91av

ONCE upon a time, the evil queen was “the fairest of them all”—her
mirror told her so. Then a new beauty came on the scene. Even the mirror could
not resist her milky skin, blood-red lips and hair as black as ebony. When the
evil queen asked her mirror who was the fairest in the kingdom, it now responded
“Snow White”. And the mirror never lies.

In another land, far, far away, some clever people in white coats are also
consulting their mirrors. But they want to know about intelligence, not beauty.
Confident in their standing as members of the smartest species in the animal
kingdom, they are using mirrors to delve into the minds of other creatures. By
testing how animals react to their reflections, scientists hope to discover how
they think. The problem is, they can’t agree about what it means to recognise
your reflection. Some say it’s the true test of self-awareness—the ability
to understand one’s own existence—but others argue that the mirror’s
message has been misunderstood.

The tale begins one morning in the 1960s, about 100 years after Charles
Darwin first put two orang-utans in front of a mirror and watched them kiss
their reflections. Graduate student Gordon Gallup was shaving and as he stared
back at his own reflection, he had an idea. “It dawned on me—gee, wouldn’t
it be interesting to try to determine whether there were other species that
could recognise themselves in mirrors?”

Chimpanzees seemed the obvious first subjects, being our closest living
relatives. Gallup found that, confronted with their mirror image, chimps soon
realise that the reflection is their own. He followed up this finding with a
simple but elegant experiment that was published in 1970. After administering a
general anaesthetic, he painted brightly coloured marks on the chimps’ eyebrows
or near their eyes—places they couldn’t see without looking in a mirror.
On awakening, the chimps noticed the change in their reflection right away,
touching the dyed areas, then smelling and looking at their fingertips.

Since then, researchers have subjected dozens of species to Gallup’s mirror
test—including dogs, cats, birds, elephants, and more than 20 species of
monkey. So far, the only ones that have conclusively passed are the great apes:
chimpanzees, orang-utans, and one gorilla (Koko, the human-reared ape that has
learned sign language). But this elite club may include more members than just
our close cousins.

In the early 1990s, Gallup and his colleagues Laurie Marino from Emory
University in Atlanta and Diana Reiss from the New York Aquarium found that
bottlenose dolphins reacted to mirrors. “It turned out to be suggestive but
inconclusive,” says Reiss. One difficulty is interpreting the mark test in an
animal that lacks arms or hands to explore its body. But a study soon to be
published by Reiss and Marino may provide a definitive answer to the dolphin
question.

Mirror studies have also identified many animals that “understand” what
mirrors do, yet cannot solve the riddle of their own reflection. They react to
their image as if it’s another member of their species. Monkeys, pigeons,
parrots, elephants, ducks, chickens and even fish can use mirrored information
to find hidden objects or to solve a puzzle. In one of the most extreme cases,
Gallup reared a pair of rhesus monkeys from two months old, exposing them to a
mirror 14 hours a day, 7 days a week, for 18 years. “They had to have been the
most mirror-experienced creatures in the cosmos,” he says. “But when it came to
responding to themselves they were at a complete and total loss.”

Animals that do pass the mirror test behave in a stereotypical way towards
their reflection. First, they act as if they are seeing another individual. But
very quickly—in as little as five minutes—they begin to engage in
unusual body movements. Mesmerised by this mimic in the mirror, they slowly,
deliberately and repetitively move their arms, legs or head. Then they do the
same sort of thing with their face, making wild contortions. When they grasp the
equivalence between themselves and their mirror image, they begin to explore
their body, using their hands or feet to systematically inspect those parts they
couldn’t otherwise see. “It’s really something to see,” says Daniel Povinelli,
an anthropologist from the University of Southwestern Louisiana in Lafayette.
“They’ll get into these exaggerated postures in front of the mirror, from the
most studied, serious explorations of their teeth and eyes to hilarious body
contortions to look at their anogenital region.”

Gallup, now at the State University of New York in Albany, believes that
passing the mirror test is good evidence that an animal has a concept of self.
“If you didn’t know who you were,” he says, “how could you possibly know who it
was that you were seeing in a mirror?” He also believes that recognising a
mirror image, in principle, allows an individual to do three things. First, to
think about itself in relationship to past, present and future events. “In other
words, you can begin to engage in mental time travel,” he says. Secondly, to
contemplate the inevitability of its own demise. And thirdly, to use your
experiences to make inferences about the psychological state of others. So a
chimp can make the leap from feeling pain when hit by a rock, to knowing that
another chimp will feel pain when pummelled.

Gallup acknowledges that these three traits have been tricky to test. There
have, however, been experiments that suggest that animals who pass the mirror
test are capable of “mind reading”. These tests look for signs of what Gallup
calls “introspectively based social strategies”, such as empathy, gratitude,
deception, role-playing and pretending. “All of those capacities require the
ability to track mental states in other individuals,” he says. In one test, for
example, chimps had to choose one of two humans to help them find some hidden
food. While the animals themselves could not see where the food was being
hidden, they could see that only one of the humans had a full view of the
process. When asked to choose a helper, the chimps overwhelmingly chose the
human who knew where the food was hidden. Monkeys fail these sorts of tests. But
even chimps don’t pass them all.

According to Gallup, the best support for his mind-reading idea comes from
mirror studies with human infants. At about six months, children begin to react
to their mirror image as if it were another child. They start to show evidence
of mirror self-recognition at 18 months, and by 24 months about 65 per cent
recognise themselves. At the same stage in their development they begin to use
the words I, you and me, and exchange roles in play. Most compelling, says
Gallup, is the finding that as children start to recognise themselves, they also
begin to infer the mental state of others. Before about 18 months, a child who
sees another child crying will join in. But after 18 months, that child will get
help from an adult or try to comfort the distressed child.

Gallup is working with his colleagues to pinpoint the area of the human brain
that mediates self-awareness and mental states such as gratitude and deception.
He says both seem to reside in the same region: the right prefrontal cortex.
“And it turns out that the part of the human brain that’s growing the most
rapidly at 18 to 24 months is the prefrontal cortex. We haven’t proven any of
this, but the data is very consistent with the proposition that self-awareness
paves the way for the ability to make inferences about others.”

But not everyone agrees with Gallup’s conclusions. “On the surface, such
findings seem to imply that chimps can understand the mental state of others,”
says Povinelli. “Because the chimp at some superficial level engages in a
similar behaviour to us, we force our human interpretation upon them. It’s the
ultimate anthropocentric conceit.” In fact, says Povinelli, even mirror studies
with our humans show how little we know about what self-recognition means. Take
a look at your reflection. When you say, “Ah, that’s me,” what pops into your
head? A lot of baggage: your desires and wants, your past, your personality
traits, aspects of your body, and much more. “Are all of those captured when an
animal or child exhibits the kinds of behaviours in front of mirrors that have
been called self-recognition?” asks Povinelli. “I don’t think so.”

Povinelli describes experiments that he and his colleagues are conducting to
unravel the mysteries of self-perception in young children. These involve
secretly hiding a sticker on top of a child’s head and then playing back the
event on video. If young children really have the same self-recognition as
adults you would expect them to say: “That’s me and you put a sticker on my
head.” What the researchers found, however, was that two-year-olds reach up for
the sticker if they watch the video in real time—the equivalent of seeing
themselves in a mirror. But if you introduce even a brief delay, they say the
most bizarre things. You ask them who it is and they say: “That’s me. But why is
he wearing my shirt?” You point to the sticker on the monitor and ask them what
it is, and they say: “That’s a sticker on his head.” It’s not until they’re
about four years old that children pass a delayed test. “If we find those kind
of surprises in members of our own species, imagine what we’re going to find
when we seriously look at others,” Povinelli reported in 1998.

What’s more, human infant mirror studies reveal no correlation between
whether they pass the mark test and whether they understand what mirrors do. If,
for example, they face a mirror and an entertaining toy is silently suspended by
a string behind them, every infant smiles at the mirror image. About half of
them will turn around to see the real toy. And on a mark test, about half will
touch the mark. But it is not the same half. “Now, for most traditional
intuitive theories of what’s going on, that’s almost incomprehensible,” says
Povinelli.

It does make sense, he argues, if you abandon the assumption that infants or
chimps understand anything at all about what mirrors actually do. As they stand
in front of a mirror, what they say is, “That out there in space, whatever that
is, is equivalent to me.” “In other words,” says Povinelli, “they have a concept
of their own body and what it’s doing—called a kinaesthetic self-concept.”
He believes orang-utans hold the key to understanding the origins of this
self-concept. They are unexpectedly good at the mirror test, given that they are
more distantly related to humans than are gorillas. What’s more, unlike other
primates, orang-utans do not live in large social groups, challenging the common
argument that self-awareness evolved to allow social interactions that involve
predicting and gauging the behaviour of others. “You have these lonely
orang-utans who are masters of self-recognition, but who are pretty much living
solitary lives. They are oddballs,” says Povinelli. He thinks he knows why.

Like our earliest ancestors, orang-utans are almost exclusively tree
dwellers. What’s more, unlike monkeys, they are big: males weigh up to 80
kilograms. They have to be acutely aware of their own bodies if they are to move
safely through the forest canopy. Povinelli speculates that arboreal acrobatics
just like those performed by orang-utans led early apes to evolve a more
explicit understanding of the position and movements of their body. This enabled
them to plan the effects of their actions on the environment around them.
Self-recognition in mirrors, he says, is an “incidental by-product” of this
ability.

Povinelli may be going out on a limb, but other researchers also question
traditional interpretations of the mirror test. Many believe that it
underestimates the mental capacities of animals that fail it. “We have to stop
saying that this is the test for self-awareness,” says Marc Hauser from Harvard
University. After all, he adds, animals don’t have to be self-aware to behave in
a way that suggests—incorrectly—the ability to infer the thoughts of
others. Look at the birdbrained plover, which fakes an injured wing to lure a
potential predator away from its nest. On the other hand, even among humans,
self-recognition is not synonymous with self-awareness. Blind people obviously
can’t see themselves in a mirror, and some people with a type of brain damage
called prosopagnosia can’t recognise their own faces. Yet both are
self-aware.

Hauser believes that some animals may fail the mark test not because they
lack a concept of self, but because they are not interested in their new
coloured splodges. A few years ago, he painted bright green marks on the arms of
cotton-top tamarins under anaesthetic. When they awoke, the primates didn’t even
bother to touch their new markings. But in another experiment Hauser dyed the
white tufts of fur on the tamarin’s heads flamingo pink, apple green or lagoon
blue, and this time they did appear interested in their reflections. No one has
been able to replicate the results of the second experiment, however.

Other researchers point out that monkeys may fail the mirror test partly
because they find their image too threatening. “It’s like the gunfight at the OK
Corral,” says James Anderson from Stirling University, who has been trying for
20 years to get monkeys to recognise themselves. “They’re too fixated on the
eyes to understand that they’re looking at themselves.” Anderson tried to get
around that by fixing the mirrors at 60° angles so that the monkeys could
only see their profile. “The monkeys did spend a lot more time looking at their
reflection, bobbing their heads from side to side.” Yet they still did not seem
to recognise themselves.

The fact that animals never encounter mirrors in the wild has led some
researchers to suggest alternative tests of self-awareness. In a recent study,
Robert Seyfarth and Dorothy Cheney from the University of Pennsylvania looked at
social recognition in baboons. They examined how pairs of females reacted to the
recorded sounds of other adults in a fight. If the adults were not their own
kin, they didn’t react. If the cries came from one of the female’s relatives,
however, the other female would look at her. And if both were related to the
brawlers, the two females looked at each other. The researchers concluded that
baboons know individuals and their calls, and also the family relationships of
each individual. This knowledge of their social selves, says Seyfarth, raises
the possibility that they have a sophisticated sense of self.

Other animals may be failing the mirror test because they are not as visually
attuned as primates. Cats and dogs, for example, are largely dependent on their
sense of smell, which could explain why they tend to show little interest in
their reflection.

Some researchers are now looking for alternative tests of self-awareness for
the many animals that may recognise themselves through the smell of their own
pheromones or the sound of their calls. In one study birds did seem to react
differently to recordings of their own songs than to those of neighbours or
strangers. Though, as Hauser points out, a recording of your voice sounds quite
different from what you hear when you are speaking.

Gallup admits that it should be possible to design olfactory or auditory
tests that are analogous to the mirror test. But he still believes his original
test is valid in most animals. “If I close my eyes, cover my ears, or hold my
nose, my sense of self doesn’t fade or disappear,” he says. “My concept of self
is not tied to the visual, auditory or olfactory modality.”

Others, including Povinelli, are more sceptical about what mirrors reveal.
Unlike the wicked queen in Snow White and the Seven Dwarfs, they are
not prepared to take a mirror too seriously.