When Photographers Are Neuroscientists

In the summer of 1867 American photographer Carleton Watkins hauled a mammoth wooden camera through the wilderness of Oregon, taking pictures of the mountains. To prepare each negative, he poured noxious chemicals onto a glass plate the size of a windowpane and exposed it while still wet, developing it on the spot. Even then, his work was not complete. Because wet-plate emulsions are disproportionately sensitive to blue light, his skies were overexposed, utterly devoid of clouds. Back in his San Francisco studio, Watkins manipulated his photos to resemble the landscapes he’d witnessed. His finished prints were composites, embellished with a separate set of cloud-filled negatives.

Nearly a century and a half later, Elena Dorfman—another American photographer porting a large-format camera—spent several summers in the rock quarries of Kentucky and Indiana, landscapes as dramatic as Watkins’s Oregon. Dorfman had none of the old limitations. Her digital Hasselblad instantaneously captured 32-megapixel photos in full color. But it didn’t satisfy her. In postproduction she created composites on her computer, layering as many as 300 images to obtain effects unlike anything seen in nature.

In one sense, Dorfman was doing the opposite of what Watkins achieved with his library of clouds. While his intervention made the mountain vistas more meteorologically accurate, she intentionally introduced physical impossibilities, from conflicting perspectives to rearranged geology, much as a painter might fictionalize a scene. Yet in another sense, each photographer was artfully striving for truth about how we experience the natural world, laboring against the inadequacies of the camera. For Watkins, the constraints were physical. For Dorfman, they’re neurological, a discrepancy between photomechanical depiction and how the brain processes what the eyes perceive. Like many modern artists, she has a strong intuitive grasp of how the visual brain works and understands that her pictures can be made more stimulating by seeding them with conflicting perspectives.

Examining Dorfman’s trickery—and the artifice of other photographers from Weegee to David Hockney—provides a valuable way of learning about how the brain handles contradictory or uncertain visual information. Equally important, the neuroscience of vision helps elucidate what makes these photographs so artistically compelling. Enlisting photomanipulation, these photographers achieve one of the most oft-praised qualities in art—ambiguity—in ways that neuroscience is now beginning to understand.

Dorfman intentionally introduced physical impossibilities, much as a painter might fictionalize a scene.

In a 2004 paper titled Consciousness and Cognition, University College London neurobiologist Semir Zeki argued that the brain experiences ambiguity as the “certainty of many, equally plausible interpretations,” much like an optical illusion alternately seen as approaching and receding. Zeki holds Johannes Vermeer’s paintings as supreme examples of artistic ambiguity, citing Girl with the Pearl Earring to support his aesthetic judgment. “She is at once inviting yet distant, erotically charged but chaste, resentful and yet pleased,” he writes. And, as Zeki notes in his 1999 book, Inner Vision, Vermeer’s art is widely admired because the artist is able to express all of that “in a single profound painting, profound because it is so faithfully representative of so much.”

Too bad Vermeer didn’t have a camera. Photography may be more suited to capturing multiple conflicting certainties than painting. A painting can only blend visual information that has been preprocessed by the unreliable brain. Each photograph, on the other hand, is an independent fact. In layering them without resolving their contradictions, as Dorfman does, photographers approach the tenuous internal quality of truth.

On the surface, photographic ambiguity is an oxymoron. The camera sees the world with total fidelity to physical laws. It’s a deterministic system, the performance of which can be perfectly predicted, given the optics and the chemistry or necessary electronics. Human perception is far messier, hacked together by a visual system riddled with inconsistencies, as Zeki documents in Inner Vision. Perception is a product of evolution rather than engineering. Nobody thought out in advance how all the parts would work together. It isn’t really a system so much as a Rube Goldberg contraption where the parts just manage to interact. For instance, every eye has a blind spot where the optic nerve passes through the retina, carrying signals to the brain, yet this blind spot isn’t noticeable to the viewer. As National Institute for Physiological Sciences neurophysiologist Hidehiko Komatsu details in a 2006 Nature Reviews Neuroscience review, the brain fills in the blank in what is perceived by interpolating visual information from areas of the retina surrounding the hole: The resulting sight is an approximation.

In layering images without resolving their contradictions, photographers approach the tenuous internal quality of truth.

In fact, most everything about how we see comes down to heuristics. In a 2005 Nature paper, Harvard neuroscientist Patrick Cavanagh observed that “our visual brain uses a simpler, reduced physics to understand the world.” In other words, we aren’t using all the complex mathematics of textbook optics to analyze the relative position and characteristics of objects. We’re taking a few imperfect cues and improvising. The visual information isn’t totally accurate, but it’s good enough to get by and can be more useful than perfect knowledge. (You don’t need to calculate the velocity of a falling rock to know that you should step aside. In fact, if you start making measurements from where you’re standing, you probably won’t survive to complete them.)

Painters have an intuitive grasp of these heuristics, taking visual shortcuts such as showing more of an object than can be seen in a single glance or muddying insignificant background details. Those distortions and uncertainties make their pictures less true to life but more true to us: They reflect how we see. On that basis, Cavanagh claimed that “artists act as research neuroscientists,” informing us about how the brain works.

That may be inevitable for painters, since their paintings derive from the perceptual apparatus of their own eyes and brains. The heuristics of the mind guide the naive hand. The reduced physics can be circumvented only with intense effort. For instance, achieving unflawed one-point perspective, in which all objects recede to a single vanishing point (like a converging set of railroad tracks) requires rigorous training in overcoming the eyes’ constant movement and refocusing. A geometrical system must be superimposed on the canvas, and even Renaissance masters were notoriously hazy about the correct projection of shadows. Photographers have the opposite problem. One-point perspective comes in the box. The camera’s view is too absolute. To be neuroscientific—and to make pictures that resonate in painterly ways—photographers must actively emulate the distortions and uncertainties inherent in vision by purposely making their pictures deformed or ambiguous.

Distortions and uncertainties make their pictures less true to life but more true to us: They reflect how we see.

In 2006 MIT neuroscientist Pawan Sinha and colleagues showed that faces with exaggerated features are more immediately recognizable than straight depictions—despite being less accurate—because they deviate more from average human appearance. Paradoxically, you increase the intelligibility of the image by decreasing the verisimilitude. Pablo Picasso instinctively exploited this phenomenon by selectively distorting facial features, intensifying the presence of portrait sitters such as Gertrude Stein. In the mid-20th century, the American photographer Weegee applied a similar tactic in his caricatures of public figures, including Lyndon Johnson and Marilyn Monroe, whose features he twisted by physically warping his negatives and using trick lenses. Johnson’s sloping nose was obscenely elongated; Marilyn’s luscious lips were puffed up to become positively lascivious.

Ambiguity has also taken cues from modern painting. David Hockney, a painter by training, has revisited Cubism with a camera in photo collages constructed from dozens of snapshots. The snapshots are all close-ups of the same subject—say, a city street or a swimming pool—each taken from a different angle and all overlapped to render the contours of the original scene inexactly. As in a Cubist painting, no single point of view can account for all the visual information. Yet the ambiguity about where you’re standing and the uncertain positioning of the scenery don’t make the image confusing: They make it all-encompassing.

Visual research by Harvard neurobiologist Margaret Livingstone and colleagues helps explain why Hockney’s method works. When we encounter an unfamiliar setting, our eyes dart around, acquiring snapshots of everything in sight before we begin to analyze what we’re seeing. This tactic maximizes our acquisition of information before we know what we want to know and prevents us from prematurely filtering out details that may later prove important in terms of how we handle a crisis. But as a consequence, each snapshot is independent. That means that, while we’re looking, we aren’t aware of any inconsistencies between images. Evolutionarily, noticing discontinuity simply wasn’t important. As Livingstone and Wellesley College neuroscientist Bevil Conway wrote in a 2007 review in Current Opinion in Neurobiology, “Perspective and reflections change second to second as we move our eyes across a scene. Therefore, there would have been little biological benefit to incorporating the rules for global perspective or illumination into our visual computations.” Hockney’s photo-cubism embodies and exploits our obliviousness to show us more than can possibly be captured with the camera’s one-point perspective. We have the sense of being right there with him.

When we encounter an unfamiliar setting, our eyes dart around, acquiring snapshots of everything in sight.

Dorfman’s layered photographs work on the brain in a similar way, though they have less in common with Cubism than with the post-Impressionist paintings of Paul Cézanne. Instead of kaleidoscopically fragmenting the picture plane, Cézanne fused the visual information gleaned from multiple disparate perspectives. In his still lifes, for instance, a bowl of fruit might be shown at an impossible tilt that allows you to see all the apples and the bowl’s contact with the table. The reduced physics lets the artist accentuate the scene’s essentials, even if his place of observation is ambiguous. It’s a naive point of view, the naturalness of which is proven by how commonly it’s found in drawings by children.

In her most ambitious pictures, Dorfman takes this power of synthesis to an extreme by transparently layering scenes viewed from completely different directions or up close and at a distance. In a single two-dimensional image, Dorfman presents the visual information we’d glean by exploring a three-dimensional landscape over many hours or days. Looking at one of her pictures, we feel as if we’ve actually experienced the place.

And in a sense, we have. Dorfman’s photographs give us access to her experiences and her ambivalence toward a place that’s at once beautiful and ruined. Through her photographs, we feel her ambivalence. We don’t only observe the multiple incompatible certainties that Zeki describes—we internalize them.

Critic and artist Jonathon Keats is most recently the author of Forged: Why Fakes Are the Great Art of Our Age (Oxford University Press).

This article was originally published in our “Uncertainty” issue in June, 2013.