Chapter Five

Sight Itself

Visual rhetoric has typically looked at things seen, not the act of seeing. But rhetoricians must follow the path of light and neural pulses through bodies if we are actually going to understand the rhetorical functions of the visual.

A macro photograph of a human eye, showing brown, yellow, and blue tones making up the iris.

Figure 9. Macro Photography of human eye. January 4, 2017. Photograph. Unsplash. @V2osk.

And now for a magic trick: What do you see in the figures below?

You’ve likely seen these images, or other ambiguous figures, before, which means you know this is a trick question. Depending on how you look at the picture, it might look like a rabbit, facing to the right with ears extending left. Or, you might see a duck, facing left with a slightly open beak. Once you know that it might resemble either, you’re more likely to be able to switch back and forth between views.1

Alt Text: Three black and white visual illusions.

Figure 54. Three ambiguous figures, right to left.
1) W. E. Hill. My Wife and My Mother-in-Law, 1915. Illustration. Library of Congress.
2) Rabbit and Duck, 1892. Illustration. Wikimedia Commons.
3) Charles Allan Gilbert. All is Vanity, 1892. Illustration. The Illusions Index.

If you’ve seen any of these ambiguous figures before, you’ve likely encountered them as a game. The ways that sight can “play tricks” on us are, indeed, often fun. However, they also have a lot to tell us about why sight is itself a rhetorical function. The variations in how these images are perceived shows that sight works within the complex of physiological functioning, neurological structures, and sociocultural training. This complex of influences is much more persistent and subtly influential than the direct instruction to notice both a rabbit or a duck. Thanks to those factors, for example, if it’s Easter, you’re more likely to see the rabbit, but if it’s October, you might see the duck first.2

Sometimes, the results of sight itself as a rhetorical function are fun and innocuous; sometimes, they are deadly. And therefore this chapter will be a jarring one as we explain the rhetoricity of sight itself from within that complexity—including the dangerous slip from rabbit-duck to a split-second decision about whether someone stopped for speeding is reaching into their pocket for a wallet or a gun. The rhetorical function of sight itself is at work across the whole spectrum of ways that viewers can be fooled by their own sense of vision.

Over time, such errors and discrepancies in vision have been the subject of research in physics, physiology, and eventually, psychology, where the gaps between sight and reality provided rich ground for establishing a growing and distinct discipline. As Nicholas Wade explains, perception was thought to have no physiological explanation in the nineteenth century, and the problem of optical illusions—an emergent form from popular print culture, often for the purposes of humor or games—was taken up more seriously as evidence of the need for the new discipline of psychology.3 Ultimately, psychological research also failed to fully explain visual error—or the broader, as yet understood, mechanisms of sight. How humans see, and why seeing is so manipulable, remain open questions. From the biological, physical, and social sciences to the arts and humanities.

For all that sensation is presently de rigueur in the humanities, it can be easy to forget that vision is a sense. Perhaps because the “pictorial turn” preceded the “affective turn” or because the legacy of Western modernity has turned vision into a matter of disembodied cognition, recent humanistic work on sensation, embodiment, and feeling has tended to prioritize senses other than vision. Though there are important recent exceptions, within rhetorical studies, “visual rhetoric” and “sensory rhetorics” have largely gone in separate directions.4 But, we argue, understanding the rhetorical function of the visual requires attending to its bodily, sensory nature—no matter how clear or accurate that vision is. Illusion, as it opens vision to ambiguity through somatic means, paradoxically allows us to bring sight itself before your eyes. As Bejamin Firgens writes, to notice visual illusions "is to notice the illusory foundations of our sensoria that we usually ignore in favor of our preferred sureties about the world."5 Or, in the terms of this book, visual illusions make clear that—via embodiment—sight itself is a rhetorical function.

Figure 55. 3D Model of an eye. 2017. Image. Sketchfab. Tom Hodes.

Habitually (and ironically), visual analysis and criticism tend to presume a thinking, largely disembodied viewer. But recent visual scholars remind us that viewers are feelers as much or more than thinkers. “We do not see with our eyes alone,” Janet Vertesi notes, “Learning to see requires both bodily skills and instrumental techniques.”6 Tina Campt, likewise, treats seeing as a multisensory act that extends well beyond interpretation of stimuli. Campt “challenge[s]... the equation of vision with knowledge”7 and proposes “listening to images” as a means for gaining “access to the affective registers through which these images enunciate alternate accounts of their subjects.”8 Sight is a matter of sensory input that, in the process, implicates and incorporates full bodies—including the neural pathways of the brain.

If rhetoricians don’t grapple with sight’s embodied nature, we will miss a great deal of what makes the visual rhetorical. The visual is fully embedded in physical, cognitive, social, and cultural matrices. What lookers see, how they see, and what they do with sight all informs the rhetorical function of sight itself. And the many matrices at work when viewers create certainty out of ambiguity are ideal for illuminating the details of that embodied rhetorical function. Sight is about bodies as complex, situated, and visceral. It is also about bodies as habituated and trained. Sight itself is riddled with ambiguity, with bias, and with rhetoricity.

In this chapter

To understand the rhetorical functions of sight itself, we begin with a brief and much-too-simple discussion of what sight actually is (and isn’t), complicated throughout by ambiguous figures and optical illusions. Then, working through psychological and sociological research on racialized perception and implicit bias, we confront the question of whether viewers actually see what they’re looking at. Finally, we consider what viewers do with their seeing. Taken together, these discussions make clear that rhetoricians need to approach sight as physical, visceral, and sensory if we have any hope of fully grasping all the other rhetorical functions of the visual.

What happens when a person sees something?


People who study rhetoric often delight in discovering rhetoric in unexpected places. We are no different. And so, Christa was overjoyed to open books about visual perception and find them peppered with commentary that could as easily have been written by a rhetorician. “Seeing,” E. T. Jaynes writes, “is not a direct apprehension of reality…Quite the contrary: seeing is inference from incomplete information.”9 When processing the reams of evidence taken in through an eye, Stone explains, “the brain, drawing on some 3.8 billion years of inherited evolutionary experience, does what it does best: it makes an intelligent guess.”10 Sight itself, in other words, is a function of the probable, not the actual. Just like rhetoric.

Figure 56. Diagram of an eye. December 1, 2017. Graphic. Wikimedia Commons.

Stone tells us that even though scientists know a great deal about “optics, the eye, and the brain,” sight itself is still a mystery: “the truth is that nobody knows exactly how we see.”11 But, even with incomplete knowledge, what is known about sight is evocative for visual rhetoric and worth learning for anyone interested in studying the rhetorical functions of the visual. Our overview of what happens, physically, to produce sight is rudimentary, but it highlights some key elements and the points where uncertainty and probability come into the mix most prominently.

We begin by noting that sight is not a universal human experience. Not everyone perceives the world through their eyes. Not everyone who sees sees in the same way or sees the same thing. This brief tour of the science of sight then, is a composite of useful information about sight but not a full accounting of how vision works, and certainly not an account of how “we” see. Such a simple and complete accounting is not possible.

Eyes appear to have evolved independently in multiple phyla starting some 543 million years ago.12 Different species have eyes that are more or less elaborate, and different evolutionary needs have led to different features across species. But eyes have been around for a long time and, once they appeared, they developed quickly as means of competition and survival. A simple eye, producing images through a single unit (one eye rather than compound eyes), “can form an image on the back of the eye or retina with or without a lens.”13 The human eye, while “simple” because contained in one unit, is a complex ball of gel with cornea, pupil, lens, and iris out front and retina, optic nerve, and macula in the back. Photoreceptors on the retina process light entering through the cornea and lens. Rods help supply low light vision and cones register color—though the nature of the lens means not all colors can be in-focus at the same time.14 Because of its physical characteristics and evolution, the eye can be “tricked” about color, shape, size, and focus. For example, the image below is a variation on the Munker illusion, which is an illusion of assimilation. What is perceived is shaped by the information around it, in this case, color. In this image, the spheres behind the horizontal bars are the same color: a bright yellow. Yet, they appear to be peach and lime green because they are interpreted alongside the colors of the horizontal stripes.

Figure 57. A color assimilation Munker illusion. 2021. Image. David Novick and Akiyoshi Kitaoka. "The Confetti Illusion." Journal of Illusions.

Sight—in all eye-using creatures—makes the world from an approximation of what photoreceptors receive. “At first,” Stone writes, “this seems to imply that the eye is not telling the brain everything it knows.”15 “However,” he continues, “what the eye does not tell the brain, the brain does not need to know.”16 Images taken in through the eyes contain a great deal of repetition. Additionally, the human eye itself is far more sensitive to light than most digital cameras—the retina has 126 million photoreceptors. However, the optic nerve, which must carry that information, has only 1 million fibers. The information from the photoreceptors, then, gets “squeezed” through the optic nerve’s limited cables.17 Sight requires noticing, as well as ignoring, some aspects (more on that in chapter 6).

In other words, human sight—even in its most extraordinary and effective forms—doesn’t produce a direct copy of the world before the eyes. “Each time we open our eyes we are confronted with an overwhelming amount of information,” Marisa Carrasco explains.18 “Despite this, we experience a seemingly effortless understanding of our visual world. This requires separating the wheat from the chaff, selecting relevant information out of irrelevant noise.”19 This ability to reduce details is extraordinarily useful: eliminating redundancy allows the eye to transmit large, complicated images to the brain along the finite resource of the optic nerve.20 It also, however, means that sight is a guessing function not a direct apprehension of the world even though viewers often experiencec what is seen as a measure of reality. This guessing function, and how it requires the elimination of what is deemed unnecessary information, can be made recognizable through attention to illusion. As Firgens writes, illusions help us "notice the usually transparent meaninglessness of all meaning-making processes."21 This point is well illustrated in the class of illusions based on "motion-induced blindness," in which a moving shape or pattern stimulates the suppression of other stationary objects. For anyone staring at the red "X" in the center of the video below, the moving blue dots will remain in focus as the yellow stationary ones disappear after a few seconds. When the moving blue dots realign, the yellow dots will appear visible again. Like most (if not all) illustions, the mechanisms that trigger this phenomenon are multiple or unknown, but our work to make sense of the world through incomplete information is constant.

Two circles overlaid with horizontal stripes. The circles appear different colors though they are the same.

Figure 58. Motion-induced blindness illusion. 2023. GIF. Masataka Suzuki, Rikako Noda, and Kazufumi Takahasi. Journal of Illusion.

Equally importantly, human brains have stored information about the world that they can draw on to complete pictures, opening all sorts of opportunities for ambiguity, illusion, and rhetoric. In order to see the world, Stone explains, “the brain has to interpret the retinal image, which is full of lies and omissions. The brain’s task is to see beyond these, to ignore the lies and fill in the omissions.”22 This interpretation is, of course, quite imperfect. It requires the brain to make judgments and determine probabilities based on previously acquired information. “Fortunately,” Stone continues, “the lies and omissions don’t change from day to day but are law-like and systematic, and this systematicity allows the brain to correctly interpret all but the most ambiguous retinal images.”23 Yet, even those laws and systems can lead to illusion-driven misinformation.

Sight is a multi-directional process, though our overview so far might seem to suggest otherwise. Sight is not (merely) the result of light entering the eye and projecting an image onto a static screen. Instead, sight is a constant, ongoing, and dynamic process. Even the physical structures of the eyes are not static. They are affected by and responsive to what they see, as when the pupil widens in a dark room to allow in more light. Yet, even this familiar process is not as simple as an environmental stimulus followed by an embodied response. A recent study in Frontiers in Neuroscience shared a new illusion, the “expanding hole,” that shows that the pupillary light reflex “depends on perception, not necessarily reality.”24 In the still image below, the blurry black circle appears to be expanding. This is because the pupil reacts by widening in anticipation to a change in light, “even if this ‘light’ is imaginary like in the illusion”—what and how we see changes because of what we expect to see.

Image of a blurry black circle centered on a white background covered in smaller black circles appears to be expanding.

Figure 59. "Expanding hole" optical illusion. May 29, 2022. Image. Bruno Laeng, Shaoib Nabil, Akiyoshi Kitaoka. Fronteirs in Neuroscience.

Despite Stone’s suggestion that the eye and brain mostly interpret the world “correctly” except in ambiguous cases, researchers have increasingly found that ambiguity is a major factor in visual engagement with the world around us. And that ambiguity extends beyond the fun of illusions. Scholars working on race and visual perception, especially, tell us that all is not right with the lies, omissions, best guesses, and probabilities that permeate visual processes. As Eberhardt et al. explain, visual ambiguity "get[s] resolved in ways most functional and meaningful to perceivers.”25 The “function” and “meaning” used to make sense of ambiguity are culturally situated, and they are informed by politics, gender, and race. The “law-like and systematic” omissions that human brains use to interpret signals from the eyes? They include white supremacy and its corollaries.

Do viewers actually see the thing they're looking at?


The picture below is the opening screen of one of Harvard’s “Project Implicit” Implicit Social Attitudes tests. This specific test was designed to assess participants’ underlying racial biases. Under normal testing conditions, if you are like the vast majority of the US population, you would likely have an easier time associating faces that you perceive as white with positive words and faces you perceive as Black with negative words.26 Whatever biases you have, you likely won’t notice their effect while looking at the pictures captured here. These illustrations don’t replicate the actual test conditions, after all. However, research suggests that neural pathways, developed over time and habituated in a white supremacist society, tend to replicate racism as data from photoreceptors in the eyes travel into and through the brain.

Screenshot of instructions to identify children’s faces as either African American or European American and words as pleasant or unpleasant.

Figure 60. The opening screen of Project Implicit's Implicit Social Attitudes Test. 2025. Screenshot. Mahzarin R. Banaji and Anthony G. Greenwald. Project Implicit.

Sighted children as young as three months old visually register racial markers. What they see affects their friendship choices and beliefs about people well before they enter school.27 Teaching evaluations, notoriously, are better at capturing students’ perceptions of their instructor’s race and gender than the instructor’s pedagogical performance.28 Both white college students and white police officers unconsciously look to and for Black faces when they have been primed to think about crime.29 These tendencies, typically termed “implicit bias,” are persistent and deeply seated, particularly among white people. They continue to appear even among white people who have worked to address their own racism, sexism, and other prejudices. Bias is embodied, and it is powerfully persistent in how viewers deal with the inevitable ambiguities of sight.

Of course, implicit bias isn’t only a matter of vision. Even so, tests for implicit bias, research on it, descriptions of it, and proposals for how to curb it invariably rely on sight as the primary trigger for implicit bias (even as scholars across multiple fields who study blindness and racial perception have complicated and contextualized the presumption that race is a visual feature).30 Tests purporting to identify implicit bias, like the one that we’ve represented here, typically ask test subjects to associate pictures of people—phenotypically racialized, gendered, disabled, or otherwise bodily marked—with positive or negative words. Explanations of how implicit bias happens also tend to emphasize visual sense: seeing a female- or racially coded name and treating a CV more critically;31 viewing a picture of a white person or a person of Asian descent while listening to a lecture and making different judgements on intelligibility;32 watching a Black person reach into their pocket and registering risk rather than nonchalance. Likewise, many solutions aimed at limiting the effects of implicit bias—whether they work or not—tend to rely on controlling vision: performing arts auditions are done “blind;” training brings implicit bias “before the eyes” of search committees; data is gathered, visualized, and published to ensure accountability.33

The features that make implicit bias so powerful and so difficult to change are also built into the bodily mechanisms and processes of sight. This fact does not mean that implicit bias is inevitable or natural. It does, however, point our attention to the embodied ambiguities that are foundational sight itself and to the rhetoricity of what viewers perceive. When viewers look at the world around them, what they see emerges from an amalgamation of physiological, psychological, and sociocultural factors. Jordynn Jack makes this point forcefully with regard to scientific vision, tracking the “pedagogy of sight” through which early microscope users learned to identify and interpret what they saw through the lens.34 But pedagogies of sight aren’t merely implicated in external visual technologies. Human vision as a whole is learned, and the world that is seen is seen through an enormous amount of physical, cognitive, and social processing.

Project 61. Project Implicit Race Test asks users to classify words and images as either good/bad or African-American/European-American. 2025. Mahzarin R. Banaji and Anthony G. Greenwald. Project Implicit.

Such processing begins early in life and sediments over time. “The brain develops in infancy certain assumptions about the world based on all the sensory information it receives,” Jaynes and Bretthorst write. “For example, nearer objects appear larger, have greater parallax, and occlude distant objects in the same line of sight; a straight line appears straight from whatever direction it is viewed, etc.…We hold tenaciously onto [these rules] because they have been successful in correlating many different experiences.”35 But we also sometimes hold onto those rules more tenaciously than we should. What is seen is not necessarily what is. Instead, it is the best approximation of what is based on a confluence of sensory input and previous experience. Frequently, the physical impressions left on the retina by the light reflecting off of objects could correlate to a wide variety of shapes, but experience tells us that some shapes are more probable than others.36 When forming the image seen, the brain selects what is most likely and presents it as what is. As Stone puts it, “experience is required to disambiguate images.”37 But experience is racialized, gendered, socialized, politicized, and filled with implicit and explicit bias.

Indeed, Stone continues, “the brain is forced to interpret an image that is fundamentally ambiguous by adding its own bias.”38 This recognition has spawned an almost alarming number of articles about visual perception that begin with the phrase “Believing is Seeing.” In an interview, Anthony Greenwald—psychologist and expert on implicit bias—explains just this point with regard to racial profiling. He notes, “When a black person does something that is open to alternative interpretations, like reaching into a pocket or a car’s glove compartment, many people—not just police officers—may think first that it’s possibly dangerous. But that wouldn’t happen in viewing a white person do exactly the same action.”39 Everyday movements—simple and straightforward, like reaching into a pocket—are socioculturally and physio-cognitively ambiguous because of white supremacy. Seeing those movements requires the body and brain to draw on previous experience and activate biases. And so viewers—especially white viewers in the United States—see in the very same action a Black man reaching for a weapon and a white man pulling out his wallet. Likewise, in their research on visual associations between criminality and Blackness, Eberhard, Goff, Purdie, and Davies found that study participants were statistically quicker to recognize a weapon when primed with a Black face than with a white one and that biased connections between Blackness and criminality were bidirectional: when white study participants viewed images associated with criminality, they tended to notice Blackness and when primed with Black faces, they were more attentive to images of objects associated with crime.40 “Bidirectional associations operate as visual tuning devices by determining the perceptual relevance of stimuli in the physical environment. That is, given the processing capacity limitations that all perceivers face, these associations determine which information is important and worthy of attention and which is not. So, for example, the association of Blacks with crime renders crime objects relevant in the context of Black faces and Black faces relevant in the context of crime.”41 “Experience is required to disambiguate images,” Stone tells us, but the experience used to make vision work is as biased as sight itself.42

Figure 62. Three screenshots of later test blocks, which ask users to classify images and words into pairings. In the first screen, the picture needs to be identified as either "Bad or African American" or "Good or European American." The next two images reverse these associations, presenting words and faces that should be identified as either "Bad or European American" or "Good or African American." 2025. Screenshot Series. Mahzarin R. Banaji and Anthony G. Greenwald. Project Implicit.

If sight is never about perceiving exactly what is before the eyes but, rather, a matter of physical presumption, reduction, and interpretation, then the rhetorical functions of the visual extend all the way back to the cornea and lens. Implicit bias is not evidence of vision’s failure; instead, it is an indication of its endemic rhetoricity. Long-standing presumptions that viewers would see what was real if they simply looked hard enough or in the right ways or with enough anti-bias training are all eviscerated by the physiology, psychology, and sociology of sight itself. “Social imagination is not only affected by physical difference but it is active in creating and maintaining the perception of those differences,” Eberhardt et. al. explain.43 This recognition does not excuse or justify bias—implicit or otherwise. Instead, it asks scholars interested in visual rhetoric to treat all vision as infused with bias, all sight as always already rhetorical, and all rhetorical vision as situated, biased, embodied, and racialized.

The obligation for rhetoricians once we recognize sight itself as a rhetorical function of the visual is to attend to the patterns that predetermine what is seen. Stone writes, “Your knowledge alters how you see things. The things you see become part of your knowledge. Your knowledge alters how you see things. You can never see the things themselves.”44 Jaynes adds, “We will not relinquish successful hypotheses as long as they work; the only way to make one change these assumptions is to put one in a situation where they don’t work.”45 The biases built into sight itself are important (and important to rhetoricians) “not only because they can lead perceivers to make mistakes occasionally but also because they can guide, generally, how perceivers come to organize and structure the visual stimuli to which they are exposed.”46 Along the way, “stereotypic associations” driven by visual perception “help people to respond to their environment by rendering certain social groups and objects especially relevant and thus worthy of attention.”47 Sight itself, embodied, racialized, biased, is constitutive rhetoric.

Implicit bias, understood through a generative framework of sight itself, reminds us that perception not only cues judgment, it is judgment. Perception and judgment are simultaneous, interdependent bodily matters. “All that you see, all that you have seen, and everything you will ever see,” Stone writes, “is delivered to your brain as a stream of digital pulses whizzing along the fragile threads of salty, fluid-filled cables that are your nerve fibers.”48 Even when we treat vision as a sense that crucially involves the brain, we must also treat its fleshiness—soma, endoneurium, fiber. And what our bodies do in making, gathering, processing, and responding to those fleshy signals has enormous consequence for what we see—or don’t see—around us.

What should we do with sight?


So, what ought rhetoricians do with this information? How ought the recognition that sight itself is a rhetorical function change analytical practice? Here, we suggest three ways that rhetoricians might begin to address the rhetorical functions of sight itself.

1. Eschew presumptions about how or what "we" see.

Such presumptions are not only ableist, they also miss the message that sociocultural contexts matter immensely. The implicit biases of one place and time do not map neatly onto those of another time. Neither does sight. In the United States, white viewers process what they see differently than Black viewers do when looking at race. Results in implicit bias tests, for example, are markedly less biased for Black and Latinx test-takers than white and Asian-American test takers. These differences in sight are not "natural" or "inborn"—they are caused by the particular factors of life in a society profoundly shaped by white supremacy. And they extend to instances of vision that seem so innocuous or neutral as to be imagined as disconnected from sociocultural contexts, but aren’t. To add nuance to this recognition of differential sight, we present our final magic trick: Which line in the image below is longer?

Two horizontal lines of the same length. The top line ends with arrows pointing inward while the bottom has arrows pointing outward.

The Müller-Lyer Illusion. The Illusions Index.

Christa and Brandee both perceive the top line to be longer than the bottom line. We are wrong. Both lines are the same length. Our error was once believed to be universal. Researchers thought that everyone who looked at this image would also see the top line as longer. For a variety of reasons, this illusion has come to be one of the most famous and widely studied, and the reach of those studies carried this illusion beyond the Western world and worldviews. When showing the illusion to viewers across the world, researchers had to acknowledge that what had been presumed to be a universal feature of sight was, in fact, cultural and contingent. Viewers in Western and Westernized cultures learned to see the world in ways that prepared them to make this visual mistake. Viewers in other contexts hadn’t necessarily learned those biases. They sometimes saw the two lines as equivalent, foiling the illusion (more on this in a moment).

In the same way, rhetoricians looking at pictures, describing them, and analyzing them ought to have less confidence about what those pictures show and what viewers will see. They ought to withhold confidence not just because other viewers might see different things (this outcome is actually less likely) but because those other viewers might all see the same thing. That sameness is telling. It is as much an approximation and indication of rhetoric at work as any differences would be. Whether they see the same thing or different things, viewer and analyst both take the part for the whole. They make meaning from partial information and base judgements on probability. And even when that invention doesn’t introduce error or bias, it still grants what is seen a reality and certainty that it doesn’t necessarily merit.

2. Track histories of sight and vision.

Rhetoricians interested in the visual can and should track histories of sight and of vision. Greenwald explains:

"The mechanisms by which our brains form associations and acquire them from the cultural environment evolved over long periods of time, during which people lived in an environment that was consistent. They were not actually likely to acquire something that they would later have to unlearn, because the environment wasn’t going to change. So there may have been no evolutionary pressure for the human brain to develop a method of unlearning the associations."49

If this is the case, then rhetoricians need to know more about the contexts in which habits of sight emerged and need to recognize the patterns that have developed over time. And then, we need to pursue the habits of perception and judgment that can help disrupt such patterns and provide the impetus necessary to force a change in well-established ways of seeing. Both Greenwald, writing about implicit bias, and Jaynes and Bretthorst, writing about perception, note that visual bias will only change when a way of seeing becomes completely untenable for the viewer. Visual rhetoricians, then, have an obligation to research, identify, and pursue the conditions under which racist, sexist, xenophobic, ableist, and other biased visual habits can be made perceptually unacceptable.

Similarly, rhetoricians should pay close attention to how histories of sight theorize and frame how vision works. The Müller-Lyon illusion here, again, proves instructive. When the Müller-Lyon was one of the illusions brought on the Cambridge Anthropological Expedition to the Torres Straits in the late nineteenth century, researchers turned it into an interactive version that used a sliding panel overlaid with another on the lid of a box. Participants were then asked to slide the top panel in until they believed that both lines were equal.

Diagram of the design of the Müller-Lyon box to be used in the Cambridge Expedition experiments.

Figure 64. Diagram of the box and lid used to test the Müller-Lyon illusion. 1901. Illustration. Alfred Haddon et al.Reports of the Cambridge Anthropological Expedition to Torres Straits.

Researchers noted all kinds of variations in these test results, including between children and people who were previously aware of the illusion. The comparison that researchers were most interested in, though, was between Western and non-Western vision. They were especially anxious to explain the finding that Western viewers were more susceptible to the illusion. One early twentieth century analysis explained:

“The fact that the illusion seemed to be distinctly less marked to the Papuan than to the European may possibly be due to the fact that the former concentrated his attention more completely on the special task he was given to perform, viz., to make the lines AB and BC equal to one another, and tended to disregard the other lines present in the figure. The European, on the other hand, probably recognizes at once that he is dealing with more than the simple problem of the relative length of two lines and tends to regard the figure as a whole.”50

The report continues to consider why there was a wider variation in responses from Europeans (who were, thus, more often wrong and more varied in how wrong they were), and it too relies on racist characterizations and associations. The report states, “it is possible that in the simpler mental features [the people of Murray Island] may present more uniformity than is found among the members of a highly civilized community.”51 One piece of this theory of “civilized” vs “uncivilized” characteristics that persisted through at least the 1960s was that Eurpoeans lived in “carpentered” environments, which meant they interacted with more right angles and square corners and were, thus, more susceptible to the illusion.52 There is, in actuality, no settled explanation for the Müller-Lyon illusion, but the explanations chosen over time should raise the attention of rhetoricians, especially for how racist and other biased ideas can underlay and be perpetuated by our theories about how vision works.

3. Track and critique evolving technologies of vision.

While the eye and the brain provide important nuances to sight itself, we rarely see through and with our eyes and brains alone. The various ways that technology, and computation in particular, are enmeshed with visual processes means that rhetoricians must consider how sight itself is always already technological (as well as how the technological is always already rhetorical). As James Dobson explains, computer vision—automated processes by which digital images are recognized, processed, and analyzed—was developed to mimic how human vision was understood to work. Human vision is, in short, “the precondition for machine vision.”53 As vision, imaging, and perception technologies increasingly become integrated into processes of human seeing—and make “unwarranted claims to objectivity,”— it will be important to understand them as implicated in the biases, flaws, and gaps that pervade human vision.54 Just as Safiya Umoja Noble and others have noted the biases built into how algorithms “see” data and perpetuate selection biases based on programming, visual rhetoricians ought to approach sight itself as “disaggregating fleshy bodies into…discrete packets of codified information” upon which yet other brains and bodies build judgment.55 The tools of visual rhetoric must thus extend in new ways not only into study of implicit bias and code but also into medical and scientific imaging, visual pedagogy, and political imagination. It can, likewise, complicate presumptions about the nature and effects of “bringing before the eyes” any particular social issue or problem.

Addressing sight itself as a rhetorical function of the visual will require visual rhetoricians to pursue new training and engage new subjects in new ways. From our viewpoint, that disconcerting, discomfiting work will be well worth it.



Footnotes

  1. Kyle E. Mathewson, “Duck Eats Rabbit: Exactly Which Type of Relational Phrase Can Disambiguate the Perception of Identical Side by Side Ambiguous Figures?” Perception, 47, no. 4 (2018): 466-69. ↩

  2. Peter Brugger and Susanne Brugger, “The Easter Bunny in October: Is it Disguised as a Duck?” Perceptual and Motor Skills, 76, no. 2 (1993): 577-78. ↩

  3. Nicholas J. Wade, “Early History of Illusions,” In The Oxford Compendium of Visual Illusions, ed. by Arthur Shapiro and Dejan Todorovič (Oxford: Oxford University Press, 2017), 3-37. ↩

  4. A notable exception is Debra Hawhee’s “Looking Into Aristotle’s Eyes: Toward a Theory of Rhetorical Vision,” Advances in the History of Rhetoric 14, no. 2 (July 2011): 139–65. . ↩

  5. Benjamin Firgens, "Autokinesis and the Sense of Meaninglessness," In Sensory Rhetorics, ed. Steph Ceraso and Jonathan W. Stone (University Park, PA: Penn State University Press, 2026). ↩

  6. Janet Vertesi, Seeing like a Rover: How Robots, Teams, and Images Craft Knowledge of Mars. Chicago ; London: The University of Chicago Press, 2015: 9. ↩

  7. Tina M. Campt, Listening to Images (Durham: Duke University Press, 2017), 6. ↩

  8. Campt, Listening, 6. ↩

  9. Edwin T. Jaynes, Probability Theory: The Logic of Science, ed. G. Larry Bretthorst (Cambridge: Cambridge University Press, 2003), 133. ↩

  10. Stone, Vision, 2. ↩

  11. Stone, Vision, 1. ↩

  12. Stone, Vision, 17, 21. ↩

  13. Stone, Vision, 21. ↩

  14. Stone, Vision, 24. ↩

  15. Stone, Vision, 30. ↩

  16. Stone, Vision, 31. ↩

  17. Stone, Vision, 185. ↩

  18. Marisa Carrasco, “Visual Attention: The Past 25 Years,” Vision Research 51, no. 3 (July 2011): 1484. ↩

  19. Carrasco, “Visual Attention,” 1484. ↩

  20. Stone, Vision, 32. ↩

  21. Firgens, "Autokinesis," 159. ↩

  22. Stone, Vision, 15. ↩

  23. Stone, Vision, 15. ↩

  24. Bruno Laeng, Shoaib Nabil, and Akiyoshi Kitaoka, “The Eye Pupil Adjusts to Illusorily Expanding Holes,” Frontiers in Neuroscience 16 (2022). ↩

  25. Jennifer L. Eberhardt, Nilanjana Dasgupta, and Tracy L. Banaszynski, “Believing Is Seeing: The Effects of Racial Labels and Implicit Beliefs on Face Perception,” Personality and Social Psychology Bulletin 29, no. 3 (March 2003): 360. ↩

  26. Betsy Mason, “Curbing Implicit Bias: What Works and What Doesn’t,” Knowable Magazine, June 4, 2020. ↩

  27. Jessica Sullivan, Leigh Wilton, and Evan P. Apfelbaum, “Adults Delay Conversations About Race Because They Underestimate Children’s Processing of Race,” Journal of Experimental Psychology: General, 150, no. 2: 395-400. ↩

  28. Anish Bavishi, Juan M. Madera, and Michelle R. Heblt, “The Effect of Professor Ethnicity and Gender on Student Evaluations: Judged Before Met,” Journal of Diversity in Higher Education 3, no 4. (2010): 245–56; Troy Heffernan, “Sexism, Racism, Prejudice, and Bias: A Literature Synthesis of Research Surrounding Student Evaluations of Courses and Teaching,” Assessment & Evaluation in Higher Education 47, no. 1 (2022): 144-54. ↩

  29. Jennifer L. Eberhardt, Phillip Atiba Goff, Valerie J. Purdie, and Paul G. Davies, “Seeing Black: Race, Crime, and Visual Processing,” Journal of Personality and Social Psychology 87, no. 6 (2004): 876–93. ↩

  30. Asia Friedman, "’There are Two People at Work That I'm Fairly Certain are Black': Uncertainty and Deliberative Thinking in Blind Race Attribution," The Sociological Quarterly 57, no. 3 (2016): 437-46; Osagie K. Obasogie, Blinded by Sight: Seeing Race through the Eyes of the Blind, (Redwood City, CA: Stanford University Press, 2013); Osagie K. Obasogie, "Do Blind People See Race? Social, Legal, and Theoretical Considerations," Law & Society Review 44, No. 3-4, 2010: 585-616. ↩

  31. Heather K. Davison and Michael J. Burke, “Sex Discrimination in Simulated Employment Contexts: A Meta-Analytic Investigation,” Journal of Vocational Behavior 56, no. 2 (April 2000): 225–48; Amanda J. Koch, Susan D. D’Mello, and Paul R. Sackett, “A Meta-Analysis of Gender Stereotypes and Bias in Experimental Simulations of Employment Decision Making,” Journal of Applied Psychology 100, no. 1 (January 2015): 128–61. ↩

  32. Donald L. Rubin, “Nonlanguage Factors Affecting Undergraduates’ Judgments of Nonnative English-Speaking Teaching Assistants,” Research in Higher Education 33, no. 4 (August 1992): 511–31. ↩

  33. Not all solutions to implicit bias currently in circulation are as visual in nature. For example, one of the most effective solutions is discretion elimination—allowing doctors, teachers, judges, and police officers (among others) less latitude in their decisions. ↩

  34. Jordynn Jack, “A Pedagogy of Sight: Microscopic Vision in Robert Hooke’s Micrographia,” Quarterly Journal of Speech 95, no. 2 (May 2009): 192–209. ↩

  35. Jaynes, Probability, 133. ↩

  36. Stone, Vision, 5. ↩

  37. Stone, Vision, 6. ↩

  38. Stone, Vision, 7-8. ↩

  39. Mason, “Curbing.” ↩

  40. Like much other research on vision, the exact mechanisms of this function are debated. For example, Payne et. al, 2005 tested two hypotheses for these handgun/tool bias: perceptual illusion and control response, in other words, perceptive vs executive failure. In testing the response time viewers had to these images, Payne et. al. 2005 found that executive failure, or errors that “can occur even when perceptions of the objects are intact,” were perhaps more likely than perceptive failure. Thus, many complex processes, including but exceeding “the eye” work together to produce vision. Keith B. Payne, Yujiro Shimizu, and Larry L. Jacoby, “Mental Control and Visual Illusions: Toward Explaining Race-Biased Weapon Misidentifications,” Journal of Experimental Social Psychology 41, no. 1 (Jan 2005): 36-47. ↩

  41. Jennifer L. Eberhardt et. al., “Seeing Black: Race, Crime, and Visual Processing," Journal of Personality and Social Psychology 87, no. 6 (2004): 877.” ↩

  42. Stone, Vision, 6. ↩

  43. Eberhardt et al., “Believing,” 369. ↩

  44. Stone, Vision, 155. Italics original. ↩

  45. Jaynes, Probability, 133. ↩

  46. Eberhardt et al., “Seeing Black,” 878. ↩

  47. Eberhardt et al., “Seeing Black,” 890. ↩

  48. Stone, Vision, 1. ↩

  49. Mason, “Curbing.” ↩

  50. Alfred C. Haddon et al., Reports of the Cambridge Anthropological Expedition to Torres Straits, vol 2. (Cambridge: Cambridge University Press, 1901), 126-7. ↩

  51. Haddon et. al., Reports, 127. ↩

  52. D. Donaldson and F. Macpherson, "Müller-Lyer Illusion," in F. Macpherson, ed., The Illusions Index. July 2017. ↩

  53. James E. Dobson, The Birth of Computer Vision (Minneapolis: University of Minnesota Press, 2023), 4. ↩

  54. Dobson, The Birth, 7. ↩

  55. Sarah Elwood and Agnieszka Leszczynski, "Feminist Digital Geographies," Gender, Place & Culture 25, no. 5 (2018): 632. ↩