Controlling (Mental) Images and the Aesthetic Perception of Racialized Bodies

1. Introduction

In her analysis of the Black dancing body, Brenda Dixon Gottschild notes that classical ballet remains the “last bastion of white dance primacy” (2003: 131). It wasn’t until 2015 that Misty Copeland became the first Black woman to be promoted to principal dancer for the American Ballet Theatre. Black female dancers in the big ballet companies in the U.S. remain in the single digits. In the U.K., the three biggest ballet companies employ only three main Black dancers: Francesca Hayward at Royal Ballet, Céline Gittens at Birmingham Royal Ballet, and Precious Adams at English National Ballet. Ballet’s pervasive whiteness cannot simply be explained by a lack of classically trained Black dancers. In the U.S. and the U.K., initiatives like Dance Theatre of Harlem, founded in 1969, and Ballet Black, founded in 2001, work in advocacy, outreach, and professional and early training programmes, so that there is no shortage of Black dancers training in classical ballet. What can explain their continued lack of inclusion in mainstream companies?

Gottschild emphasizes that racialized bodies aren’t inherently fit or unfit for any one kind of dance. Sure, we find variations in bodily features among dancers. But all bodies can be trained to move in a way that accomplishes classical ballet’s aesthetic ideals of geometry of line, lightness, and verticality. In Gottschild’s words, “[i]t’s really more about what we like to see than what the dancing body can be taught to do” (2003: 103). The lack of Black ballerinas at the front of the main ballet companies can partly be explained by racist attitudes affecting the aesthetic appreciation of their bodies.¹ A ballerina is delicate, elegant, vulnerable, graceful, ethereal. These are features that have been historically associated with white women. Black women have been historically regarded as aggressive, hyper-sexual, and even grotesque. Of course, it might be difficult to find dance critics nowadays who would make explicitly racist statements publicly. But it isn’t strange to find critics, artistic directors, choreographers, and dancers who claim that Black dancers simply “don’t have the right lines”, or that their “style” of dancing or “strength” aren’t suited for classical ballet. The reasons offered in excluding Black dancers are aesthetic: the excuse is that their bodies are aesthetically unsuitable for ballet.

The aesthetic evaluation of human bodies is part and parcel of our everyday lives. Not only do we judge Hayward’s elegance as Odile in Swan Lake, but we take notice of a stranger’s beauty while walking down the street, or even a colleague’s scruffiness as they walk into a meeting. Aesthetic evaluations leak into other aspects of our everyday lives, and they partly determine individuals’ access to social and material goods. Unfortunately, these everyday settings reproduce the racialized aesthetic disregard found in classical ballet. For example, Black hair and hairstyles continue to be discriminated in the workplace; and, as in the case of ballet, the excuses offered are aesthetic: Black hair and hairstyles are criticised as inelegant and unsophisticated.²

Aesthetic evaluations routinely disregarding non-white bodies isn’t simply a problem of aesthetic norms designed to favour white bodies. Aesthetic evaluations depend on the aesthetic properties we regard objects as having. And it is widely agreed that aesthetic properties are directly accessed in our experience of aesthetic objects. These racist attitudes are really a problem of aesthetic perception:³ Black ballerinas look inelegant; Black full, curly hair looks messy. Philosophers working in critical race aesthetics, like Alia Al-Saji (2019), Mariana Ortega (2019), Paul C. Taylor (1999; 2016), and George Yancy (2016), argue that this is because there is no such thing as the naked eye: perception of racialized bodies is affected by the social structures in which they are appreciated. The aim of this paper is to argue how social structures can affect aesthetic perception, by deploying tools from analytic aesthetics and philosophy of mind. I argue that mental imagery acquired through the interaction with aesthetic phenomena structures the perception of non-aesthetic properties of bodies (e.g., curly hair), so that aesthetic properties consistent with racist stereotypes (e.g., inelegant hair) are attributed to individuals.

The structure of the paper is as follows. Section 2 examines how the aesthetic appreciation of racialized bodies reproduces the workings of racism. Section 3 characterizes aesthetic properties, and Section 4 offers a proposal for how they might be perceived. Finally, Section 5 applies this proposal to the case of racialized bodies to explain how racist mental imagery impacts the perception of aesthetic properties of racialized bodies.

2. Aesthetic Evaluations and Racial Hierarchy

Taylor argues that aesthetic practices have historically racialized beauty, so that aesthetic evaluations go hand in hand with racial hierarchy. Beauty has been historically defined by the white-European tradition in terms of the physical features white people are considered more likely to have (Taylor 1999: 17–18). Whiteness is thus construed as an aesthetic ideal. Human beings are regarded as more beautiful the more their features resemble the physical features associated with white identity: the fairer the skin, the flatter and silkier the hair, etc., the more beautiful one is. Establishing whiteness as an aesthetic ideal is, of course, part of the project that establishes whiteness as a moral ideal (Yancy 2008; Taylor 2016; 1999). Physical ugliness is taken as a sign of moral ugliness.

The established aesthetic hierarchy is thus problematic because of its implications for whether we recognise others as moral equals. Under the assumption that there is an intimate relation between beauty and goodness, excluding members of a given racial group from positive aesthetic properties carries the implication that their moral value is diminished. The concern for the aesthetic appreciation of racialized bodies is therefore grounded on the implications it carries for equal moral respect. Moreover, as Taylor points out (2016: 115), aesthetic evaluations partly determine individuals’ access to social and political institutions, and ultimately, partly determine their access to social and material goods.⁴

Someone could still object to the idea that people of colour are excluded from positive aesthetic evaluations by arguing that, while they mightn’t be traditionally regarded as beautiful, non-white bodies are aesthetically appreciated in other ways. Many different aesthetic properties are valued in other aesthetic objects. Some artworks are valuable not for being beautiful, but for being fascinating or disturbing. Likewise, someone could argue that non-white bodies are aesthetically appreciated, even if this involves ascribing different aesthetic properties, such as exotic or exuberant. Relatedly, someone could also object that people of colour are commonly celebrated as sexually desirable.⁵

Nevertheless, expressions of sexual desire, and the aesthetic properties it might ascribe, are unlike judgements of beauty in important ways. Taylor notes that there has always been a dialectical relationship between aversion and sexual desire in contexts where beauty is racialized (2016: 116–18). In these contexts, while judgements of beauty are reserved for people whose physical attributes are those traditionally regarded as white features, non-white individuals might still be regarded as having a “special kind” of sex appeal, which might involve the attribution of other aesthetic properties. Nevertheless, this sexual desire isn’t marked by affection, warmth or admiration that we would think should characterise intimate interpersonal relationships, but by violence and forced submission. Taylor directs us to Christina Sharpe’s (2010) account of “monstrous intimacies”: social transactions that are marked by violence and forced submission. This is particularly relevant, for example, when we consider the history of sexual violence toward peoples of colour. This brings us to why it isn’t enough to counter that non-white bodies are aesthetically appreciated and desired even if this involves the attribution of different aesthetic properties. The problem lies in what these different aesthetic properties say about individuals’ moral worth. Some aesthetic properties, such as beauty, elegance or vulnerability, involve moral recognition and motivate moral concern. Other aesthetic properties, like exotic, intimidating, or animalistic, reinforce patterns of violence, dominance, and dehumanization.

One might hope that as we strive for racial equality, beauty norms will soon no longer involve features associated with white identity. But as Taylor argues, the processes of racial categorization are “underwritten by aesthetic perception—by the affectively and symbolically loaded workings of immediate experience. Black people look dangerous, or unreliable, or like bad credit risks” (Taylor 2016: 22). Janell Hobson (2003) argues that racist attitudes have historically led to Black female bodies being perceived as grotesque, ugly, aggressive, and intimidating; and Yancy (2016) notes that Black male bodies have been historically perceived as beastly, violent, and brute. The problem is that, given that we take perceptual experience to be prima facie justified, from perceiving Black bodies as “grotesque”, “aggressive” or “beastly”, the assumption is that that’s just how Black bodies are. The perception of Black bodies as aesthetically deformed serves to justify the core assumption of white supremacy that “the Black body is night, doom, darkness, and danger; it is deceptive and devious; it is a site of vice and moral depravity” (Yancy 2016: 245). The problem lies thus not simply in racialized beauty norms, but in the aesthetic properties racialized bodies are perceived as having. It makes no sense to think beauty norms can stop excluding people of colour, if the problem is that non-white bodies continue to be perceived as having aesthetic properties that simply don’t ground beauty, like being grotesque, lewd, aggressive or exotic.

At this point, it is worth distinguishing between verdictive and substantive aesthetic properties (Zangwill 1998). Verdictive aesthetic properties concern overall aesthetic merits; for example, someone being beautiful or ugly. Substantive aesthetic properties are those on which verdictive aesthetic properties depend; for example, someone being graceful, elegant, delicate, vulgar, clumsy, lewd, etc. Substantive aesthetic properties depend on non-aesthetic properties of objects; for example, a ballerina is elegant because she has long limbs. Verdictive aesthetic properties depend on substantive aesthetic properties; for example, a ballerina is beautiful because she is elegant.

Even if the problem of the aesthetic appreciation of non-white bodies partly consists in racialized beauty norms, the application of these beauty norms involves verdictive aesthetic properties, which, in turn, depend on substantive aesthetic properties. The application of aesthetic norms thus ultimately depends on the perception of the relevant substantive aesthetic properties. Gendered norms of beauty involve, for example, a woman being perceived as delicate, elegant, vulnerable, etc. The problem of the racialization of beauty involves, therefore, not only the application of beauty norms, but the perception of the relevant substantive aesthetic properties (e.g., delicacy, elegance, vulnerability) in racialized bodies.

Perhaps one would be tempted to explain, for example, anti-Black aesthetic evaluations by alluding to bodily features. Since substantive aesthetic properties depend on non-aesthetic properties of objects, the aesthetic evaluation of bodies ultimately depends on their features: given certain physical features, we attribute certain substantive aesthetic properties, on which, in turn, verdictive aesthetic properties depend. Nevertheless, bodily differences aren’t enough to explain racialized aesthetic evaluations. Although aesthetic properties depend on non-aesthetic properties, there are no necessary or sufficient non-aesthetic properties that serve as conditions for ascribing aesthetic properties (Sibley 1959: 424–26). This means that having bodily features traditionally associated with white identity is neither necessary nor sufficient for positive aesthetic evaluations. There are, nevertheless, non-aesthetic features that characteristically count for or against certain aesthetic properties (Sibley 1959: 427–29). But since this is about what is characteristically the case, we still need an explanation for why in the case of racialized bodies, non-aesthetic features like light skin or flat hair characteristically count towards being delicate or elegant, rather than towards being insipid and washed-out. That is, we still need a story for why, when evaluating racialized bodies and given certain non-aesthetic properties, certain aesthetic properties obtain.

To explain the move from non-aesthetic properties of bodies to racialized aesthetic evaluations, critical race philosophers argue that the perception of racialized bodies is affected by the social structures in which they are appreciated. Taylor, as we saw, speaks of the “affectively and symbolically loaded workings of immediate experience” (2016: 22). Likewise, Yancy argues that the aesthetic disregard for non-white bodies is built on a “white gazing” that isn’t “an atomic act or an inaugural event that captures, in an unmediated fashion, the bareness, as it were of ‘objects’”, but “a specific historical practice, socially collective and intersubjective, a process that is dutifully maintained” (2016: 243). According to this picture, the differences in the aesthetic appreciation of white and non-white bodies can be explained by the way social structures shape our ways of seeing, and, thus, by the way the dominant racial project determines how racialized individuals are perceived.

Critical race philosophers argue that these ways of seeing are partly shaped by a social imagery that mediates the so-called immediate experience of racialized bodies. Drawing on María Lugones, for example, Ortega (2019) argues that coloniality constructs particular kinds of beings by means of an “imaginary” that shapes the operations of perception. Al-Saji (2019), on her part, builds on Franz Fanon to argue that normative white ways of seeing are sustained by racial imageries that are partly constituted by images in artistic and cinematic production. Likewise, bell hooks argues that popular and artistic images play a central role in how racialized individuals are seen and engaged with, so that “the real world of image-making is political” (hooks 2015: 5).

We can understand the role of this social imagery in shaping racialized aesthetic perception with the help of the notion of controlling images. Patricia Hill Collins (2002; 2020) argues that controlling images are interlocking stereotypical images that attach to individuals belonging to specific social categories in given social structures. These controlling images normalize social hierarchies in intersecting power relations. Collins pays particular attention to controlling images associated with Black women: “From the mammies, jezebels, and breeder women of slavery to the smiling Aunt Jemimas on pancake mix boxes, ubiquitous Black prostitutes, and ever-present welfare mothers of contemporary popular culture, negative stereotypes applied to African-American women have been fundamental to Black women’s oppression” (Collins 2002: 7). Nevertheless, she argues that all social groups in social organizations have controlling images that apply to them in order to reproduce prevailing power hierarchies (Collins 2020: 79). The controlling images attached to some social categories might be less visible than others due to their seemingly positive content, but they still serve the function of maintaining power relations by upholding certain ideals. This is the case, for example, of controlling images associated with white femininity.

While critical race philosophers offer the crucial insight that aesthetic perception is affected by the social structures in which racialized bodies are appreciated, and that controlling images play a determinant role in shaping ways of seeing, this picture still says nothing about how racist imagery and social structures could be making a difference. In the rest of the paper, I aim to show that analytic aesthetics and philosophy of perception have the resources to complete this picture, and to further bolster the case for thinking that the aesthetic perception of racialized bodies is affected by social structures.

3. Aesthetic Properties

Aesthetic properties have been traditionally thought to be directly experienced. They aren’t inferred even if we can use non-aesthetic properties as explanatory reasons after we have perceived them. For example, we might explain that a painting is dynamic because of the distribution of its figures, brushstrokes or colours. But it isn’t the case that we perceive the brushstrokes, colours, or figures, and then infer that the painting is dynamic. One avenue to explain that aesthetic properties are directly experienced is to claim that they are perceptible: ⁶ they feature in perceptual experience of perceptual objects (e.g., Lamarque 2010; Levinson 2006b; Ransom 2020; Sibley 1959; 1965; Stokes 2018; Walton 1970; 2020). The claim that interests me isn’t that aesthetic properties are perceptual, in that they can only be accessed through perceptual experience, or in that they only depend on low-level perceptual properties. Instead, I examine the claim that they’re perceptible, in that, in the case of perceptual objects, they are directly experienced because they can feature in perceptual experience. This kind of aesthetic perceptualism interests me for two reasons. First, although perceptualism is controversial, very few alternatives have been proposed to explain that aesthetic properties are directly experienced.⁷ Second, this perceptualism fits well with the insights examined in the previous section according to which racialized aesthetic evaluations are so pervasive because of how racialized bodies are perceived. Moreover, it leaves room to accommodate the insight that aesthetic perception is affected by social structures.

Aesthetic perceptualism can take off by arguing that aesthetic properties are gestalt-like or emergent properties (Sibley 1965; Walton 1970; 2020). They arise when low-level properties, like colours, lines, or shapes, are perceived organised in specific ways, but are irreducible to these. Aesthetic properties are directly experienced because they are brought into presence in perception, when low-level properties are perceived under specific configurations. For example, Jerold Levinson (2006b) characterises aesthetic properties as ways of appearing, properties that reveal their natures in and through appearances, and that are perceiver-relative and condition-relative: they only manifest in the right conditions to perceivers with the appropriate sensory-perceptual-cognitive apparatus. The idea is that: 1) objects have certain non-aesthetic properties; 2) in the right conditions, these are perceived under a given structured configuration by a subject with a given sensory-perceptual-cognitive apparatus; 3) aesthetic properties are brought into perceptual presence. What I argue in the following sections is that mental imagery provides the structured configuration under which aesthetic properties of racialized bodies are brought into perceptual presence.

Given that we take perceptual experience to be prima facie justified, understanding aesthetic properties as being perceived could be problematic for an antiracist aesthetics for the body because it carries the implication that objects either have or fail to have the relevant properties. A defender of the racialized aesthetic hierarchy could simply claim that if white bodies are perceived as elegant and delicate it’s simply because they are. Racialized beauty standards would be explained, not by racist attitudes, but by the fact that some bodies, which just so happen to be white, are perceived as having aesthetic properties that we value, and others, which just so happen to be non-white, as having aesthetic properties we disvalue. While the antiracist critic can respond that racial prejudice leads to the attribution of racialized aesthetic properties, this just reveals a conflict. But if aesthetic properties are indeed directly perceived, why would we find such conflicts?

When it comes to aesthetic matters, these conflicts have come to be expected. This can be explained by the fact that aesthetic properties are perceiver and condition relative. Frank Sibley (1968) acknowledges that the perception of emergent properties, like aesthetic properties, depends on other cultural, experiential, and even emotional factors. More specifically, Levinson notes that variations in the attribution of aesthetic properties could be due to differences in either: a) perceptual sensibilities, understood as the “disposition to receive phenomenal impressions of certain sorts from various constellations of perceivable non-aesthetic features”; or b) attitudinal sensibilities, understood as the “disposition to react to phenomenal impressions with attitudes of favour or disfavour”, which are modifiable over time and involve a culturally formed component (Levinson 2006a: 331). In the end, Levinson favours the latter: although it could be possible to find variations in perceptual sensibilities, we have little reason to think that such differences are what cause conflicts in aesthetic evaluations (Levinson 2006a: 333).

In the case of racialized bodies, the fact that attitudinal sensibilities involve a culturally formed component could also favour them for explaining racialized aesthetic evaluations. The problem is that, in some cases, for example, anti-Black aesthetic evaluations go against subjects’ antiracist commitments. Part of the problem with racist aesthetic evaluations is that they’re built on something more subtle than overtly prejudiced views, something that can coexist with antiracist commitments. Consider again the dismal number of ballerinas of colour in the biggest ballet companies. It’s true that we still find examples of explicitly racist remarks against Black ballerinas.⁸ But ballet institutions have also repeatedly declared an explicit commitment to equality and diversity. As said before, reasons to exclude dancers of colour aren’t usually explicitly racist; they are disguised as aesthetic reasons.

Moreover, conflicts in aesthetic perception of racialized bodies are slightly different than the cases normally discussed in the literature. It isn’t simply about conflicting evaluations by two different critics who perhaps don’t share cultural, experiential, or emotional elements. Rather it’s about one critic evaluating the same kind of object with similar non-aesthetic properties, for example, Black and white ballerinas, and arriving at anti-Black aesthetic evaluations.

In the case of racialized bodies, the conflict in attribution of aesthetic properties needs to be explained both in terms of perceptual and attitudinal sensibilities. If we cannot explain the discrepancy by alluding to straightforward differences in attitudinal sensibilities, it seems like the discrepancy could be better explained in terms of the properties different bodies are perceived as having. In other words, the answer might be that non-white bodies simply aren’t perceived to be the relevant ways as to consider them beautiful, and that this leads to attitudes of disfavour. This is one way to interpret Taylor’s claim that aesthetic perception underwrites racial categorization: the fact that racialized bodies look a certain way is what gives reasons to respond with attitudes of favour or disfavour. This would mean that the divergence in the attribution of aesthetic properties could be better explained by alluding to perceptual sensibilities. Nevertheless, we would still need to explain how we could find, among human perceivers, variations in perceptual sensibilities.

In the following sections I argue that we have reasons to think that we might find differences in perceptual sensibilities due to top-down influences on perception. And this means that, although racist aesthetic evaluations might be explained by looking at perceptual sensibilities, these, in turn, need to be explained by attending to deeply engrained racist attitudinal sensibilities. In Section 4, I argue that racist mental imagery acquired from the white-European aesthetic tradition impacts the perception of aesthetic properties of racialized bodies. But first, let’s examine the role of mental imagery on aesthetic perception more broadly.

4. Mental Imagery and Aesthetic Properties

Aesthetic properties are high-level properties that depend on low-level properties of objects, like colour, pitch, or texture. Low-level properties refer to the paradigmatic features associated with specific sense modalities; for example, shape and colour for vision, or volume and pitch for sound. High-level properties, on the other hand, include a wide range of complex features, like natural or artefactual kind properties, agential properties, aesthetic properties, etc. Since high-level properties aren’t among the features associated with specific sense modalities, it isn’t clear that we perceive them. For example, while we undoubtedly perceive the colour or texture of a tree, it isn’t clear that we perceive an object being a tree. One could easily claim that only low-level properties are perceived, and that one arrives at high-level properties through other cognitive processes. Thus, the traditionally accepted claim that aesthetic properties are perceptible requires support.

Some people argue that high-level properties are represented in perceptual experience because perception is subject to various top-down influences.⁹ This means that perceptual states are affected by other mental states and that, as a result, the content of perceptual experience changes. For example, it isn’t that we perceive colours and shapes, and then infer that the object in our garden is a tree; rather, because of the influence of background knowledge on visual experience, “being a tree” is part of what we perceive. According to some, high-level properties are perceptible because the influence of other cognitive processes changes the phenomenal character of perceptual experience.¹⁰ As has been done for other high-level properties, one can argue that aesthetic properties are directly perceived because perceptual experience is affected by other cognitive processes. This is crucial for our purposes. If aesthetic perception in general results from the intervention of other mental states, this offers support for the claim that aesthetic perception of racialized bodies results from the intervention of other mental states, namely, those acquired from the interaction with given social contexts.

Dustin Stokes (2014) argues that aesthetic properties are perceptible because concepts or beliefs affect perceptual experience, and that experts are better at appreciating and judging artworks because their background knowledge makes them better at perceiving them. Stokes points to empirical evidence that suggests that concepts and beliefs influence colour perception. For example, in Christoph Witzel et al.’s (2011) study, subjects were presented with images of well-known traditionally coloured objects (e.g., a blue Smurf or the red Coca-Cola logo) in random colours, and were asked to shift the colour until objects were matched to an achromatic grey. The study found that participants adjusted the images toward the opponent colour to counteract the impression of apparent object colour against the grey background. This suggests that subjects continue seeing these traditionally coloured objects in colour even though they aren’t presented as such.

In line with this evidence, Stokes argues that background knowledge about aesthetic practices impacts aesthetic perception. Further, he notes that this isn’t a controversial position in aesthetics. Kendall Walton (1970) already argued that whatever aesthetic properties we perceive artworks to have depends on art-historical facts. According to Walton, the appreciation of artworks depends on the categories to which we perceive them as belonging. These categories refer to perceptually distinguishable ways of classifying artworks, like media, genres, styles, forms, artistic movements, etc. Categories under which works are perceived depend on art-historical facts, and they determine which non-aesthetic properties are aesthetically relevant and which aesthetic properties we perceive.

If the appreciation of works depends on perceiving them as belonging to specific categories, and these categories depend on art-historical facts, this could mean, as argued by Stokes, that art-historical knowledge impacts aesthetic perception. But background knowledge needn’t be cashed out in terms of propositional attitudes. Walton himself seems to disregard the propositional route when claiming that perceiving art categories doesn’t involve the consideration of art-historical facts, and that perceiving aesthetic properties doesn’t simply follow from knowing the correct category (1970: 365–66). ¹¹

Nevertheless, one can still construe aesthetic perception as involving top-down influences by turning to mental imagery. Here, I understand mental imagery as imagistic mental representations, representations with quasi-perceptual phenomenal character, that might involve contents in a single sense modality—for example, visual or olfactory—or multimodal contents—for example, visual and olfactory. In understanding mental imagery in regards to its sensory contents, I follow Margherita Arcangeli’s (2020) distinction between attitude and content senses of mental imagery. In its content sense, mental imagery doesn’t entail that we adopt an imaginative attitude toward the relevant representations. Mental imagery can feature as the content of concepts and other psychological attitudes, like belief and desire. Mental imagery can help explain aesthetic perception. Since aesthetic properties are high-level, the intervention of other cognitive processes could explain that they feature in perceptual experience. Mental imagery is a better candidate than propositional states to impact perceptual experience of aesthetic objects.

First, mental imagery has the right kind of phenomenal character to seamlessly interact with perceptual states. In proposing an indirect mechanism of cognitive penetration, Fiona Macpherson (2012) examines how the phenomenal character of quasi-perceptual states (e.g., imagining, daydreaming, or hallucinations) can impact perceptual states. Consider Witzel et al.’s (2011) study. Instead of thinking that propositional beliefs about the colour of these objects impacts perception, Macpherson proposes that subjects visualize coloured objects as they know them, and that the phenomenal character of these quasi-perceptual states interacts with the phenomenal character of their visual experience (Macpherson 2012: 51). Macpherson supports this by alluding to other cases where the phenomenal character of a state is the result of both perceptual and quasi-perceptual states, such as the Perky effect, in which perceptual experiences are mistaken for imaginative states, or cases in which hallucinations affect visual experiences. The result in these cases isn’t two different phenomenal states, a perceptual and a quasi-perceptual state, but one state with phenomenal character that is the combination of the other two (Macpherson 2012: 51–52).

Second, mental imagery is a better candidate to interact with perceptual experience because of the richness and fineness of grain of these quasi-perceptual states. Consider again subjects who behave as if they continue seeing pictures of well-known traditionally coloured objects in colour when that isn’t the case. They don’t just behave as if they continue seeing the Smurf picture in blue; rather, they behave as if they continue seeing the Smurf picture in that Smurf-like blue. So it isn’t simply a propositional state in the form of “Smurfs are blue” that affects perceptual experience. Rather, what seems to explain these results is that subjects have an imagistic mental representation that looks like that Smurf-like shade of blue. Mental imagery in its content sense allows that subjects might have an imagistic belief that Smurfs are that Smurf-like shade of blue. What I want to highlight is that it’s an imagistic, rather than a propositional, mental state affecting perceptual experience. It seems more likely that perceptual experience is affected by the phenomenal character of mental imagery looking “like that”.

Explaining aesthetic perception by means of mental imagery affecting perceptual experience is consistent with Waltonian contextualism. According to Walton, we perceive aesthetic objects as belonging to specific art categories, and perceiving them under these categories leads to perceiving the relevant aesthetic properties. As we interact with aesthetic practices, and as we become familiar with specific art categories, we acquire a broad repertoire of mental imagery. This mental imagery can provide the configuration under which non-aesthetic properties of objects are perceived in order for aesthetic properties to manifest. Mental imagery allows us to perceive non-aesthetic properties organized in such a way that aesthetic properties are brought into perceptual presence. Mental imagery can provide the organization under which non-aesthetic features of objects are perceived in two complementary ways. First, mental imagery of other exemplars of a given art category can work as an endogenous attentional cue in directing attention to specific features of perceived objects, say standard, contra-standard, and variable features (Walton 1970). Second, mental imagery can interact with perceptual experience, organizing non-aesthetic properties so that aesthetic properties emerge. Next, I propose how this picture helps in understanding racialized aesthetic evaluations.

5. Controlling (Mental) Images

The case of racialized bodies is very similar to the case of artworks because racial categories are artefacts, “things that we humans create in the transactions that define social life” (Taylor 2004: 86). Racialized bodies are cultural products that, like artworks, are also imbued with meaning. Processes of racialization just are ways to assign meaning to differences in human bodies to draw inferences to non-physical matters, such as psychological, cultural, or moral features, and by which social goods are distributed (Taylor 2004: 16–17, 24). As such, mental imagery acquired through the interaction with aesthetic practices is a good candidate for structuring the perception of non-aesthetic properties of bodies so that aesthetic properties are directly experienced.

I propose that the process of aesthetic perception of a racialized body works something like this. Bodies are perceived as belonging to a given racial category. Mental imagery of racial stereotypes, cultural representations that ascribe a set of properties to a group and that are acquired through the interaction with cultural practices, provides the organization under which specific bodily features are perceived. One perceives aesthetic properties that are consistent with racist stereotypes because of how this mental imagery structures perceptual experience. Just like mental imagery of exemplars of different art categories is acquired through the interaction with aesthetic practices, mental imagery that creates and reinforces racial stereotypes is acquired through the immersion in given cultural practices. This acquired mental imagery involves racist tropes and racialized standards of beauty made available by the white-European aesthetic tradition that dominates cultural practices.

In Section 2, we saw that critical race philosophers argue that perception of racialized bodies is affected by controlling images that sustain systems of oppression. Controlling images are able to play this role because of how they encode racist attitudes. For example, images of Black “welfare mothers” found throughout popular culture include as their content dehumanizing racist scripts that position Black women as lazy and irresponsible. Images of racialized individuals thus have an ideological content under white supremacy. Artistic representations of racialized individuals are tasked with sustaining white supremacist racial scripts by offering stereotypes and stock figures of people of colour. Racialized aesthetic evaluations are also explained by the lack of what hooks calls oppositional images, representations of people of colour that challenge racist scripts, and that could structure perception of and engagement with racialized individuals in different ways (hooks 2015: 75–76). We lack oppositional images because non-white aesthetic practices and artists have been historically excluded from the institutions of the artworld (Taylor 2016: 49–50). Taylor highlights that instead of complex and diverse representations of people of colour, we find “archetypical personifications of anti-black prejudices, defined by single, characteristic traits—servility, buffoonery, sexual rapaciousness, brutishness, and so on—rather than by the complex configurations that make for unique personalities” (Taylor 2016: 52).

Controlling images in aesthetic practices encode this ideological content because these representations don’t just present their content in a neutral way. Artistic representations are criterially prefocused: they highlight and obscure specific features of their content as to call for specific responses (Carroll 2003). When it comes to artistic representations of racialized bodies, these are already represented as having specific aesthetic properties: they are aesthetically prefocused. This aesthetic prefocus of racialized bodies isn’t trivial. As we saw in Section 2, establishing whiteness as an aesthetic ideal is central for the white supremacist project. The aesthetic prefocus of represented racialized bodies is part of the ideological content of controlling images because aesthetic differences are used to mark intellectual and moral differences among races.

Compare, for example, how white and Black women are represented side to side in paintings like Titian’s Diana and Actaeon or Manet’s Olympia. While Diana and Olympia are represented as the focus of their respective painting with a glowing skin that makes them look ethereal and vulnerable, the unknowable Black women at their side are represented with dull, faded skin, not only in the background, but in servile poses, at the service of white beauty. Artistic representations of racialized bodies already make salient certain features that are consistent with racist stereotypes, while downplaying and obscuring other features that would contradict them. Controlling images are thus affectively, symbolically, and aesthetically loaded. Artistic representations have at their disposal various strategies to encode these cultural meanings: the disposition of racialized bodies in respect to representational focus, light and colour, attire, poses and demeanour, and even symbolic and allegorical iconography. Aesthetic practices play a significant role in sustaining controlling images. Subjects internalize these representations as they interact with white-European aesthetic practices. These controlling mental images are evoked when encountering racialized bodies and structure aesthetic perception.

While high-level content of perception is controversial, it’s uncontroversial to think that mental imagery represents high-level properties. Mental imagery not only represents, for example, an object being certain colours and textures, but an object being, for example, an oak tree. We can take this further: mental imagery can represent a majestic oak tree. So, in the same way that artistic representations have ideological content, racist attitudes are encoded in mental imagery. Mental imagery of racialized bodies preserves the ideological content with which controlling images are imbued: I can invoke a visual mental representation of the ethereal and delicate white women of Renaissance paintings. So in the acquired mental imagery of racialized bodies, certain aesthetic properties have already been made salient. This mental imagery affects the perception of racialized bodies. This is how aesthetic properties of racialized bodies are brought into perceptual presence.

Racial categories under which bodies are perceived activate the associated mental imagery, which directs attention to certain non-aesthetic features of bodies. Mental imagery might affect perception beginning at this stage, affecting how bodily features on which it calls attention are perceived. For example, there is evidence that racial categorization impacts skin colour perception (Levin & Banaji 2006). But this stage isn’t enough to explain why negative aesthetic properties emerge when evaluating non-white bodies: we still need a story for why, for example, from perceiving darker skin in someone categorized as Black, negative aesthetic properties emerge. The mental imagery account can take us further. Mental imagery is aesthetically, symbolically, and affectively loaded. So what I propose is that the ideological content of mental imagery structures the perception of non-aesthetic properties of racialized bodies so that aesthetic properties consistent with racist stereotypes emerge.

Consider the case of Black ballerinas. The racial category under which they are perceived activates racist mental imagery. This mental imagery brings attention to specific bodily features, for example, the extension of their backs. Mental imagery might affect perception at this stage, leading to perceiving Black ballerinas as having a more extended back than their white counterparts. But this isn’t the whole story. The racist content of mental imagery further affects perceptual experience of non-aesthetic properties, like the extension of the back, in such a way that aesthetic properties consistent with racist stereotypes come into perceptual presence. Black ballerinas’ lines look lewd, rather than delicate, because of the prevalence of controlling images portraying Black women as sexually aggressive. The mental imagery account can also explain the asymmetry in numbers of male and female dancers of colour hired by the main ballet companies. Racist mental imagery might lead to perceiving Black male dancers as powerful or animalistic. But these aesthetic properties don’t disqualify them from classical ballet because strength and power are valued in male ballet dancers.

Of course, one could object to this picture by saying that mental imagery doesn’t feature in one’s experience of racialized bodies. So, how could mental imagery be affecting aesthetic perception of racialized bodies? First, recall that, as noted by Macpherson, in cases where perceptual and non-perceptual states interact, one experiences a single phenomenal state that results from a combination of the two. Second, mental imagery can be, in fact, unconscious and unavailable to introspection. For example, Bence Nanay notes that mental imagery can be both conscious and unconscious (Nanay 2013: 104–5), just as is the case of perception; and he posits that racist behaviour in antiracist individuals can be explained by alluding to pragmatic mental imagery, mental states that attribute action properties to objects (Nanay 2013: 126). Similarly, Ema Sullivan-Bissett explains implicit biases by alluding to unconscious imagistic and propositional imaginings (Sullivan-Bissett 2018). Controlling images work so well in sustaining systems of oppression because they are so pervasive that one mightn’t realize they are playing a role in our interactions.

This points to a further reason to think that mental imagery is a better candidate than propositional states in affecting perceptual experience to bring racialized aesthetic properties into perceptual presence. Like perceptual experience, mental imagery is incorrigible in a way that propositional beliefs aren’t taken to be. So even when someone might have embraced antiracist commitments, racist mental imagery might continue to structure perception of bodily features to bring certain aesthetic properties into perceptual presence rather than others.

This isn’t to say that beliefs aren’t playing a part in explaining racist attitudes. Understanding mental imagery in its content sense allows that it can feature as the content of other psychological attitudes like belief. What I want to point out is that imagistic representations might be able to explain the inconsistency one sometimes finds between racist aesthetic evaluations of bodies and antiracist commitments. While one might hold antiracist propositional beliefs, racist mental imagery might feature as the content of other beliefs: racialized bodies look like this. Further, these imagistic mental states are available to affect immediate experience.

6. Concluding Remarks

In this paper, I have offered support for the claim that aesthetic evaluations that disregard non-white bodies result from ways of seeing shaped by social structures. I have argued that mental imagery is responsible for providing the configuration under which non-aesthetic properties of racialized bodies are perceived so that aesthetic properties consistent with racist stereotypes are attributed to individuals. Mental imagery is a good candidate to explain how aesthetic properties of bodies are perceived because it has the right kind of phenomenal character to interact with perceptual states, and because of its richness and fineness of grain relative to propositional attitudes. Moreover, I argued that alluding to mental imagery can explain why aesthetic evaluations that disfavour individuals of colour can coexist with antiracist beliefs.

While it could be assumed that aesthetic evaluations are the result of bodies having certain features, I have argued that aesthetic appreciation of racialized individuals depends on the ideological load of representations we consume. Aesthetic evaluations systematically disregard individuals of colour because racist mental imagery affects perceptual experience so that non-white bodies aren’t perceived as having the relevant aesthetic properties. Mental imagery acquired through the interaction with predominant aesthetic practices impacts the aesthetic perception of racialized bodies because artistic representations present racialized bodies aesthetically prefocused. This means that representation matters. Aesthetic practices craft the mental imagery that is available to impact the aesthetic perception of racialized bodies. Positive, complex, and diverse representations of people of colour are needed to have mental imagery available to structure the aesthetic perception of racialized bodies differently.

One might worry that the mental imagery account might lead to relinquishing responsibility over racist aesthetic evaluations, and to complacency in regards to racist attitudes that seem beyond one’s control. But this doesn’t follow. While mental imagery affecting aesthetic perception might be itself out of our control, we are responsible for interrogating the source of aesthetic evaluations, especially when we should, by now, be well aware of the broad range of biases impacting cognitive processes. Moreover, we are responsible for representations we choose to consume, produce, and amplify, as well as for interrogating the sources of racist stereotypes. The mental imagery account thus points to the relevance of what hooks calls the oppositional gaze (hooks 2015: 117–25): we should critically engage with aesthetic practices with an “understanding and awareness of the politics of race and racism” (2015: 123).

This opens the door for concluding on a less grim picture. Rather than thinking that we are doomed to fall prey to racist aesthetic perception, the determinant role of predominant representations of racialized individuals in structuring so-called immediate experience highlights the relevance of community-wide action. It isn’t enough to disavow racist views in private. Instead, antiracist commitments demand that aesthetic communities, both creators and consumers, adopt an oppositional approach in our engagement with aesthetic practices by which we critically interrogate what meets the eye.