artifact. It may be an ancient, hardwired,
universal, and potent behavior-driver, on a par with hunger or pain,
wrought through eons of evolution that rewarded reproductive winners and
killed off losers. If beauty is not truth, it may be health and
fertility: Halle Berry's flawless skin may rivet moviegoers because, at
some deep level, it persuades us that she is parasite-free and
consequently good mating material. Acquired, individual preferences
factor in, but research increasingly indicates that their influence is
much smaller than many of us would care to know. While romantic writers
blather about the transcendence of beauty, Elizabethan poet Edmund
Spenser more than 400 years ago pegged the emerging scientific thesis:
"Beauty is the bait which with delight allures man to enlarge his kind."
Implications of human-beauty research range from the practical-providing
cosmetic surgeons with pretty-people templates-to the political and
philosophical. Landmark studies show that attractive males and females
not only garner more attention from the opposite sex, they also get more
affection from their mothers, more money at work, more votes from the
electorate, more leniency from judges, and are generally regarded as more
kind, competent, healthy, confident, and intelligent than their
big-nosed, weak-chinned counterparts. (Beauty is considered such a
valuable trait by some that one entrepreneur recently put up a Web site
offering to auction off the unfertilized ova of models.)
Human attractiveness research is a relatively young and certainly
contentious field-the allure of hyperfemales, for example, is still hotly
debated-but those on its front lines agree on one point: We won't conquer
"looks-ism" until we understand its source. As psychologist Nancy Etcoff,
author of the 1999 book Survival of the Prettiest, puts it: "The idea
that beauty is unimportant or a cultural construct is the real beauty
myth. We have to understand beauty, or we will always be enslaved by it."
the modern era of beauty studies got a big push 20 years ago with an
awkward question in a small, airless room at Louisiana State University
in Baton Rouge. Psychology graduate student Judith Langlois was defending
her doctoral dissertation-a study of how preschool children form and keep
friendships-when a professor asked whether she had factored the kids'
facial attractiveness into her conclusions. "I thought the question was
way off the mark," she recalls. "It might matter for college students,
but little kids?" After stammering out a noncommittal answer-and passing
the examination-she resolved to dig deeper, aiming to determine the age
at which human beings could perceive physical attractiveness.
Nature or Nurture?
"The 50 Most Beautiful People in the World" assess the source of good
looks By Robert Sapolsky
As a scientist doing scads of important research, I am busy, very busy.
What with all those midnight experiments in the lab, all that eureka-ing,
I hardly have any time to read professional journals. Thus, I only lately
got the chance to peruse People magazine's most recent compilation of
"The 50 Most Beautiful People in the World." It was fabulous. In addition
to offering helpful grooming tips, the issue grapples with one of the
central conundrums of our time: Which is ultimately more influential,
nature or nurture? "About beauty," opine the editors, "the arguments can
be endless." No such shilly-shallying for the Chosen Ones themselves: The
50 Most Beautiful and their inner circles appear to harbor militant
ideologues in the debate.
Consider first the extreme nurturists, who eschew the notion that
anything is biologically fixed. There's Ben Affleck, who in service to
stardom has slimmed down, pumped up, and had his teeth capped. Affleck is
clearly a disciple of John Watson, famous for the nurture credo: "Give me
a child and let me control the total environment in which he is raised,
and I will turn him into whatever I wish." It's hardly surprising that
Affleck's celebrated affair with Gwyneth Paltrow, clearly of the genetic
determinist school (read on), was so short-lived.
A nurture viewpoint is also advanced by TV star Jenna Elfman, who
attributes her beauty to drinking 100 ounces of water a day, eating a
diet based on her blood type, and using a moisturizer that costs $1,000 a
pound. Jaclyn Smith, the erstwhile Charlie's Angel, maintains her beauty
has been preserved by not smoking, not drinking, and not doing drugs.
However, even a neophyte student of human developmental biology might
easily note that no degree of expensive moisturizers or virtuous living
would get, say, me on People's pulchritudinous list.
Naturally, similarly strong opinions emanate from the opposing, nature
faction-the genetic determinists among the Most Beautiful. Perhaps the
brashest of this school is Josh Brolin, an actor whose statement could
readily serve as a manifesto for his cadre: "I was given my dad's good
genes." Similar sentiments emerge from the grandfather of the
aforementioned Paltrow, who avows that she was "beautiful from the
The very epitome of the natalist program, in which genetics forms an
imperative trajectory impervious to environmental manipulation, is TV
host Meredith Vieira. People's editors cite various disasters that have
befallen her-shoddy application of makeup, an impetuous and unfortunate
peroxide job on her hair-and yet, it doesn't matter. She is still
beautiful because of her "phenomenal genes."
One searches the pages for a middle ground, for the interdisciplinary
synthesist who perceives the contributions of both nature and nurture. At
last, we espy Monica. The single-name singer, we are told, has an
absolutely wondrous skill for applying makeup. This, at first, seems like
just more nurture agitprop. But where does she get this cosmetic
aptitude? Her mother supplies the answer. With Monica, Mom says, "it's
something that's inborn." One gasps at the insight: There is a genetic
influence on how one interacts with the environment. Too bad a few more
people can't think this way when figuring out what genes have to do with,
say, intelligence, substance abuse, or violence.
Langlois, who had joined the faculty at the University of Texas at
Austin, devised a series of experiments. In one, she had adults rate
photos of human faces on a spectrum from attractive to unattractive. Then
she projected pairs of high- and low-rated faces in front of 6-month-old
infants. "The result was straightforward and unambiguous," she declares.
"The babies looked longer at the attractive faces, regardless of the
gender, race, or age of the face." Studies with babies as young as 2
months old yielded similar results. "At 2 months, these babies hadn't
been reading Vogue magazine," Langlois observes dryly.
Her search for the source of babies' precocious beauty-detection led her
all the way back to nineteenth-century research conducted by Sir Francis
Galton, an English dilettante scientist and cousin of Charles Darwin. In
the late 1870s, Galton created crude, blurry composite faces by melding
mug-shot photographs of various social subgroups, aiming to prove that
each group had an archetypal face. While that hypothesis fizzled-the
average criminal looked rather like the average vegetarian-Galton was
shocked to discover that these averaged faces were better looking than
nearly all of the individuals they comprised. Langlois replicated
Galton's study, using software to form digitally averaged faces that were
later judged by 300 people to be more attractive than most of the faces
used to create them.
Human beings may be born "cognitive averagers," theorizes Langlois. "Even
very young infants have seen thousands of faces and may have already
constructed an average from them that they use for comparison."
Racial preferences bolster the idea, say some scientists. History shows
that almost universally, when one race first comes into contact with
another, they mutually regard each other as homely, if not freakish.
Etcoff relates that a delegation of Japanese samurai visiting the United
States in 1860 observed that Western women had "dogs' eyes," which they
found "disheartening." Early Western visitors to Japan thought the
natives' epicanthic folds made the eyes appear sleepy and small. In each
case, Etcoff surmises, the unfamiliar race most likely veered from the
internal, averaged ideal.
But why would cognitive averaging have evolved? Evolutionary biology
holds that in any given population, extreme characteristics tend to fall
away in favor of average ones. Birds with unusually long or short wings
die more often in storms. Human babies who are born larger or smaller
than average are less likely to survive. The ability to form an
average-mate template would have conveyed a singular survival advantage.
Inclination toward the average is called koinophilia, from the Greek
words koinos, meaning "usual," and philos, meaning "love." To Langlois,
humans are clearly koinophiles. The remaining question is whether our
good-mate template is acquired or innate. To help solve the mystery,
Langlois's doctoral student Lisa Kalakanis has presented babies who are
just 15 minutes old with paired images of attractive and homely faces.
"We're just starting to evaluate that data," says Langlois.
But koinophilia isn't the only-or even supreme-criterion for beauty that
evolution has promoted, other scientists argue. An innate yearning for
symmetry is a major boon, contend biologists Anders Moller and Randy
Thornhill, as asymmetry can signal malnutrition, disease, or bad genes.
The two have found that asymmetrical animals, ranging from barn swallows
to lions, have fewer offspring and shorter lives. Evolution would also
logically instill an age preference. Human female fertility peaks in the
early 20s, and so do assessments of female attractiveness. Between 1953
and 1990, the average age of Playboy centerfold models-who are presumably
selected solely for sexual appeal-was 21.3 years. Similarly, Johnston has
found that the beauty of a Japanese female face is judged to be at its
peak when its perceived age is 22.4 years. Because men are fertile
throughout most of their adult lives, their attractiveness ratings-while
dropping as they age past their late 20s-remain relatively higher as
their perceived age increases. As Johnston puts it, "Our feelings of
beauty are exceptionally well tuned to the age of maximum fertility."
Still, a species can stagnate without some novelty. When competition for
mates is intense, some extreme traits might help to rivet a roving eye.
"A male peacock is saying, 'Look at me, I have this big tail. I couldn't
grow a tail this big if I had parasites,' " says Johnston. "Even if the
trait is detrimental to survival, the benefit in additional offspring
brought about by attracting females can more than compensate for the
decrease in longevity." The concept seems applicable to humans, too,
because it helps to resolve a nagging flaw in average-face studies. In
many of them, "there were always a few individual faces in the population
that were deemed even prettier than the average," says Etcoff. "If
average were always best, how could that be?"
Psychologist David Perrett of the University of St. Andrews in Scotland
aimed to find out by creating two averaged faces-one from a group of
women rated attractive and another from men so judged. He then compared
those faces with averaged faces constructed from a larger, random set of
images. The composites of the beautiful people were rated more appealing
than those made from the larger, random population. More surprising, when
Perrett exaggerated the ways in which the prettiest female composite
differed from the average female composite, the resulting face was judged
to be even more attractive.
"It turned out that the way an attractive female face differs from an
average one is related to femininity," says Perrett. "For example, female
eyebrows are more arched than males'. Exaggerating that difference from
the average increases femininity," and, in tandem, the attractiveness
rating. In the traffic-stopping female face created for this experiment,
200 facial reference points all changed in the direction of
hyperfemininity: larger eyes, a smaller nose, plumper lips, a narrower
jaw, and a smaller chin.
"All faces go through a metamorphosis at puberty," observes Johnston. "In
males, testosterone lengthens the jaw. In females, estrogen makes the
hips, breasts, and lips swell." So large lips, breasts, and hips combined
with a small jaw "are all telling you that I have an abundant supply of
estrogen, so I am a fertile female." Like the peacock, whose huge tail is
a mating advantage but a practical hindrance, "a small jaw may not, in
fact, be as efficient for eating," Johnston says. But it seems attractive
because it emphasizes la difference; whatever survival disadvantage comes
along with a small jaw is more than made up for by the chance to produce
more babies, so the trait succeeds.
Along with his morphing program, Johnston approached the hyperfemale
hypothesis through another route. Starting with 16 computer-generated
random female Caucasian faces, he had visitors to his Web site rate the
attractiveness of each face on a scale of one to nine. A second
generation of faces was then computed by selecting, crossing, and
mutating the first generation in proportion to beauty ratings. After
10,000 people from around the world took part in this merciless business,
the empirically derived fairest-of-them-all was born. Facial measurements
confirm that she is decidedly hyperfemale. While we might say she is
beautiful, Johnston more accurately notes that the face displays "maximum
Johnston's findings have set off a ruckus among beauty scientists. In a
paper titled "Attractive Faces Really Are Only Average," Langlois and
three other researchers blast the notion that a deviation from the
average-what they term "facial extremes"-explains attractiveness better
than averageness does. The findings of Perrett and his team, she says,
are "artifacts of their methodology," because they used a "forced-choice"
scenario that prevented subjects from judging faces as equally
attractive. "We did the same kind of test, but gave people a rating scale
of one to five," says Langlois. "When you do it that way, there is no
significant difference-people would tell us that, basically, the two
faces looked like twins." Langlois argues that if extremes create beauty,
"then people with micro-jaws or hydrocephalic eyes would be seen as the
most beautiful, when, in fact, eyes that are too big for a head make that
But for Etcoff, circumstantial evidence for the allure of some degree of
hyperfemininity is substantial. "Female makeup is all about exaggerating
the feminine. Eye makeup makes the brow thinner, which makes it look
farther from the eye," which, she says, is a classic difference between
male and female faces. From high hair (which skews facial proportions in
a feminine direction, moving up the center of gravity) to collagen in
lips to silicone in breasts, women instinctively exaggerate secondary
female sex characteristics to increase their allure. "Langlois is simply
wrong," declares Johnston. In one of his studies, published last year in
Psychophysiology, both male and female subjects rated feminized pictures
as more attractive. Further, male subjects attached to
electrical-brain-activity monitors showed a greater response in the P3
component, a measure of emotional intensity. "That is, although both
sexes know what is attractive, only the males exhibit an emotional
response to the feminized picture," Johnston says.
And what about male attractiveness? It stands to reason that if men
salivate for hyperfemales, women should pursue hypermales-that is, men
whose features exaggerate the ways in which male faces differ from female
ones. Even when adjusted for differing overall body size, the average
male face has a more pronounced brow ridge, more sunken eyes, and bushier
brows that are set closer to the eyes. The nose and mouth are wider, the
lower jaw is wider and longer. Ramp up these features beyond the norm,
and you've got a hunk, right?
There's no question that a dose of this classic "maleness" does
contribute to what is now called handsome. Actor Brad Pitt, widely
regarded as a modern paradigm of male attractiveness, is a wide-jaw guy.
Biologically speaking, he subconsciously persuades a female that he could
chew more nutrients out of a leafy stalk than the average potential
father of her children-a handy trait, in hunter-gatherer days anyway, to
pass on to progeny.
But a woman's agenda in seeking a mate is considerably more complex than
simply whelping strong-jawed kids. While both men and women desire
healthy, fertile mates, a man can-and, to some extent, is biologically
driven to-procreate with as many women as possible. Conversely, a woman,
"thinks about the long haul," notes Etcoff. "Much of mate choice is about
finding a helpmate to bring up the baby." In several studies, women
presented with the hypermale face (the "Neanderthal type" as Etcoff puts
it) judged its owner to be uncaring, aggressive, and unlikely to be a
Female preferences in male faces oscillate in tandem with the menstrual
cycle, suggests a study conducted by Perrett and Japanese researchers and
published last June in Nature. When a woman is ovulating, she tends to
prefer men with more masculine features; at less fertile times in her
monthly cycle, she favors male faces with a softer, more feminine look.
But amid the hoopla that this widely publicized finding generated, a
critical fact was often overlooked. Even the "more masculine" face
preferred by the ovulating women was 8 percent feminized from the male
average (the less masculine face was 15 to 20 percent feminized).
According to Perrett's study, even an averagely masculine face is too
male for comfort.
In matters of human beauty, hardwired preferences matter but can be
overcome. Novelist George Eliot (the pen name of Mary Ann Evans) was
strikingly homely, but her magnetic character inspired Henry James to
write in a letter: "She is magnificently ugly-deliciously hideous. She
has a low forehead, a dull grey eye, a vast pendulous nose, a huge mouth,
full of uneven teeth, and a chin and jaw-bone qui n'en finissent pas. . .
. Now in this vast ugliness resides a most powerful beauty which, in a
very few minutes, steals forth and charms the mind, so that you end as I
ended, in falling in love with her." To further complicate the
male-appeal picture, research indicates that, across the board in mating
species, an ugly guy can make up ground with status and/or wealth. Etcoff
notes that female scorpion flies won't even look at a male unless his
gift-a tasty bit of insect protein-is at least 16 square millimeters
wide. The human situation isn't all that different. Anthropologist John
Marshall Townsend showed photos of beautiful and homely people to men and
women, and described the people in the photos as being in training for
either low-, medium-, or high-paying positions-waiter, teacher, or
doctor. "Not surprisingly, women preferred the best-looking man with the
most money," Etcoff writes, "but below him, average-looking or even
unattractive doctors received the same ratings as very attractive
teachers. This was not true when men evaluated women. Unattractive women
were not preferred, no matter what their status."
it's all a bit bleak. talk to enough psychobiologists, and you get the
impression that we are all rats-reflexively, unconsciously coupling
according to obscure but immutable circuitry. But beauty researchers
agree that, along with natural selection and sexual selection, learned
behaviors are at least part of the attractiveness radar. In other words,
there is room for individuality-perhaps even a smattering of mystery-in
this business of attraction between humans.
"Human beauty really has three components," says Johnston. "In order of
importance, there's natural selection, which leads to the average face
and a limited age range. Then there's sexual selection," which leads men,
at least, to be attracted to exaggerated feminine traits like the small
lower jaw and the fuller lips. "Finally, there's learning. It's a
fine-tuning mechanism that allows you to become even more adapted to your
environment and culture. It's why one person can say 'She's beautiful'
and another can say, 'She's not quite right for me.' "
The learned component of beauty detection is perhaps most evident in the
give-and-take between races. While, at first meeting, different racial
groups typically see each other as unattractive, when one race commands
economic or political power, members of other races tend to emulate its
characteristics: Witness widespread hair straightening by American blacks
earlier in this century. Today, black gains in social equity are mirrored
by a growing appreciation for the beauty of such characteristically black
features as relatively broader noses and tightly curled hair. "Race is a
cultural overlay on beauty, and it's shifting," says Etcoff.
She adds that human appearance is about more than attracting sex
partners. "There was a cartoon in the New Yorker. A mother and daughter
are in a checkout line. The girl is saying to the cashier, 'Oh, no, I do
look like my mother, with her first nose!' As we make ourselves more
beautiful, we take away things like family resemblance, and we may
realize that's a mistake. Facial uniqueness can be a wonderful emotional
tag. Human beings are always looking for kinship as well as beauty."
Midway between goats and gods, human beings can find some accommodation
between the notion that beauty is all and that it is nothing. "Perhaps
it's best to enjoy the temporary thrill, to enjoy being a mammal for a
few moments, and then do a reality check and move on," writes Etcoff.
"Our brains cannot help it, but we can."
RELATED WEB SITES:
To participate in an experiment in human attractiveness or to see the
results of previous experiments, visit Johnston's Web site:
This archive was generated by hypermail 2b29 : Mon Oct 16 2000 - 15:00:44 PDT