I have known for years that my “good side” is my left side — it began with an advantage and is gaining on the right all the time. In my teens, a mole appeared in the middle of my right cheek. It’s not small enough, pigmented enough, or close enough to the mouth or eye to qualify as a “beauty mark,” providing Barthesian punctum, the poignant accident; it merely mars the profile. I have a snaggletooth, my upper right lateral tooth, which began twisting out of line as soon as my braces were removed. It’s cute, to a point. It sometimes catches on my lip when I smile. One morning just before my 27th birthday, I fainted and fell into a French door, breaking one of the panes with my face. The glass sliced my chin open, from lower lip to jaw, requiring 15 stitches. The scar is on the right. My right side seems to be aging faster, inexplicably — the nasolabial fold is more pronounced — though the left side gets more sun exposure in the car, and when I sleep on my side, I usually turn to the left.

It was only recently, however, that I realized I automatically look at my good side in the mirror. Because vision is stereoscopic — the two “beams” of your gaze converge at a point, rather than remaining parallel — you can’t look someone in both eyes at once; you have to pick one eye at a time. Standing in a hotel bathroom, its mirror huge and immaculately clean, I suddenly noticed that when making “eye contact” with myself, I look at my left side. Reflexively, I gaze into my left eye.

There’s a high probability that your face also has a good side, and that your good side is the left. Studies have shown that photographs of faces manipulated to be left-symmetrical (so both sides are the left side) are routinely judged to be more attractive than right-symmetrical faces, both in terms of reported “pleasantness ratings” as well as according to unconscious responses such as pupil dilation. One theory supposes that this is because the left side of the face shows more emotion: “Our results suggest that posers’ left cheeks tend to exhibit a greater intensity of emotion, which observers find more aesthetically pleasing,” researchers wrote. “Our findings provide support for…the notions of lateralized emotion and right hemispheric dominance with the right side of the brain controlling the left side of the face during emotional expression.”

The photographer Julian Wolkenstein did a series of “Symmetrical Portraits” in 2010, showing each subject’s right-symmetrical and left-symmetrical photos side by side. The portraits are somewhat unsettling; the resulting faces look related but not identical, like sisters but not twins. Also, perfectly symmetrical faces dip into the uncanny valley, appearing alien or clone-like.

Maybe we prefer our mirror image to photographs, especially candid photographs, because the photograph is objective; the reflection is enhanced

Comparing the half-clones, I find the emotion theory plausible. One blond woman is more conventionally attractive on one side — her eyes more catlike and farther apart, squinting slightly in a Tyra Banks–style “smize”; her nose narrower and pointier at the tip; her cupid’s bow more defined — and more androgynous on the other, her overall countenance less taut. The androgynous side looks somehow less intelligent — but then, we routinely gauge more attractive people to be more intelligent; it’s called the halo effect.

In a popular Quora thread, the top answers to the question “Why do I look good in the mirror but bad in photos?” all revolve around the “mere exposure effect,” which states that we prefer things simply because we are more familiar with them. Photos often capture unfamiliar angles, but even taken head-on, like a mug shot, they show us our true face, not the reversed face we see in the mirror. It’s the reflection that’s inaccurate, but to us, the unreversed face looks wrong.

The best test we have for self-consciousness — which may be the same as consciousness — is something known as the mirror test. It’s administered by putting a subject under anesthesia and then marking their face with an odorless paint. When the subject comes to, they are presented with a mirror. If the subject looks in the mirror and touches their face to investigate or remove the mark, we know they are self-conscious — they recognize the image in the mirror as themselves.

The mirror test has been performed on many different species. Chimpanzees pass the mirror test, but monkeys don’t. Results with gorillas have been mixed. Dolphins and magpies also pass the mirror test. Children pass the mirror test, but not until the age of about 18 months, at which point they also begin to demonstrate “self-emotionality,” or self-conscious emotions, such as pride and shame.

I am fascinated by so-called “mirror delusions,” which can occur with brain damage or degenerative diseases like Alzheimer’s. These patients cease, in a sense, to pass the mirror test; they no longer recognize themselves in their reflections, and believe their reflection is someone else, often an enemy. In Stranger in the Mirror, Robert V. Levine describes a woman named Yolanda, who became convinced that her reflection was a woman named Ruth. “I want her to come over but she doesn’t want to come over,” Yolanda said of Ruth. “I’m tired of talking to her through the window.” She was described by her doctors as seeming “fond” of Ruth. Another woman, Donna, developed a much more antagonistic relationship with her reflection; she felt haunted by the “ugly hag” in the mirror, who would mock and mimic her. A third woman, R.D., felt similarly stalked by an “old bag” who “never told me her name”: “When we get home, you know what? We’re gonna find her waiting right around the windows. In the windows where she watches…I can’t stand her.”

Two of these women refer to the mirror as a window: If we don’t see the reflection as ourselves, we may not notice that the glass is reflective. However, these patients typically persist in recognizing others in mirrors; they have not forgotten what mirrored surfaces do, nor do they have generalized prosopagnosia (the inability to recognize faces). It is only themselves they fail to recognize. Brain scans and autopsy reports reveal that patients like these almost always have right-side frontal lobe damage.

Another set of studies has shown that there is more right-brain activity when we encounter images of ourselves (what scientists call “the self-face”) or hear our own voices. According to psychologist Julian Paul Keenan in The Face in the Mirror, we also recognize our own faces faster when our right brains are active (which is tested by constraining the image to one visual field or asking the subject to respond with their left hand). This leads to the conclusion that the self, the sense of self, is not distributed throughout the brain but located in the right hemisphere.

I wonder, since the right brain controls the left side of the face, if we are drawn to left-symmetrical faces because they display — not more emotion — but more selfhood? Do I look at my left side because I locate my self there?

Think of a memorable experience from your childhood. Do you embody your childhood self in the memory, or do you see it as though from across the room, or a bird’s eye view?

There is something fundamentally different about mental images drawn from memory — when you picture a friend or a celebrity, say — and your self-image. The brain seems to build a self-model, a representation of your own body within your mind, so robust that you may glimpse or even confront your own “avatar” in certain fringe states of consciousness — dreams, memories, and fantasies where you watch yourself in the third person — and through the course of “glitches” in conscious experience, such as out-of-body experiences (known in the literature as “OBEs”) or autoscopy, the spooky phenomenon of seeing your own double.

Certainly what we see in general is not entirely the result of vision. The mind builds a model of our environment based on sensory input, which functions well for our purposes but is not identical with the “real world.” Attention and assumptions play a large role in this, which is what causes “change blindness.” In one well-known study, participants asked to focus on a task didn’t notice when a man in a gorilla suit walked through the room. It’s almost as though people are seeing a cached version of the world that, to save processing power, hasn’t been refreshed.

Only when looking at the photo do I see the asymmetry — so I keep my expression neutral, the neutral face I confront over sinks

But according to cognitive scientist and philosopher Thomas Metzinger, our mind not only builds a world model but also a model of the self to exist in that world model — our inner avatar. In The Ego Tunnel, Metzinger posits that we need this avatar to experience selfhood. All experience is mediated, but we don’t experience the mediation as such, in part because we identify so completely with this avatar. What we experience as direct access to the actual physical world through our actual physical body is really just an extremely immersive user interface. Rather than experiencing the world directly, we move through life in a kind of continual virtual reality.

This feeling of self-identification can be extended outside the body. In the famous rubber hand experiment, subjects were induced to identify with a rubber hand on the table when their own hand was out of view. The illusion is apparently so easy to recreate that you can do it as an impromptu party trick. Race car drivers say that they don’t have to think about the dimensions of their car — when driving they feel like the car is their body. The same goes for people who use a cane every day or have artificial limbs. They completely internalize the new dimensions of their “body.”

This provides evidence for our mental ability to identify with an avatar—it’s as though selfhood can float outside the body and latch onto something else. In a typical OBE, the “astral body,” a doubled self, seems to exit the body through the head and hover near the ceiling, so it can view the now “empty” body below. Of course, you have not actually left your body; instead the self-model seems to be replicating. These “replicants” may contain errors. Metzinger mentions an epileptic who, during a seizure, saw himself from the outside, wearing the same clothes but with “curiously” combed hair, whereas he knew his own to be uncombed.

On the other hand, there are phantom limbs. Amputees frequently continue to feel the presence of their missing limb and even feel pain in it, suggesting that the mental self-model can be so persistent and strongly ingrained that changes to the physical body are difficult to incorporate into a new mental model — that the mind is not as plastic as the body. Or perhaps it’s that the sense of self expands more readily than it retracts, that the mind is resistant to reducing the scope of the self. I am reminded of the poet Anne Boyer remarking on Twitter that she did not identify with recent photos of herself because her hair was missing, following treatment for breast cancer. Of course, I thought: phantom hair.

How accurate, then, are our mental models? There’s a pop-psych idea that people in general (women, especially) are plagued by low-level body dysmorphia, believing ourselves less attractive than we really are. But there’s evidence to the contrary. One study found that people were more apt to “recognize” themselves in photos when their images had been enhanced — that is, Photoshopped to appear more attractive. This points to an innate vanity. Maybe we prefer our mirror image to photographs, especially candid photographs, because the reflection more closely aligns with our self-model. The photograph is objective; the reflection is enhanced. Further, change blindness could explain why we suddenly appear old to ourselves on camera. We have grown old in real life but have been blind to those changes in the mirror; our mental model has not adjusted. When the delusional woman stops seeing her reflection as herself, her idealized self, she sees instead a mean old hag to be avoided.

My favorite variety of mirror delusion is known as “negative autoscopy.” When this rare condition occurs, the patient — like a vampire — cannot see their own reflection in the mirror. To me, this has devastating implications, suggesting that what we see when we look in the mirror is not what is reflected at that moment but what we expect to see: our self-model. If we sustain damage to the area of the brain responsible for the self-model, we may be unable to construct a reflection.

And how much damage can the self withstand? The Cotard delusion, also known as “walking corpse syndrome,” causes patients to stop using the first-person pronoun and deny that they exist. This led one sufferer, known as “Mademoiselle X,” to believe she was incapable of dying; she starved herself to death. Physiologically, it is thought to be related to the Capgras delusion, a form of prosopagnosia where those with familiar faces are experienced as impostors. But here it is the self-face that is seen as an impostor. We no longer see or believe external confirmation that our mental self-model is real — a delusion which can lead to a catastrophic loss of self.

The proliferation of self-facing digital cameras makes it easier than ever to produce self-portraiture, to confront non-mirror-images of ourselves. When I tell my friend C that I’m writing about vanity, she tells me that she has taken a photograph of herself every day since she was 27, when she purchased a laptop with a built-in webcam. “I am obsessed with knowing how my face ages,” she says, “and I want to know what my basic expression is.” I have noticed that people who take and post selfies on the internet tend to choose photos from the same angle and showing the same expression. I think we must choose the photos that look most like our self-image; that self-image is then reinforced and reified by the photos.

If I try, with my phone camera, to reproduce the image that I see in the mirror, I fail. The “live” camera shows me a mirror image, but once I take the photo, the image gets flipped, unreversed, my real face. My mirror image looks symmetrical to me, because I look at my left side and unconsciously assume symmetry. Only when looking at the photo do I see the asymmetry — the deeper crease on one side by the nose, the right eyelid drooping slightly. So I usually turn my face to the right, showing more of my left side. I keep my expression neutral, the neutral face I confront over sinks.

Some months ago, my friend A, then working on his dissertation, recorded me speaking about poetry on his expensive new DSLR camera and cut the footage into a short film. Some of the footage was shot outside a bar in daylight, that soft, late-afternoon light referred to dreamily by director-types as “magic hour.” This footage in particular horrified me. Supposed magical properties notwithstanding, the light seemed to magnify every flaw on my person: the facial scar, the sun damage, the second-day hair limp and stringy in the breeze, the tendency to frown and sneer while concentrating. It was not just that I found the angle or lighting unflattering, not quite to my standards — my reaction was vehement. I felt the person in this film was hideously ugly, much uglier than my idea of myself, but more so, uglier than anyone I know. Though I knew it to be irrational, deathlessly vain, I was shaken to the core. I got drunk and cried.

Later, I read that people react strongly to images of themselves that they don’t consciously recognize, often with powerful dislike. In one study cited by Keenan, men and women were shown images of their own faces and body parts (their hands, for example) without being informed that is what they would see and asked to rate their response. Some recognized themselves in the images and some didn’t, but “non-recognition resulted in highly extreme ratings (for example, highly attractive and highly unattractive)…subconscious recognition appears to have a strong influence.” Perhaps my reaction to the video was a combination of recognition and non-recognition. I knew full well the woman speaking was me but could not square the visual with my mental self-image — the serene, impassive, almost expressionless face I find in a mirror or cultivate on camera.

Being called beautiful has made me vain, in the sense of self-consciousness, much as being photographed makes me temporarily over-aware of my body and my interface

My friend T once called me “beautiful by the conventional metrics,” then added that he didn’t mean it as faint praise — clearly intuiting that women, “intellectual” women at least, prefer to think of themselves as “unconventionally attractive.” Until then, I had. But I am white and blond and thin. Had I flattered myself by thinking my looks interesting, even while identifying with my more conventionally attractive images? “Conventionally beautiful and not conventionally beautiful both sound like insults,” I said.

I like to be called beautiful in the moment, of course, but later I half-resent it. I want, somehow, for others to think me beautiful without having to think of myself as beautiful — the myth of beauty without vanity. Being called beautiful has made me vain, not in the Narcissus sense, but in the sense of self-consciousness, much as being photographed makes me temporarily over-aware of my body, my interface — “this mortal coil,” I sometimes think of it. It is not what Shakespeare meant at all, but when I hear the phrase, I imagine my physical form is a skin I can shed like a snake.

I asked C to send me a few of her photos. She complied, writing:

I’m sending you four pictures of myself — two new, two old. The difference is probably obvious. Here’s what I see in them: I am simply more interested in life now than I was then. My lips used to be fuller and have gotten thinner. I think I used to look less like a specific person, more like a generalized girl, now I look like me more exactly.

“I used to look less like a specific person… Now I look more like me exactly.” I can see what she means. In the more recent shots, her gaze connects; it is piercing and intelligent. In the older photos, she seems to be trying to catch herself off-guard. I think of a woman I know who worked in door-to-door sales one summer; she was coached to be found looking off into the distance, as though distracted, when her mark opened the door. This was supposed to make her seem unassuming, less threatening. The goal was to get inside.