I don’t know what it’s like for you, but there are days when it feels I’m like meeting someone for the first time. Her features seem foreign to me, and that, in its way, is not so far from the truth.
I don’t know what it’s like for you: there are days when I am most comfortable if the sight is brief. Best if I have a specific task, like brushing my teeth or plucking at the ragged curve of my eyebrows until one bends to match the other; best if I can file the required report and move on, before too much is seen: Go ahead and wear that shirt. It looks well on you. No, there is no scratch on your cheek. It must have been a momentary twitch of a nerve… Yes, you look as tired as you feel. More tired. There it is. It’s not that I am ashamed, understand; my self-esteem is not a dress that has fallen and must be tugged back up. It’s not that I never stare; oh, stare I do. But there’s something unnerving about it.
I don’t know what it’s like for you. For me it’s a question of manners. Too direct a gaze creates an impossible challenge: which pair of eyes will drop first? I know that both are mine. Yet how strange is what I perceive—that I am at home inside one set of arms and legs, and at the same time these very limbs are hanging quite happily on a separate frame. That I am twinned.
I live quite comfortably with this contradiction, of course; but I suppose I haven’t always.
Babies aren’t born with what psychologists (somewhat ploddingly) call “mirror self-recognition.” It takes many months before they’re able to draw an unfaltering line between their reflections and themselves, to comprehend that the stare that meets their own so fearlessly does not belong to another human being. It’s not just a question of waiting until certain inevitable developments take place in the brain, either—though that is important. A light bulb doesn’t just blaze on one day and transform stranger into self. No; in fact, developing the ability to recognize one’s own body in the mirror seems to be a surprisingly rational undertaking, and one that builds over time.
In 1979—the year I was born, naked of a sense of self— two scientists named Lewis and Brooks-Gunn tumbled a series of burbling 12-month-old babies in front of a mirror, to see what they could see. The vast majority of them, the experimenters observed, engaged in something they called “contingent play”—so named because the movements of a reflection are contingent upon one’s own movements.
Having noticed that there was a being opposite them in the glass, and having perceived that the behavior of this being seemed oddly familiar, the babies would proceed to carry out clever studies of their own. Staring at their reflections, they would perform the same series of movements over and over again, each time watching intently to see if the strange creature in front of them would follow their lead correctly. They bobbed their heads up and down, bounced their chubby bodies enthusiastically, carefully waved their arms back and forth, all the while with eyes growing wide as they began to clarify and confirm the fact that they possessed a perfectly synchronous imitative partner who would do all that they did at just the same moment. These early play sessions seem to be a necessary first step towards claiming one’s reflection as one’s own.
(They are not sufficient, for it is possible to recognize that your movements dictate those of another without recognizing that the two are one and the same. The full understanding that the face opposite you in the mirror is your own does not generally arrive until late in the second year of life, according to subsequent studies. When that understanding comes, it can truly be described as self-consciousness. One common test of mirror self-recognition is to dab a spot of rouge on a child’s nose, then place them in front of a mirror. A sheepish, or frustrated, rubbing at the spot is the positive indicator researchers are looking for.)
But listen; am I the only one who is astounded by the canny, systematic tests those children conducted? Am I the only one who went straight to the mirror to reenact them, nursing a tiny thrill and half-hoping to catch my other self shifting her neck just a heartbeat too late? Because I’ll tell you what the Lewis and Brooks-Gunn study says to me. It says that seeing yourself does not come easily.
Let’s put it this way: to know an apple, say, is straightforward. Hold it in your palm; take in its dangerous crimson; scrutinize its glossy skin. It is entirely self-contained. Its apple-y nature is self-evident. To know your own face in the mirror is different. You have to slide into the apprehension sideways, gather together a body of physical evidence and reason your way towards the truth:
When I nod, she nods. When I stare, she stares back. Her arms follow my arms; her legs stretch as far as mine. This plant does not move when I move; it is not part of me. This other person moves without my say-so; he is not part of me. Only she, with her skin so brown and her feet curling under her like frightened mice—only she moves with me. So. This is who I am. These are the things I am made of. These are my boundaries in space.
I don’t remember collecting those proofs. I don’t remember building my sense of self like this, brick by brick with my baby-brain. But I believe that I did, and you as well. And I’ll tell you something else: I believe that we’re in good company. Elephants, apes, and dolphins can learn to see themselves through contingent play, too.
Also, robots. Robots can learn to see themselves. Are you smiling yet? Listen, at the very least, one robot can that I know of—its name is Nico. I read about Nico in this charming paper, published last year. In it, two Yale computer scientists show how, with the help of three algorithms that deftly compare data to experience, a robot “can learn over time whether an item in its visual ﬁeld is controllable by its motors, and thus a part of itself.”
First, Nico spends some time—four minutes, to be precise—waving its arm back and forth and carefully noting the shape of its own movements. Then, it’s ready to look itself in the eye, so to speak. Nico is placed in front of a mirror, whose contents are captured in a streaming image by a wide-angle lens embedded in what would be Nico’s right eye. Carefully monitoring that video stream, the robot continues to motor its arm around in random directions, checking for precisely contingent movements in the reflected scene. It consults the algorithms in its memory, calculating the probability that what it sees is really Nico. Very quickly, then, the robot is able to accurately determine whether it happens to be looking at itself, an inanimate object, or an animate other.
Once it has understood the form of its own arm, learned the way in which its joints shift position—once it has traced the essential outline of its own metallic body—Nico can be said, in a very real sense, to recognize itself. And after that understanding has set in, no one (not even a sly researcher insinuating himself into the scene and painstakingly mimicking Nico’s movements) can fool it. Nico knows exactly what it is.
But achieving that knowledge demands two things, both of which are clearly spelled out in the title of the Yale paper: time and reasoning.
Seeing yourself doesn’t come naturally; it’s not fundamental to your understanding of the world in general. And it can’t be accomplished simply by having someone else tell you who you are in the glass; it’s not a fact you swallow, but a judgment you come to. At first—ask a baby; ask a robot—it’s not at all silly to narrow your eyes at that odd-looking stranger and wonder why they’re copying what you do. At first, surely it’s right and proper to be suspicious of the shade in the mirror.
When it comes right down to it, I mean, you might be wrong about the whole thing.
Here’s the thing: When I think of myself, what comes to mind is less a single clear and shining image of my own face than a shifting sensation of me-ness: a complex amalgamation of memories, ideas, and sensory impressions. I am the one around whom my husband’s arms wrap, the pressure of his musculature against my own clearly defining the shape of my body. I am the one who lay at the foot of my parents’ bed as a child, listening to the hum and click and drip of their ancient air-conditioner and imagining the sounds growing larger and larger until they merged with my own heartbeat. I am the one who frets for hours before phone calls, sweaty and pale, who dances while she cleans, who hates hair in her face and still remembers the sharp, dusty taste of the whiskey sours she used to drink because she liked the way they made her tongue twist up inside her mouth.
I am the one who feels the way I feel, thinks the way I think, not—or not just—the one who looks the way I look. And how do I look, anyway? No matter how many tests I run, no matter how much I grow to trust the image before me in the glass, the sight of my own face is always mediated through layer after layer of tin, silver, glass, copper, paint. I’ve never seen it without a mirror as middleman, without a bender and broker of light. What if the person I’m seeing isn’t who I think it is at all? Why should their mere resemblance to me be sufficient identification?
I live quite comfortably with this suspicion, of course; but not everyone does.
People with a extremely rare disorder known as Capgras delusion come to believe that those whom they love have been replaced by impostors. These strangers are identical to my mother, my sister, my brother, Capgras sufferers say, but they are not them. They are different people entirely. Their features—remarkably similar! The close resemblance is uncanny! But no; they are certainly not the ones I know. They are frauds. I do not recognize them.
The extent of the delusion is such that, confronted with their own reflections, Capgras patients are apt to startle violently. Why, I’ve never seen this person before! they may exclaim, in horror and disgust. Some engage in the same kind of contingent play that babies do, pinching themselves and waving their arms, keeping a chary eye on the ghoul in the mirror—but unlike babies, they will not be satisfied by the paltry evidence of their own eyes. And when repeated gazes into a mirror call up the same disagreeable stranger again and again, people with Capgras may accuse their likenesses of deliberately appearing in their lives solely to stalk and torment them. Capgras delusion steals a person’s ability to see their own true selves, and replaces it with an uninvited guest who cannot—will not—leave them alone. I confess, I sympathize.
But how exactly does this happen? Capgras patients are otherwise, for all intents and purposes, normal—whatever that means. Their vision is not impaired, and neither is their cognitive functioning; nor are any aspects of their memory. Their negative emotional response to their loved ones and their own reflections is bizarre, to say the least, but in some sense it’s also perfectly lucid and reasonable. It matches, after all, precisely the way you would expect someone to react if everyone in their inner circle of intimates had been replaced by an impostor. (And wouldn’t you yell if your beloved reflection suddenly turned into someone you knew, deeply and profoundly, wasn’t you at all?)
So what causes this extraordinary disconnection between vision and belief, between seeing a person, recognizing their features, and correctly identifying them as someone whom you know and love? The inimitable UCSD behavioral neurologist V.S. Ramachandran has a lovely theory about this. Look, he says: Sensory information about objects the eye sees is transmitted from the retina into visual centers in the temporal lobes. Here, the object is identified: This looks like a teapot, this looks like a poodle, this looks like my sister. Capgras patients can accomplish this part of seeing perfectly well.
But after an object has been identified, the brain continues to work. It sends its decoded information to the limbic system, which is a complex network of brain structures that enables the perception and expression of emotions. One of the first places this visual information passes through is the amygdala (the name means almond-shaped, which it is). Ramachandran explains that the amygdala is responsible for labeling the emotional content of what the eye has seen. This object is beloved, the amygdala concludes, and should trigger affection; this one is despised, and should trigger hate.
This visual data, then, becomes colored with a layer of emotional interpretation; it travels on towards other structures in the brain. At its final stop, what began with a glance at a face sets in motion at last the physiological responses that enable a person to actually experience the appropriate emotion: things like a speedier heart rate, higher blood-pressure, and a light film of sweat covering the skin. (For what is emotion but the brain, talking to the body, talking to the brain?)
You might imagine that a rather odd sensation might occur if this process were disrupted somewhere after the point where an object is decoded and before the point where the emotion that ought to be associated with it is actually experienced. Ramachandran did. He asks:
Is it possible that in this patient there has been a disconnection between the face area of the temporal lobes and the part concerned with the experience of emotion? Perhaps the face area and the amygdala are both intact, but the two areas have been disconnected from each other. When (the patient) looks at his mother, even though he realizes that she resembles his mother, he does not experience the appropriate warmth, and therefore says ‘Well, if this is my mother, why is it I’m not experiencing any emotion? This must be some strange person.’
I think of a person like this, and how they must feel when they stare at the mirror, eyes fixed full upon their own faces and hearts as hard as stone. Do they never, now, experience the comfort of being alone with themselves?
It hasn’t been proven, so far, the hypothesis that the Capgras delusion is caused by the neurological disconnect that Ramachandran describes—but its elements catch at my heart. It’s not enough to simply identify a person in order to truly know them. The brain needs more. It’s not enough to match the movements of a reflection to your own in order to recognize it as yourself. The brain needs more. Seeing yourself does not come easily. It requires time. It requires reason. And, beyond all that, it requires some measure of affection.
Knowing this, I return to my counterpart in the mirror—the one who still seems so strange to me sometimes—and am moved to tenderness. I look on her for a long moment, studying the shape of her lips, the brown of her eyes. I forgive our separation, forget the times when her eyes have challenged mine or mine hers, and gaze.
Because if I do not see her, how will I love her?
And if I do not love her, how can I see her?
If you’re fascinated by reflections, I can do no better than to recommend the deep and intricate treatment of the subject in Mirror, Mirror: A History Of The Human Love Affair With Reflection. I found my copy in the stacks of Powell’s Books in Portland, OR, on my honeymoon.