Institutional Racism without Racist Ideology: A Critique of Tommie Shelby's Marxist Theory of Racism

1.1. Ideology

Tommie Shelby develops a conception of racism over the course of three journal articles and one book chapter (2002; 2003; 2014; 2016).¹ The latter, entitled “Injustice,” is the first chapter of his book, Dark Ghettos: Injustice, Dissent, and Reform. In this work, Shelby succinctly articulates his theory of racism:

In its most basic form, racism is an ideology: [1] a widely held set of associated beliefs and implicit judgments that misrepresent significant social realities and [2] that function, through this distortion, to bring about or perpetuate unjust social relations. These beliefs and judgments form a kind of system of thought, which influences how adherents understand their social life and identities. Ideologies purport to be forms of knowledge (factual and normative) and so are amenable to critique. They don’t just contain false beliefs; more often they also obscure relevant information, organize facts in a misleading way, or rest on fallacious reasoning. Because of such epistemic flaws, ideologies constitute a distorted or biased outlook that conceals social injustices… [3] What is distinctive about racist ideologies is that they invoke or presuppose the problematic idea of race, a concept that attaches social meaning to visible inherited physical characteristics, continental origins, and biological ancestry. (Shelby 2016: 22, 23)

I introduce three terms for purposes of clarity. Racial cognition is cognition (including beliefs, implicit judgments, and reasoning) that invokes or presupposes the idea of race. Racial ideology is intersubjective racial cognition, or more precisely, a widely held set of associated beliefs and implicit judgments.² Racist ideology is racial ideology that satisfies all three of Shelby’s conditions of racism, as stated above. As I read Shelby, these conditions are necessary and jointly sufficient for racist ideology. I call [1] the distortion condition, namely, that a widely held set of associated beliefs and implicit judgments—call this network of beliefs, B—misrepresent significant social realities. I call [2] the injustice condition; it states that the set of cognitive attitudes B functions, through this distortion, to bring about or perpetuate unjust social relations. Finally, I call [3] the race condition; it states that the set of cognitive attitudes B invokes or presupposes the idea of race.³

Before clarifying each of Shelby’s conditions, it should be pointed out that people who harbor racist beliefs are often unaware that they do. One reason for this is that some of our attitudes are characteristically unconscious (“implicit”). Another reason is that racist beliefs can be learned passively, such that it is only upon reflection or in the course of experience that a racist belief comes across the awareness. Shelby explains:

the locus of ideology is common sense, that reservoir of background assumptions that agents draw on spontaneously as they navigate the complexities of social life and the demands of human existence. These assumptions are often held without full conscious awareness, creating various forms of cognitive and affective unconscious bias. For this reason people can actually be surprised to learn that they harbor racial prejudices or implicitly accept degrading racial stereotypes. (Shelby 2016: 23–24)

I now clarify the three essential conditions of racist ideology.

The distortion condition specifies that a belief or implicit association cannot count as racist ideology unless it has the property of cognitive distortion. This obviously includes false belief, but more commonly involves other cognitive and epistemic errors, such as inconsistency, oversimplification, exaggeration, half-truth, equivocation, circularity, neglect of pertinent facts, obfuscation, misuse of authoritative sources, hasty generalization, and so on (Shelby 2003: 166). Insofar as cognitive distortion is ideological, what is distorted is “our perception of social reality” (Shelby 2003: 178), particularly the social reality of oppression (Shelby 2003: 181). The idea is that ideology distorts our ability to understand an oppressive phenomenon as oppressive; or else it distorts our apprehension of the cause/s of oppression.

To illustrate, imagine an oppressed racial group that is falsly stereotyped as monstrous, criminal, or subhuman. What makes the belief ideological is not merely that the racial group is unfavorably (or even unjustly) depicted, but that the depiction is false and that this false belief is linked to some further distortion about oppression. For example, this particular belief might incline one to scapegoate and falsely blame some member of the stereotyped group for a social ill (say, a criminal act or a looming social threat). In other cases, racial stereotypying is irrelevant to the ideological belief’s oppressive function, as when racial subjugation is rationalized as the inevitable consequence of human nature; or when the burdensome plight is deemed an acceptable (albeit unfortunate) byproduct of a more fundamental societal commitment or priority; or it might be justified as a “necessary evil” which society must endure in order to tackle some greater evil. In yet other cases, ideological belief does not so much justify oppression as conceal it from view, as in narratives that misdirect attention away from the issue of oppression by concealing pertinent facts.⁴ Ideological beliefs can produce distorted normative judgments, including judgments that mistakenly ascribe poor moral character traits to an individual or group. Finally, ideology can distort our practical beliefs about which remedies are most conducive to repairing an oppressive social ill (e.g., if racial inequality is mistakenly perceived to be a consequence of cultural inferiority, the remedy for racial inequality may be inacurate).

The injustice condition of racist ideology requires that the distorted understanding of oppression perpetuate the condition of oppresion. Consider the practice of racial profiling. Suppose that two police officers interrogate a Black man, believing him to be “pushing drugs.” Questioning him, they caution him to “stop acting” as though he’s out for a casual walk and they insist he “give up his merchandise.” Though they purport to have good reason to suspect criminal behavior, they are in fact on a fishing expedition. The real motivation for the interrogation (whether conscious or not) is the belief that young Black males from poor neighborhoods are thugs, drug dealers, and criminals. Upon repeated and failed attempts to successfully address the officers’ hostile line of questioning to their satisfaction, the young Black man decides to walk away from the situation. He is immediately apprehended and arrested by the officers. In this instance, the racist ideology of the officers perpetuates anti-Black oppression by contributing to the wrongful treatment of young Black men—including harassing and humiliating this young man, as well as undermining his autonomy.⁵ In addition, the wrongful treatment is socially unjust because the officers’ racist ideology perpetuates injustice by reinforcing racial stigmatization.

The race condition of Shelby’s definition of racism requires that social meaning be assigned to inherited physical characteristics, continental origins, and biological ancestry. For example, being categorized as a drug dealer has social meaning. Social meaning can be understood in terms of its moral and social dimensions. The moral dimension is the significance of assigning a moral property to a group. A property P is morally significant in the sense that it has implications for how we treat (or ought to treat) individuals that are thought to possess this property.⁶ For example, inherited physical characteristics, continental origins, and biological ancestry might be correlated with an undesirable property, such as low intelligence or criminal disposition. People taken to be unintelligent are bound to be treated differently than people taken to be intelligent. The social dimension consists in the fact that the association of P with a particular group is widespread in the community. Going back to my example, the stereotype that young Black males are thugs is socially significant because this stereotype is not limited to a couple of racist police officers; after all, the police officers are not the inventors of the stereotype, but learnt it from others. In other words, their possession of this pernicious belief is a consequence of their membership within a society that widely disseminates the stigmatizing trope.

1.2. Institutional Racism

We have seen that racism, for Shelby, exists when three conditions obtain. Roughly: cognitive error distorts the understanding of racial oppression; second, the distortion functions to perpetuate racial oppression; third, the distortion invokes or presupposes racial belief. Shelby deserves credit for rigorously articulating a novel cognitive theory of racism that re-invigorates the Marxist conception of “ideology.”⁷ However, there are many other entities—besides beliefs, implicit associations, cognitive errors and patterns of reasoning—that are called racist. Examples include racist persons, racist actions, racist institutions, racist societies, racist speech, racist artifacts, racist symbols, and so on. Not surprisingly, a common objection to virtually all theories of racism is their failure to accommodate the “categorial plurality” of racism.⁸ Shelby designs his own account of racism to meet this desideratum.⁹ He argues that his ideological theory can accommodate non-ideological forms of racism at the personal and institutional levels by virtue of the fact that racist ideology is causally linked to racism on both levels. In the remainder of this section I present Shelby’s account of institutional racism; in section two, I offer a critique of his account.

Shelby’s strategy for accommodating the categorial plurality of racism is to adopt what Jorge L. A. Garcia calls an “infection model” of racism (1996: 11). Infection models distinguish between racism’s core and its infections. The core is the category of entity that represents racism’s primary mode of manifestation. An infection is any category of entity that derives its racist nature from the core. For example, Shelby derives the notion of a “racist person” from his conception of racist ideology. He defines a racist person as an individual who explicitly subscribes to racist ideology; a racist person can also be someone who, unintentionally and unknowingly, is disposed to act on racist assumptions (Shelby 2016: 24). Shelby understands “institutional racism” as a derivative form of racism. Accordingly, he claims a racist institution is infected with racism when it is extrinsically racist or intrinsically racist. He seems to hold that an institution can be intrinsically or extrinsically racist, but not both.¹⁰ My critique will target Shelby’s account of intrinsic institutional racism; however, for comprehensiveness, I present his analysis of both kinds of institutional racism.

I begin with intrinsic institutional racism. Here racist ideology causes institutional racism by constituting it. That is, racist ideology becomes a cause of institutional racism when it seeps into the institution, thereby corrupting it. An institution, he explains, is “a formal system of roles and rules that enable and regulate sustained cooperative action for some specified purpose.”¹¹ He identifies three modes of corruption; that is, three modes of intrinsic institutional racism (Shelby 2016: 26):

• First, racism can exist in the goal/design of the institution: this occurs when the public rationale or officially stated goal of an institution expresses racist ideology. For example, the American institutions of chattel slavery and Jim Crow segregation were racist in their goals, for the public rationale (as well as the state’s justification) of these oppressive institutions was that Blacks are inferior to whites.

• Second, racism can exist in the content of an institutional rule or policy: this occurs when the content of an institutional rule includes racial bias or is discriminatory. Consider ostensibly race-neutral laws that are harsher on crack cocaine (associated with Blacks) than on powder cocaine (associated with whites). Even though these laws are not explicitly justified on racial grounds, they are intrinsically racist because they are manifestly biased against Blacks.

• Third, racism can exist in the application of an institutional rule or policy: this occurs when a legitimate institutional rule is either misapplied or selectively applied by an administrator harboring racial bias or discriminatory intent. For example, the policy that police officers ought to approach and question suspicious actors seems legitimate. Yet, as we have seen, a police officer who harbors racial bias or hostility might abuse this institutional norm. Such cases illustrate the third form of intrinsic racism.

Because Shelby takes racist ideology to be the core of racism, he owes us an explanation of how racist ideology can “infect” an institution—i.e., how it injects racism into the institution. His explanation is that racist ideology either informs the institution’s goal by forming the basis (rationale) for bringing it into being; alternatively, it informs the content or application of an institutional norm or policy. In each instance, racist ideology explains the unjust institutional practice.¹²

Next, let us consider extrinsic institutional racism. Shelby writes that “On the extrinsic conception,”

an institution’s policies are regarded as racist, not by virtue of the policymakers’ racist beliefs, but solely in virtue of the policies’ effects. Extrinsic institutional racism occurs when an institution employs a policy that is race-neutral in its content and public rationale but nevertheless has a significant or disproportionate negative impact on an unfairly disadvantaged racial group. …What is nonetheless wrong with the institution’s practices is that they perpetuate the negative effects of ongoing or past racist actions and thereby encourage racist attitudes and stereotypes. The underlying idea is that some groups in society are already disadvantaged by racism, and an institution that is not intrinsically racist may nevertheless play a role in keeping these groups in their disadvantaged condition, thus leading some to conclude that they occupy this low station because of the disadvantaged groups’ culpable failings or inherent inferiority. (2016: 24)

Given that racism, at its core, is racist ideology, what is the relationship of racist ideology to institutional racism? In the passage, Shelby says that an institution that is not intrinsically racist can be extrinsically racist, if it perpetuates or reinforces racist ideology; in his own words, it “perpetuate[s] the negative effects” of racism and “thereby encourage[s] racist attitudes and stereotypes.” César Cabezas (2022) has critiqued Shelby’s account of extrinsic institutional racism for its unintuitive implication that affirmative action policies are racist. He cites empirical evidence suggesting that affirmative action policies perpetuate anti-Black affect and stigmatizing anti-Black representations among whites. Hence, if we are to avoid the implication that affirmative action policies are racist, we need some other account of extrinsic institutional racism. Cabezas proposes we think of extrinsically racist institutions as, first and foremost, causing durable racial inequality (rather than racist ideology). As he writes, racist ideology rationalizes and normalizes extrinsically racist institutions, which are sources of durable racial inequality. As a consequence, “racist ideologies stabilize the racist impact of extrinsically racist institutions by making their impact appear legitimate and unproblematic” (2022: 12).¹³ I welcome this modification of Shelby’s account for reasons I explain elsewhere.¹⁴

I now turn to my critical evaluation of Shelby’s theory of racism. I am not convinced that his account of racist ideology is a successful infection model for every form of institutional racism. In particular, I contend that some instances of institutional racism evade explanation in terms of racist ideology. To defend this claim, I present a thought experiment that offers a counterexample to Shelby’s analysis. I then offer an analysis of this thought experiment that exposes the inability of Shelby’s theory to accommodate the categorial plurality of racism.

2.1. Transparent Institutional Racism (TIR): A Counterexample

In this section, I present a counterexample to Shelby’s theory of racism and, by extension, his theory of institutional racism. My case involves intrinsic institutional racism that cannot be explained by racist ideology because racist ideology is not constitutive of the institution. I call my example “transparent institutional racism.” It is transparent in that the oppressor’s use of racial categories to sustain racial oppression is widely acknowledged and its rationale is both accurate and explicit in its depiction of social reality.

Transparent Institutional Racism (TIR). Imagine a racially oppressive society that does not have a history of racism. In this society, there are two groups, the oppressor group (designated as “whites”) and the oppressed group (designated as “nonwhites”). The term “white” picks out individuals that belong to the oppressor group and are characteristically light-skinned.¹⁵ The term “nonwhite” picks out individuals that belong to the oppressed group and are characteristically dark-skinned (non-light skinned). Within each group, there is a norm of endogamy (reproduction is restricted to other members of the group) and a norm of ancestry (racial identity is transmitted by descent). Whites, being the dominant group, invoke their will to define nonwhites and their kin as their social subordinates for the benefit of whites and their kin. They introduce the norms of endogamy and ancestry to ensure that racial subjugation persists indefinitely via the intergenerational transfer of social statuses. Whites, who obviously benefit from these arrangements, nonetheless acknowledge (1) that their reason for oppressing nonwhites and their kin is to benefit whites; and (2) that the norms of endogamy and ancestry are designed to facilitate the persistence of racial oppression in perpetuity. They do not ground these social arrangements in claims of genetic, spiritual, or metaphysical inferiority; nor in claims of cultural inferiority. No one thinks (or argues), for example, that it is the fate of nonwhites to be oppressed or that they have a defective culture which makes them deserving of their status. On the contrary, the social position of each group is widely acknowledged to be an accident of history and birth which arbitrarily benefits whites at the expense of nonwhites. Even the norms of endogamy and the markers of skin color are widely recognized as arbitrary and socially constructed properties; skin color, for example, is a convenient way of distinguishing members of each group and nothing more. Therefore, neither the prevailing racial categories nor whites’ explanations of social inequality involve false, misleading, or stigmatizing representations. Moreover, the central reason why this racial system works is that the dominant racial group has the power and willingness to make it work. That is, the oppressor group acknowledges that it oppresses because of its desire to maintain a superior material advantage. Whites readily acknowledge two motives; first, they do not want to partake in the menial jobs and laborious, back-breaking tasks reserved for nonwhites; second, they do not want to share in the burdens that an egalitarian society would provide were racial equality to be the law of the land. The motivating variables that sustain inequality—power/control¹⁶ and material advantage—constitute the expressed rationale offered to explain the oppression. These reasons are not cited as moral justification of white-on-nonwhite oppression, but as prudential justification. As an explanation of white-on-nonwhite oppression the account is accurate.

That completes my description of TIR. Here is my question: Is my racially oppressive society racist? Intuitively, it seems so. Indeed, I claim that it is overwhelmingly obvious that it is racist. Yet, given Shelby’s definition, he must deny the obvious. For there are no falsehoods or other epistemic errors in whites’ explanatory account of society. That is, there is no “cognitive distortion” here. For example, whites who appeal to the categories of “white” and “nonwhite” in explaining society do not understand these terms to denote natural race-based essences. Rather, they understand these terms to denote arbitrarily constructed social groupings, which of course accurately represents the social ontology. For example, consider whites’ narrative about why racial oppression exists. Their narrative underscores the dynamics of power and privilege, domination and oppression—that is, what obtains. Whites’ understanding of the role of biology is also correct; their use of skin color in marking off whites and nonwhites is an arbitrary but efficient means of categorization, for the sake of implementing differential treatment.¹⁷ To be sure, whites are admittedly vicious in my scenario, but they are aware of this fact, do not deny their moral depravity, and do not hide behind rhetoric that obscures their representation of themselves as morally wicked. Their explanations of the racial order involve no moral (only prudential) justifications of white domination (and their prudential explanations capture their real motivations and interests).¹⁸

In virtue of what, exactly, is TIR racist? More than one plausible explanation is available given the literature on racism, and I think this further confirms TIR as a viable counterexample. For example, whites in TIR are arguably racist because they harbor race-based ill-will (and disregard) toward nonwhites.¹⁹ Alternatively, one might argue that TIR is racist independently of any connection to race-based ill-will. Consider the impact-based approach to racism. According to it, any set of practices that creates or perpetuates race-based harm is racist.²⁰ This, too, plausibly explains the racism in TIR. My aim in this paper is not to defend any particular definition of racism (indeed, the best thing to say might be that TIR is racist for more than one reason or in more than one way, as the pluralist could argue). Rather, my aim is to critically examine Shelby’s proposal for his own definition.²¹ Hence, for purposes of this paper, I do not endorse either of these accounts of racism, as I do not see that I need to establish a theory of racism to defend the following claims: (1) intuitively, TIR is an instance of racism; (2) Shelby’s theory of racism fails to accommodate this intuition.

2.2. Some Objections to TIR

I now consider four objections to my critique.

Objection 1: Is TIR psychologically plausible?

One objection to my argument is that it is cognitively implausible for people to dominate a group of human beings without their cognition involving or eventually becoming ideological in the pejorative sense which entails distortion. It seems plausible that, over time, TIR’s social stratification system would foster distortive attitudes among whites regarding their social position—for instance, how could whites avoid organically developing a feeling of entitlement to their unearned privilege? How could they not come to view these entrenched social arrangements as “natural” and “good”? Similarly, the hierarchical structure of TIR would likely produce affective responses in whites—say, contempt and hostility—that would eventually distort their understanding of their subordinates vis-à-vis stigmatization.

To this objection I offer two replies (I add an additional reply in the concluding section). First, I do not agree with the premise that TIR is cognitively implausible. It seems to me that it is not far-fetched that a group G1 might systematically disrespect another group G2—even as G1s acknowledge that they are not inherently entitled to disrespect G2s and have no moral warrant for doing so—all the while being transparent about this fact. To be sure, it seems plausible that human beings have a psychological tendency to naturalize entrenched social arrangements (i.e., regard them as appropriate). However, this tendency seems compatible with taking preventive measures to effectively preclude its manifestation. For example, weekly trainings might be implemented to remind whites that if they start to feel inclined to justify their superior social position, they should remember that these feelings are demonstrably unwarranted, since whites’ superior position is completely arbitrary and, in any case, nonwhites are fully human. Moreover, we can imagine the tendency to rationalize injustice being curbed through education. If whatever can be mystified can also be demystified, then the tendency to mystify can be curbed through active modes of managing this tendency.

Bearing the above in mind, I offer the following emendation to TIR in order to account for the natural tendency to rationalize oppression. Let us suppose that whites, being acutely aware of the human tendency to internalize stable social arrangements, implement a rigorous educational initiative to teach whites, especially the youth and the most privileged members of society (elite whites), about the actual social nature of racial divisions, with an eye toward curtailing all of the following: their natural affective responses to their social subordinates; the human tendency to naturalize oppression; the natural desire to justify the dominant group’s superior position as morally just; and, by consequence of the previous measures, their potential to stigmatize nonwhites. In short, we can build into TIR that whites take sufficient, calculated measures to prevent the germination of distortive cognitive attitudes. In this way, the psychological mechanisms that lend plausibility to the notion that TIR is incompatible with human psychology are curbed.

Second, I reply that even if TIR is cognitively implausible, contrary to what I’ve argued, it nonetheless appears to be a conceptual possibility, which arguably suffices for my argument. Suppose that a person S possesses the power to generate bombs from the palm of her hand, at will. Imagine that S uses this power to oppress a racialized group in virtue of their race. Would we not be right to deem such a person racist? It seems obvious we would and should. Now, of course, no such person exists, as generating bombs at will is physically impossible. But the fact that this scenario is contrary to fact does not undermine the conceptual point, which is about the application of the term “racism.” That TIR appears to be racist and that Shelby’s theory of racism fails to accommodate this case suggests that his theory is conceptually inadequate.

Objection 2: Does TIR involve cognitive distortion at the level of normative judgment?

One possible objection to my argument asserts that social classifications have normative force.²² For example, it might be thought that, given that “white” means a superior and “nonwhite” means an inferior, “S is a nonwhite” implies “S should be subordinated.” It might then be thought that the normative judgment that nonwhites should be subordinated is plainly false; hence, that TIR involves cognitive distortion.

I reply that the objector confuses the grammatical ought with the moral ought. Consider the proposition “The king moves (must move) one square at a time.” The proposition tells us how the king is to move. The “must” in this case is not a moral must. Rather, it tells us how the piece is to be moved in a game of chess; that is, this is how the king must move in order to count as playing chess. The rule does not state how the piece ought to be moved if there is no interest in playing chess; nor does it state how the piece ought to be moved if one is to count as being moral. Similarly, the inference from “S is a nonwhite” to “S should be subordinated” states how whites ought to treat nonwhites given that whites want to participate in the “game” of oppression. That is, the rule reveals something about the logic of social oppression in TIR. The point is that the “ought”/“must” is not linked to a moral judgment, but to a grammatical one. To know this inference pattern is to understand who the whites are: they are those whose role it is to subordinate and exploit nonwhites. This includes understanding that the whites’ dominant social position enables them to assume ownership of the spoils of nonwhite labor, and to be the primary beneficiaries of their labor. From a moral perspective, this racial institution is morally abhorrent. As stipulated in my account of TIR, whites recognize this moral deficiency, which is likewise constitutive of a proper understanding of what it is to be “white.”

Objection 3: Is TIR race-based?

A third objection to my argument is that it is not at all clear that TIR is an example of racial domination. Against Garcia’s theory of racism, Shelby argues that it is misguided to say that an attitude is race-based just in case it is based simply on the fact that the targeted person is of one “race” rather than another “and for no better reason.” For Shelby, the problem with this approach is that it

would leave the motives of the racist largely opaque, mysterious, even unintelligible. What would it mean for a racist to hate someone simply because he or she is Black? Does the racist hate Blacks because they have dark skin and kinky hair? Surely “Blackness” has deeper meaning for the racist than that—unless he or she is psychotic. (2002: 414–415)

Applying this line of objection to my own thought experiment, whites in TIR might seem to target nonwhites for no other reason than that they are nonwhite. This, however, is psychologically untenable. We would have to believe, implausibly, that every last member of the white race is psychotic.

To this, I reply that I deliberately built into my thought experiment reasons that should dissuade us from thinking that whites in TIR suffer from psychosis. The motivating variables that perpetuate inequality in this scenario—such as power, control, and material benefit—are offered as whites’ basis for targeting nonwhites. Nonwhites are not targeted for the reason that they are nonwhite, and for no other reason.²³ Rather, they are targeted because, as a group, nonwhites make for easy targets, because whites are well-positioned to dominate and exploit the group, and, ultimately, because whites have a desire and material interest in dominating them.

Against this, it might be objected that my analysis is not “racial enough.” The objection would be that nonwhites seem to be targeted for reasons other than their race. That is, they are targeted because they are particularly vulnerable to oppression, because of how they are situated in the society. If this is correct, then it isn’t because of their race that nonwhites are targeted.²⁴ I reply that racial identity seems to be a constitutive feature of the targeting of the group. After all, whites would have to be perceived as a unified group to be considered an object of targeting. The unity of the group is entirely arbitrary and socially constructed, of course, but this is part and parcel of whites’ understanding of the group.

Objection 4: Is the conception of race in TIR plausible?

In light of my reply to objection 3, one might push back that it doesn’t answer the underlying concern of the objector. I have suggested that the social structure in TIR, which facilitates “white domination,” is racist. Hence, this is a form of racial domination. Therefore, since the domination is race-based, we need a convincing account of what “race” means in the context of TIR (e.g., an account that can help explain why “white” and “nonwhite” are racial terms). Up to now, I have not provided an analysis of the idea of race that is operative in TIR, and one might worry that such an account does not exist. How, then, is my hypothetical society divided by race? What does race-based cognition consist in on TIR? The objection can be formulated as a question: If my thought experiment is an example of racial oppression, in what sense is it racial? What plausible conception of race underwrites TIR?

My answer is that the operative conception of race is Shelby’s own conception. Race in TIR consists in the assigning of social meaning to the heritable characteristics of skin color and ancestry. In my thought experiment, whites assign the meaning social subordinate to those classified as nonwhite and the meaning social dominant to those classified as white. These classifications are analogous to social class/status designations. The existence of race thus entails a certain kind of cognition—namely, the thought that people of one particular skin color and ancestry belong to the “subordinate group,” and the thought that people of a different skin color and ancestry belong to the “dominant group.” These meanings, which are written onto the skin color and ancestry of folks, go hand in hand with coordinated acts that function to materially partition the society. The result of such coordinated activity is social injustice—whites benefit materially as a group whilst nonwhites suffer materially.

Still, one might push back by suggesting that Shelby’s social significance view of “race” is controversial and arguably misguided. Supposing the objector is right that it must ultimately be abandoned, this calls my thought experiment into question. The worry is linked to the intuition that the social significance view of race must be misguided because racial/racist cognition is essentially distortive. In the next section, I argue that, given his commitment to accommodating the categorial plurality of racism, Shelby cannot swap his social significance view of race for the more common racialist view (which entails cognitive distortion). I also defend the plausibility of Shelby’s social significance view of race.

3.1. Racialist Ideology: The Challenge of Colorblind Racism

Belief in the natural superiority of one race and the natural inferiority of another is the paradigm case of racist ideology. However, Shelby rightly points out that racist ideology isn’t always packaged in the idea of natural superiority/inferiority. Indeed, he explicitly contrasts racism “during the period of American slavery,” where “black slaves were commonly thought to be docile, superstitious, easily satisfied, and obsequious” by nature with racism directed at the culture of Blacks “in the present postindustrial phase of capitalist development, [where] blacks are more often viewed as socially parasitic, full of (unjustified) anger, irresponsible, and dangerous” (2002: 417). Shelby offers similar examples elsewhere. For instance, “racist ideology underwrites the widely held view that (most) black women who receive welfare support are poor because they are lazy, irresponsible, and promiscuous” (2003: 165). Or again:

it is believed that members of some racial groups tend to be less hardworking, less law abiding, less intelligent, and so on, without insisting that all members of such groups possess these negative characteristics and without necessarily concluding that these traits are (completely) congenital. (2003: 67).

For Shelby, the salient difference between the racist ideology that was prominent in the classic period and the racist ideology that is most prominent today is that today’s racists tend to “emphasiz[e] the ineradicable cultural pathology of blacks rather than their biogenetic inferiority” (2003: 169).

Despite this shift in ideology, the new racist ideology retains a similar social function as in the past; namely, it continues to sustain racial injustice. As Luc Faucher explains, many whites today invoke culture-based stereotypes about racial groups with the aim of

exculpating whites of their responsibility for racial disparities and/or faulting cultural features of racial minorities for the existence of these disparities (for instance, blaming disparities on a lack of effort from their members, or loose family organization). … Moreover, their negative affects are not necessarily grounded in hate or hostility… (2018: 410)

I will variously refer to post-civil rights cases of racism, which mark a fundamental shift in racist ideology, as the new racism and as colorblind racism. I will refer to the racist individuals in the new racism cases as colorblind racists.²⁵ In most of his writings on racism, Shelby explicitly argues that the “new” racist ideology is problematic because it satisfies his definition of “ideology.” Not only is its content misleading or false, but it also serves the same racist function as classic racist ideology and retains the same problematic idea of race. Since Shelby is interested in articulating a unified theory of racism and since he takes the new racism cases to be genuine instances of racism, he aims to accommodate these cases in his theory. Therefore, he’s committed to the following desideratum: an adequate theory of racism must accommodate colorblind racism.

When does cultural criticism become racial criticism? Anthony Appiah (1990: 5) thinks the phenomenon of criticizing a racial group’s cultural practices and psychological dispositions rather than its natural essence can be a form of “extrinsic racism,” the idea being that it is sometimes racist to think that a racial group is defective solely in virtue of its extrinsic properties.²⁶ For instance, one might disapprove of “street culture,” not because one dislikes or hates practitioners of street culture, but because one thinks street culture is characteristically rude, obnoxious, or cruel. This becomes racial for Appiah when one thinks that the practitioners’ racial essence disposes them to street culture. He calls his proposed theory racialism, which, as he explains, is the doctrine

that there are heritable characteristics, possessed by members of our species, that allow us to divide them into a small set of races, in such a way that all members of these races share certain traits and tendencies with each other that they do not share with members of any other race. These traits and tendencies characteristic of a race constitute, on the racialist view, a sort of racial essence; and it is part of the content of racialism that the essential heritable characteristics of what the nineteenth century called the “Races of Man” account for more than the visible morphological characteristics—skin color, hair type, facial features—on the basis of which we make our informal classifications. (1990: 5)

Undergirding racialism is the idea of “heritable characteristics,” which implies that certain traits are inherited from one’s progenitors, as a matter of nature.²⁷ Lawrence Blum helpfully designates this “the claim of ‘inherency.’” Inherency is the proposition that “certain traits of mind, character, and temperament are inescapably part of a racial group’s ‘nature’ and hence define its racial fate” (Blum 2002: 133).²⁸ Roughly, racialism implies that if a person S has racial property P then S has this property necessarily (i.e., P is inherent to S).

Consider how Appiah’s account of extrinsic racism would account for colorblind racism. Since extrinsic racism presupposes racialism, both the classic and the new racial ideologies must be firmly rooted in racialism. Hence, the ideas of natural essences and inherentism must figure into the explanation of colorblind racism. It follows that the shift from natural racial essences to hardened cultural dispositions does not mark a fundamental shift in cognitive content. Instead, the shift from classic to colorblind racism mostly amounts to a change in the rhetorical expression of racialism. It seems the most plausible way to think about the shift is in terms of the move to conceal the underlying idea of racialism by dressing it up in cultural garb. Concealing something by dressing it up, of course, means that the “something” in question remains the same old thing it’s always been (like the wolf in sheep’s clothing). This makes it a case of “unacknowledged racism,” to borrow Naomi Zack’s felicitous term (2003: 255; see also Lichtenberg 1992: 92).²⁹

Does Shelby agree with Appiah that racialism is the root of all racial cognition, such that “new” cultural critiques of racial groups only count as racist if they retain a link to racialism? The first thing to recognize is that Shelby’s theory of racism presupposes racial realism. Shelby acknowledges this when he writes: “with the possible exception of the belief in the reality of ‘races,’ no one belief is essential to the legitimating function of the [ideological] belief system” (2002: 415). We next ask: what exactly is real when it is said that “races” are real? Shelby is a social constructionist. “Social construction” means that race, qua system of social meaning and practice, is a product of collective human ingenuity (a social creation).³⁰ Races are human inventions, but they are not invented as one invents a password (i.e., individual fiat) but as a group invents a social institution—via social coordination that spans a series of personal and collective decisions and interactions, mediated by social roles, norms of various types, power relations, and so on.³¹ So, his answer to our second question is that when a particular race exists, it is a particular social group that exists. A race is a social group created by human beings and dependent upon particular ideations and practices for its continued existence.

So far, we have seen that Shelby thinks races are real and that races are a particular kind of social group. Next, we need to consider what is distinctive about race; what makes a racial group different from other kinds of social groups? Racial groups differ from other groups by how they are conceived, by their cognitive or ideological content (Shelby 2009: 132). There is some textual warrant for thinking that Shelby analyzes the content of racial cognition in racialist terms. For example, in discussing contemporary cultural critiques of Black culture, Shelby writes: “Despite references to ‘culture,’ ‘race’ is still the operative (if not the ostensible) category used to identify the relevant groups” (2014: 67). Race is the operative category because, “Though the defects are now more often attributed to cultural characteristics, these are treated as thoroughly entrenched and (practically) unchangeable, which effectively racializes these purported differences” (67). On the same page he writes, “contemporary forms of racism are often nothing more than remnants of the classic doctrine [of racism], either operating implicitly in the background culture or repackaged to buttress the racial status quo” (67).

These and similar passages lend support to the interpretation of Shelby as one who thinks racialism underwrites racist ideology. First, Shelby uses terms like “ineradicable,” “entrenched,” and “unchangeable” which, at least on one reading, are plausibly interpreted as conveying that racial properties are permanent and inherent. This includes Shelby’s remark that treating properties as entrenched and unchangeable effectively “racializes” them, which seems to imply that “racialization” obtains precisely at the moment that differences are conceived as inherent. Second, Shelby alludes to racialism (an element of “the classic doctrine” of racism) being “fundamental” or still “operating” in the context of the new racism. The aforementioned considerations support the interpretation that the conception of race at work in Shelby’s understanding of racist ideology is racialism; that a social group counts as a racial one if its characteristics are treated as thoroughly entrenched and unchangeable properties.

Apparently working from the premise that Shelby is a racialist about the content of racist ideology, Faucher questions whether Shelby’s theory of racism is capable of accommodating the new racism cases. Shelby’s definition of racist ideology includes proposition [3]. This proposition asserts that racist ideology necessarily invokes the problematic idea of race. One way of interpreting this proposition is that Shelby thinks the content of the racist’s attitudes toward the constructed group is rooted in racialism. If so, the racist does not think of racial groups as socially constructed but as having inherent properties. Taking this interpretation for granted, Faucher argues that Shelby has no way of explaining why colorblind ideology is racist because colorblind racists characteristically eschew the reality of race. He writes:

it does not appear to be necessary to believe in the existence of races to be racist. Someone can sincerely profess that races do not exist, and yet still be racist if they have a malevolent attitude toward a particular racial group. Indeed, this is the reason why doxastic accounts [including Shelby’s ideological account] do not fare very well with the new forms of racism…like color-blind racism or symbolic racism. These new forms of racism are predicated on the idea that nowadays agents believe that biological races do not exist, but still entertain negative attitudes towards racialized groups. (2018: 416)

Faucher is suggesting that Shelby’s theory of racism-as-racist-ideology implies that if colorblind racists really are racist, then they necessarily believe in racial essences (despite their professed denials). The problem, Faucher argues, is that this implication is false. The colorblind racist isn’t just someone who thinks that race doesn’t matter; she is someone who thinks races aren’t real. Said differently, colorblind racism is supposed to be a new form of racist ideology, not the same old racist ideology albeit cleverly disguised. The racialist analysis of the new racism cases turns the colorblind racist into a covert and insincere classic racist. By contrast, the colorblind racist, for Faucher, may be transparent and sincere; a fundamentally distinct kind of racist. According to the sincere colorblind racist, attributed negative traits like being less hardworking, less law abiding, less intelligent, and so on, are not explained by a deeper racial essence (i.e., inherent racial properties), but by cultural deficiencies that are contingent and in principle revisable. In short, Faucher argues that the colorblind racist really does think that racialism is false.

To sum up the discussion so far. Shelby’s desideratum for a theory of racism is that it must accommodate the new racism cases. If Faucher is right that Shelby endorses a racialist conception of racist ideology, then it seems Shelby’s theory can’t meet this desideratum.³² Three considerations support Faucher’s reading. First, Shelby’s theory of racism clearly presupposes racial realism, the view that races exist. Second, his preferred brand of racial realism is social constructionism. Third, Faucher’s interpretation of Shelby’s constructionism attributes to him a racialist conception of how race is conceived by the racist. A helpful way of summing up Faucher’s reading of Shelby is as follows: whereas Shelby himself conceives of races as social kinds, he thinks racists conceive of races as natural kinds. This makes Shelby a social constructionist about race and a racialist about the form that racial ideology takes. To capture this thought, I define racialist racial ideology as the view that racist ideology necessarily involves the false belief of a social kind (a racial group) as a natural kind; this view implies that racist ideology always takes a racialist form.³³

3.2. Colorblind Racism without Racialism: Introducing the Instability Thesis

In the previous section, we considered two reasons for thinking that Shelby’s theory of racism is based on a racialist conception of racial cognition. First, Shelby’s constructionist account of race is compatible with a racialist rendering of racial/racist ideology. Second, there’s some textual warrant for attributing a racialist conception of race to him. We also saw that a racialist understanding of racial cognition renders his theory of racism vulnerable to Faucher’s objection. A racialist conception of racial ideology opens him up to the new racism counterexamples he sought to accommodate. In this section, I critically assess these considerations. My claim is that the aforementioned considerations fail, because Shelby actually adopts a wide conception of racial ideology, which accommodates the new racism cases of both the racialist and non-racialist varieties, as well as classic racialist cases of racism. After establishing this claim, I return to the implications of his wide conception of race for my thought experiment, TIR.

Regarding Faucher’s critique, I think he moves too quickly. He supposes, without argument, that the only idea of race that is or can be socially constructed is a racialist conception. But that seems false, as Michael Hardimon (2013) ably demonstrates in his discussion of various concepts and conceptions of race. Faucher needn’t assert that a racialist conception of racial cognition is the only game in town. It’s sufficient for him to assert that Shelby endorses a racialist conception of racial cognition. The problem with this argument is that a contextualized and even-handed analysis of the textual evidence suggests that Shelby is not convinced that racialism is the only game in town when it comes to racial cognition.³⁴ In particular, racial cognition without inherentism is possible. Below, I argue that throughout his writings, Shelby has remained consistent in his insistence that racist ideology is not a stable doctrine. That is, he has consistently rejected the stability thesis, as I will call it, the view that racist ideology is stable over time. For example, in his first essay on racism, Shelby writes that “racist views are part of a complex and dynamic system of ideological belief. These beliefs have greater specificity and variety than the belief in a hierarchy of ‘races’; they often shift and are reformulated given specific political contingencies, economic circumstances, and sociohistorical context” (2002: 415) Similarly, in a different essay on racial ideology, he remains consistent on this point, writing: “An ideology’s content can shift over time in response to changes in the sociopolitical context” (2016: 23).

More telling is how Shelby elects to wield his rejection of the stability thesis to criticize Blum’s conception of racism-as-racial-inferiorization. He criticizes Blum’s definition for invoking an overly rigid conception of ideology. As he explains:

[Blum] treats ideologies as explicit and fully developed doctrines whose propositional content is stable over time. However, I mean “ideology,” not as it is currently understood in everyday discourse (where it often denotes partisan political doctrines), but as it has come to us from the Western Marxist tradition. (Shelby 2014: 66)

Racialism, as Appiah conceives it, is an example of an explicit, fully developed, and stable doctrine. To be clear, what Shelby rejects is not the possibility that race can be conceived in racialist terms. Rather, his objection is against defining racism as a fully developed, stable doctrine. His position is that a stable definition of racist ideology is needlessly restrictive. Again, that is because “the content of an ideological belief system can change over time” for many reasons. One reason is that ideologies are political tools that change “in response to shifts in the cultural, political, and economic context and also, importantly, in response to social criticism” (Shelby 2014: 66). The racialist conception of race has been subjected to brutal social criticism from the scientific and anti-racist communities. This has led to respectable members of both communities eschewing the idea of racialism. Yet, racial groups continue to exist, argues Shelby, because people continue to believe in the reality of races (conceived at times as racialist groups and conceived at times as non-racialist groups); in particular, the content of these beliefs has evolved, becoming more social and cultural and less biological.

Faucher might offer the following rejoinder. If racist ideology presupposes belief in race, but the relevant conception of race is sufficiently wide to accommodate racist doctrines that are racialist and inherentist in form, on the one hand, as well as racist ideologies that are non-racialist and social in form, on the other hand, then Shelby owes us an alternative definition of “race.” Shelby owes us an explicit articulation of his wide conception of race, because it has seemed to many that racial realism entails racialism. What, then, does race consist in, for him? We’ll get to that in a moment. I want to first emphasize that his conception of race is compatible with the instability thesis, the view that racist ideology tends to be unstable over time.³⁵

The proposition that racist ideology is essentially unstable is plausible. We can better see this if we first reframe the colorblind racist’s sincerely held disbelief in the existence of races. The terms “realism” and “anti-realism” are relative terms. The substance of an individual’s anti-realism depends directly upon the substance of the realism one rejects, and the idea of race is open to more than one way of being real. The most common way race is thought to be real is cashed in racialist (naturalist) terms. What the colorblind racist sincerely rejects when she rejects racial realism is not the idea of race as such, but the idea of racialism (hence also, inherentism). The sense in which the colorblind racist is an “anti-realist” is that she thinks there are no natural or biological races. This means that it is possible for the colorblind racist to be an “anti-realist” of sorts while being a “racial realist” in Shelby’s constructionist sense. Shelby, for one, is an anti-realist about the racialist conception of race (for he thinks races are social groups rather than natural groups). He also seems to think that being in the grip of racist ideology does not necessarily entail being in the grip of racialist and inherentist cognition. The racialist form of racial cognition is not the only game in town. What racist ideology entails is belief in the reality of race, a position that is consistent with the racist conceiving of races as either social kinds or natural kinds, as the case may be. Hence, the counterexample in Faucher’s objection turns out to be no counterexample at all. It only appears to be a counterexample because Faucher mistakenly conflates racial realism with racialism; that is, he overlooks the possibility of forms of racial cognition that are non-racialist and non-inherentist.

So what is Shelby’s conception of race? Shelby proposes, you will recall, that we conceive of race as the attachment of social significance to observable heritable characteristics, ancestry and continental origins. This, I submit, is his positive and wide view of race, his alternative to the stable racialist conception. Races, for Shelby, are groups of human beings that, roughly speaking, have historically been coextensive with the classic (racialist) race categories and are designated by the same labels (e.g., “Black”). Ontologically, races are defined in terms of social relations and the nature of these relations is determined by the particulars that determine the idea of race in some actual context—that is, the actual social meaning that the relevant actors assign to observable heritable characteristics, ancestry and continental origins at a site. The content of the idea of race—or the social meaning assigned to racial groups—varies across time and place, in accordance with the social constructionist view of race but also with the instability thesis. In some instances, the social meaning of race may depict races as natural and/or stable categories; in other instances, it may depict races as social and/or unstable categories. Therefore, because of the instability thesis, Shelby’s account of racial cognition is wide enough to accommodate the classic cases of racism as well as the new cases of racism. Pace Appiah, one can be a racist without being a racialist.

3.3. Colorblind Racism is Analogous to Classic Racism

This brings us to the second consideration, the textual basis for attributing a racialist conception of race to Shelby. Is the best overall interpretation of his writings on racism that he thinks colorblind racists are racialists who are merely pretending to be cultural critics? Or does he think that at least some colorblind racists (whom he concedes are genuinely racist) are sincere in eschewing racialism and in grounding their criticisms solely in sociocultural considerations? To answer these questions, consider this passage. The colorblind racist, he says, insists

that members of some racial groups tend to be less hardworking, less law abiding, less intelligent, and so on, without insisting that all members of such groups possess these negative characteristics and without necessarily concluding that these traits are (completely) congenital. …Rather than think of these more contemporary beliefs and assumptions about racial groups as forming a different [non-racial] ideology, I agree with those who argue that it is more accurate to view them as expressions of a newly constituted racist ideology (Shelby 2014: 67).³⁶

I submit that this passage enables us to reconsider the meaning of Shelby’s claim that talk of racial properties is “unchangeable” and “ineradicable”—that is, to reframe talk which, on the surface, sounds inherentist. While his mentioning of such properties appears to be expressing the view that racist ideology presupposes racialism, the latter part of this passage suggests an alternative possibility. For example, Shelby does not say that colorblind racism is the old ideology of classic racism in disguise. Instead, he says it is a “newly constituted” ideology. This strongly suggests that Shelby thinks sincere colorblind racism is possible.

Second, in this passage, Shelby arguably makes an analogical argument based on a resemblance between classic racism and colorblind racism. The resemblance is that new racism involves treating people as though their racial properties were inherent. In other words, his suggestion is not that the newly constituted racist ideology pretends to be new but is not really new because it is but a cleverly disguised formulation of the classic doctrine. Rather, his suggestion is that the newly constituted ideology really is new and is, notwithstanding this fact, relevantly similar to racialism inasmuch as it shares a common history, having evolved out of the classic doctrine of racialism and having evolved in direct response to antiracist resistance to the classic doctrine. Because of this shared historical connection, it is not surprising that colorblind racism has the same social function as classic racism. Namely, racist ideology rationalizes racial oppression and does so in a way that, in effect, makes it permanent; for all intents and purposes, it sustains the oppression of the group even as some members might escape the group’s racial fate.³⁷ After all, if traits such as laziness and lack of drive are seen as cultural and non-inherent, and are also seen as sufficiently hardened, then these traits might as well be inherent. Oppression can be rationalized just as effectively by the sincere colorblind racist in the post-racial era as by the classic racist in the classic era. My analogical reading of Shelby is made all the more plausible by the fact that it is consistent with his expressed commitment to the instability thesis. In fact, the proposition that racist ideology is unstable and changes over time helps to explain the genesis of colorblind racism, hence also its differentiation from classic racism.

Notice that this reading has the additional virtue of explaining why Shelby thinks racialism is in some sense “fundamental” or the “linchpin” of racist ideology, namely, that, historically speaking, racialism is the doctrine that brings the modern idea of race into existence. In doing so, racialism introduces the idea of inherent properties and with it, the idea of a fated social hierarchy. In our “post-racial” moment, the notions of inherency and fate have been softened. They are no longer literally invoked by sincere colorblind racists; however, the shadows of these ideas remain for all relevant intents and purposes. By this I mean that the social function of oppression is sustained through contingent explanations of “ineradicable cultural inferiority” and “deeply entrenched” defective values. Shelby seems to think that in order for a belief to count as ideologically racist, it must bear a family resemblance to the social function of racial oppression and a traceable historical connection to the classic idea of race. This implication is interesting in its own right and deserves further attention. However, it falls outside the scope of my argument, so I will not take it up further.

Instead, I return to my main argument. We have seen that Shelby conceives of race in terms of the attachment of social significance to certain properties of human groups. We have seen that he thinks these attachments are contingent, revisable, and change over time. Consequently, the social meaning of race tends to be unstable over time. In particular, it may take a racialist form at time T1 and a cultural/non-racialist form at time T2. Racial identity appears to be forged through the resemblances and historical connection of the two ideologies. What makes both ideologies racist is that they meet Shelby’s three necessary and sufficient conditions for racism: these widely held doxastic attitudes misrepresent significant social realities and function, through such misrepresentation, to bring about or perpetuate unjust social relations.

I thus conclude that if Shelby’s conception of race—which is crucial to what appears to be his best reply to Faucher—proves adequate for accommodating the new racism cases, then his conception of race is adequate for accommodating TIR. As argued in section two, TIR is an example of race-based oppression because it invokes the idea of race; that is, whites in TIR attach social meaning to the observable physical characteristics of skin color and the ancestry of those they classify as “white” and “nonwhite.” Put slightly differently, if the new racism cases are genuine instances of racism, then TIR is a genuine instance of racism, too. Conversely, if TIR is not a genuine instance of racism because it invokes a problematic conception of race, then Shelby’s conception of racism is likewise misguided. Of course, one might agree that Shelby’s conception of race is problematic and move to replace it with a better conception—say, one that is closer to the racialist conception. The problem with retaining his conception of racist ideology whilst rejecting his conception of race, from Shelby’s perspective, is that his theory of racism loses the ability to accommodate the new racism cases. As we have seen, Shelby argues that accommodating the new racism cases requires rejecting a stable understanding of the idea of race in favor of an unstable view. And an unstable view of race is consistent with racist ideology taking both racialist and non-racialist forms. So, either way, his account fails to meet the desideratum of accommodating the full categorial plurality of racism.