Critiquing racist ideology as harmful social norms

Introduction

What do we mean when we state that racism is an ideology? Ideology answers Wilhelm Reich’s (and before him, Spinoza’s) famous question: Why do the oppressed people not revolt against their oppressors? Ideology aims to explain the pervasive phenomena of voluntary servitude because its key factor is that oppression is not recognized as such due to ideological effects. Ideology explains the persistence of harmful social arrangements. In this article, I will specifically focus on the ideology regarding racism. According to Haslanger, ‘Ideology … is intended to function as an answer to the explanatory question: what explains persistent racial injustice?’ (Haslanger 2017a, 8). In addition to its explanatory function, theorists have argued that racist ideology has a negative cognitive impact on people in a pervasively racist society. The functional and epistemic aspect of ideology is interconnected. ¹

In this article, I will argue that racist ideology should be understood in terms of racist social norms that constitute certain incentive structures. Although this view may not resolve all problems related to defining ideology, it better accounts for the changeability of ideology. To this end, I will support my position by examining two influential accounts of ideology: those of Tommie Shelby and Sally Haslanger. While I limit myself to the examination of Shelby and Haslanger, before moving on to provide a synoptic overview of the article, I would like to situate my view in the discussions between analytical philosophy working on ideology and sociologists of critique as well as more classical Critical Theorists on the nature of ideology (critique). To this, I do not consider analytical philosophy (Shelby, Haslanger, Stanley and so on) and critical theory (Frankfurt School, Lukacs, Jaeggi, Althusser and so on) to be oppositional or mutually exclusive alternatives. When my view tries to reconcile these two camps, I believe that my view should be aligned with critical theory in a broadly pragmatic sense. ² For pragmatist critical theory, rather than picking sides, the ultimate aim is to bridge the gap between analytic philosophy and continental philosophy for the sake of collective enquiry and problem-solving in the face of problematic situations. In my view, whether in the analytic philosophical works on ideology or sociologies of critique, the ultimate purpose should follow the emancipatory aspects of critical theory to diagnose various social pathologies including ideological oppressions. Beyond the methodological oppositions, my view is pragmatic and integrative because it aims to critically synthesize the strengths of various theoretical camps while overcoming their shortcomings. In addition to my pragmatist orientation, as it will become clearer as we proceed, my claim is in line with the realist view of ideology critique. ³ One of my aims is to reinforce the realist ideology critique because transforming actual social practices is not a viable option when using idealist manners. Ideology critique must be effective and feasible rather than unachievable or overwhelming.

Using this idea as a blueprint, the article proceeds as follows: First, I will begin by reconstructing Shelby’s account of racism as ideology. After analysing three dimensions of ideology (epistemic, genetic and functional), I will argue that his view is overly cognitivist. In this regard, Shelby’s view is doxastic in noting that racist ideologies consist in misguided beliefs. Rather, what is essential in ideology is its functional dimension. Here, ideologies persist because they function to stabilize and reinforce the unjust status quo of subordination and oppression. ⁴ Then, I will turn to Haslanger’s account of ideology as cultural technēs. Her view is more functional than that of Shelby because the former is based on the account of social practice and culture. Although Haslanger is right in her critique of Shelby’s cognitivist view of ideology, I argue that what she calls nonideal moral epistemology weakens her overall insight. The problem is that without considering how to intervene in concrete social mechanisms, merely knowing certain moral truths may not practically motivate subjects under ideologies. Considered together, both Shelby and Haslanger present a narrow understanding of ideology, in terms of epistemic deficiency. Although Shelby and Haslanger have dealt with the discursive superstructure of ideology, both underestimate the functional substructure of ideology, such as social and psychological motives, desires and needs. Finally, I will argue that ideology is better understood from the perspective of a racist social norm. This account is explanatorily superior to Shelby and Haslanger’s views due to the emphasis on nonepistemic, nonmoral and functional aspects of ideology. Furthermore, this conception of ideology that I develop can help us to better see the connection between ideology and social changes.

Shelby on ideology: Epistemic, genetic and functional

In what follows, I will present why racist ideology cannot be merely a matter of misrecognition, that is, false and irrational ideas concerning races. To this end, in this section, I will begin by reconstructing Tommie Shelby’s cognitivist account of racism as ideology. ⁵ Social consciousness is considered ideological due to the following three properties: epistemic, genetic and functional. Following Raymond Geuss (1981), Shelby articulates these three properties (Shelby 2003, 163–64). First, ideologies have epistemic properties insofar as they are false, distorting and illusory. Second, regarding the aetiology or history of ideas, ideologies have genetically negative features in that they are adopted because of some bad psychological origins such as class interests. Third, ideologies have a certain social function or legitimating role of specific interests or social position. In a later paper, wherein these three points were incorporated, Shelby succinctly defines ideology: ‘An ideology is a widely held set of loosely associated beliefs and implicit judgments that misrepresent significant social realities and that function, through this distortion, to bring about or perpetuate unjust social relations’ (Shelby 2014, 66).

A notable factor in Shelby’s definition of ideology is that for it to serve a hegemonic social function, it should be misleading and distorting as well. That is, a belief with a hegemonic function is racist when the belief is false. Therefore, Shelby considers ‘ideological belief to be the primary object of ideology-critique, treating talk of ‘ideological practice’ as a derivative usage’ (Shelby 2003, 157). Contrary to this view, I think that Shelby should have emphasized the functional aspect of ideology, which is related to the stability of ideology over time. Recall the explanatory question about ideology: why do ideologies persist? One idealist answer to this question attributes it to individual ignorance, irrationality and faulty reasoning (Appiah 1990; Shelby 2003, 184). Here, it is assumed that real racists would exhibit ‘a systematically distorted rationality’ (Appiah 1990, 8). Indeed, since Marx identified ideology as irrationality (the illusionary image of camera obscura), it has been commonly argued that racism is inherently irrational. For instance, Michael Dummett says that ‘No one can rationally think that the great majority of members of any racial group are intellectually or artistically inferior to the great majority of members of some other group’ (Dummett 2004, 29). If so, racism should be rejected because racist beliefs are erroneously sustained by stereotyping, category mistakes, overgeneralization, equivocation, false dichotomy and inconsistencies (Goldberg 1990; Shelby 2003). ⁶

I argue that we need to see the functionality of racist ideologies, rather than making them necessarily dependent on their irrationality. For instance, Shelby writes: ‘Regardless of what other troublesome features a set of beliefs may have, if we think the beliefs are nevertheless true or sufficiently warranted, then, surely we must accept them’ (Shelby 2003, 165). For a belief to be racist, in his view, it should be false, in addition to wrongful social functions.

Before proceeding to Haslanger’s view in the next section, note that it can be misleading to assume that racism is inherently and necessarily irrational, not to mention the polysemy of rationality itself. For rationality as a contested concept is not a necessary component of functionality. In critiquing Shelby’s view, my point is not to rationalize or legitimate racism but to examine how racism is cognitively maintained and legitimated via ideology. Although racism usually involves racist beliefs concerning racial differences as being unchangeable and inevitable, what practically matters is the function of racist beliefs, rather than their truth condition. It may also be argued that even ideological beliefs that are true but dysfunctional can exist. Insofar as racism is primarily about domination based on differing racial ascription, it need not involve a very rational and coherent set of racist beliefs. Therefore, it does not follow that there is the need for a close connection between falsity and oppressive function for something to count as ideology. Shelby’s point is that being cognitively distortive is a necessary condition of racism. By contrast, my point is that beliefs, whether true or false, can have a hegemonic social function. Racism is not simply limited to a few irrational anomalies; this implies that a mere critique of racist belief would not abolish the racist practices as a whole. More importantly, as Haslanger says, ‘universal norms of rationality will only in rare cases be sufficient to adjudicate between different forms of life’ (Haslanger 2017b, 164). The point is that the task of determining the correct notion of rationality is far from clear and this difficulty is aggravated by the existence of diverse forms of life.

Haslanger on ideology: Cultural technēs

In this section, I will examine Haslanger’s practice-based account of ideology. Sally Haslanger sympathetically builds her theory of ideology based on Shelby’s initial account. According to Shelby’s view, ideologies are fundamentally propositional, whereas for Haslanger, the explanatory framework of ideology should be less cognitivist and more expansive. Although I largely concur with Haslanger’s critique of Shelby’s cognitivist account of ideology, I will argue that it falls short of proper ideology critique. While Haslanger is right that social movements against racism are a viable solution, her appeal to nonideal moral epistemology does not solve the issue.

According to Haslanger, although Shelby is excessively optimistic about the power of scientific and philosophical inquiry over ideology, ‘“successful” ideology isn’t always false. (This is, in fact, how we often end up forming the “ideological” beliefs in question – we look around us’ (Haslanger 2017a, 4). This position not only makes ideology something true-like but also implies that what others do and expect from us is significant. For instance, we learn about racist ideology and associated behavioural norms by simply looking around us. As Simon de Beauvoir puts it, ‘The American White relegates the Black to the rank of shoeshine boy; and he concludes from this that the Black is good for nothing but shining shoes’ (Beauvoir 1972, xxx). The point is that social knowledge and the social reality it represents are interdependent such that ‘ideology (sometimes) makes itself true’ (Haslanger 2020, 16). When such racist norms and racist ideology mirror each other to some degree, reasoned dialogue by itself is insufficient for critiquing, or dislodging racist ideology. If so, contrary to Shelby’s view, ‘it is not even clear how teaching people the biology of race and a liberal theory of justice will dislodge implicit bias’ (Haslanger 2017a, 5). The cognitivist view of ideology, which focuses on explicit beliefs and shared judgements, is inadequate for considering non-propositional or sub-doxastic elements of ideology such as implicit racial bias. ⁷ Such elements in ideology do not represent or misrepresent the world in any truth-conducive way. If so, in addition to reasoned debate, the conception of ideology should be expanded to include various attitudes, dispositions and habits of our minds. This is what Shelby’s account is lacking.

If ideologies are not limited to shared cognitive failings and ideology critique does not solely consist in showing the irrationality of shared beliefs, we need a different, less cognitivist perspective. In this respect, a broadened account of the components of culturally shaped ideologies is needed. As Haslanger says, ‘ideology is not primarily a psychological phenomenon, but is also a cultural one; therefore, a change of culture, and not just a change of individual attitudes, is required to achieve social justice’ (Haslanger 2017a, 7). Rather than beginning from discrete individual minds, the starting point of ideology critique is now located in various social practices, which are ‘collective solutions to coordination or access problems with respect to a resource’ (Haslanger 2017c, 20). ⁸ Various cultural schema can be embodied or instantiated in the relevant material resources. For instance: ‘the schema of two sex categories is manifested in the design and labeling of toilet facilities’ (Haslanger 2012, 415). ⁹ In turn, those resources function to stabilize and reproduce the very schema of binary genders.

According to Haslanger’s practice-first approach, although culture and material resources interact with each other in a sort of looping effect, social practices are dependent on ideology to practically orient individual agents. Ideologies should be understood as cultural schemas, which are ‘clusters of culturally shared mental states and processes, including concepts, attitudes, dispositions, and such, that enable us to interpret and organize information and coordinate action, thought, and affect’ (Haslanger 2017c, 21). ¹⁰ Although ideology is learned through socialization, ‘in learning an ideology, we don’t just form a set of beliefs; there is a sense in which we learn a form of life’ (Haslanger 2017c, 38). That is, ideology is about the lived experience of social agents who spontaneously perceive the surrounding social world as natural and normal. ¹¹ For instance, consider the division of labour in the family. Under sexist ideology, the gendered division of labour appears inevitable, natural and normal. For this division to work, sophisticated propositional knowledge of the inferiority of women is not required. Rather, unconscious and habitual conformity to certain cultural schemas as know-how is sufficient for sustaining practices of caring labour and housework. ¹² As Haslanger writes, ‘Living together requires social fluency, skills for interpretation, interaction and coordination that we exercise “unthinkingly”’ (Haslanger 2017b, 158). Cultural schemas provide social agents with the tools for a degree of regularity and patterns of actions. Thus, ‘Although schemas are variable and evolve across time and context, their elements are sticky and resist epistemic updating’ (Haslanger 2017c, 21). Given the centrality of social practices over individual action, reasoned advice concerning particular actions may miss the point. As Haslanger states,

Insofar as my action is called for by a practice, the pros and cons of this particular choice to act are set aside.… And because we are typically fluent, ‘unthinking,’ in the social practices of our milieu… what is missing from the cognitivist account is how these attitudes are connected to our unthinking responses, our bodily comportment, the social and material realities that constitute our milieu. (Haslanger 2017a, 13)

The point is that while there is room for rational revision in social practices, most social practices are implicit, opaque, informal and unintended. This is not to deny that there are some rationalizable potentials of social practices but to call for perceptual, conative and affective shifts in cognition as well. Given that it is very hard for an individual to change entrenched practice, the key task of ideology critique is to collectively change problematic cultural schemas. Here, Haslanger argues that: ‘broad social change requires change on multiple levels: change to agents, change to culture, and change to structures, policies, and laws’ (Haslanger 2017a, 15).

When it comes to Shelby’s stipulation of racism as ideology, it is notable that explaining the persistence of racism by reference to racism as ideology is circular. Even if ideologies are oppressive, not all oppressive relations are ideological. Instead of ideological state apparatuses (ISAs), consider what Louis Althusser (2014) calls repressive state apparatuses (RSAs) such as army and prisons, which work through merely coercive measures. Although ideological and repressive oppression usually go hand in hand, there are instances of racism, which are not necessarily explicable in terms of racist ideology. To avoid the danger of merely begging the question of persisting racism, there should be distinctions between explanandum and explanans. In this vein, insofar as ideology is part of the larger web of social practices, Haslanger proposes that we think of racism as ideological formations, rather than an ideology per se. By ideological formations, she means ‘the unjust practices, institutions, behavior, and other artifacts guided or formed by an ideology are ideological formations’ (Haslanger 2017a, 16). ¹³

In turn, racism as ideological formations is connected to ‘an interconnected web of unjust social practices that unjustly disadvantage certain groups, such as residential segregation, police brutality, biased hiring and wage inequity, and educational disadvantage’ (Haslanger 2017a, 16–17). An advantage of this view is that it perceives racial inequality as a systematic phenomenon that can be characterized by dynamic homoeostasis (Mallon 2003). As dynamic homoeostasis, ideological formation is sufficiently stable to reproduce while simultaneously being changeable at the same time. For Haslanger, conceiving ideology in terms of culture is decisive because culture is a site of transformation as well as reproduction. If ideology is a causal factor in the complex of social mechanisms, Haslanger’s definition of racism as ideological formation may be useful for considering it to be dependent on other factors, such as material resources. Therefore, her account of ideology is ‘materialist because the source and structure of a discursive/conceptual frame depends on the complex network of social relations that organizes our relationship to things of (assumed) value’ (Haslanger 2017c, 25). This is also consistent with the reason why I favour the functionalist approach of ideology over the epistemic one.

It is noteworthy that Haslanger considers social movements and contentious politics to be decisive in critiquing harmful ideologies. In other words, social movements are primary agents of ideology critique in Haslanger’s view. The point is that effective social movements, rather than reasoned arguments, can create new experiences to disrupt ideologies. Although I agree that social movements are valuable for critiquing ideologies, I believe Haslanger steps back from her insight when appealing to the epistemology of moral truths. To observe this, it is important to note that Haslanger draws on what she calls nonideal moral epistemology to motivate social movements and thereby critique ideology. Although it is inevitable to detour any normative theory of morality to critique ideology, she argues against ideal theories because ‘it is not necessary to know what justice is, or have a complete moral theory, to engage in critique’ (Haslanger 2017c, 41). For the sake of the argument, Haslanger introduces two assumptions. One assumption is that we are reliably capable of having basic knowledge of good and justice. ¹⁴ Her presuppositions of non-relative moral truths are, among others, that slavery, genocide and rape are morally wrong. ¹⁵ Another assumption is that moral knowledge is gained through interactions with others. For instance, we learn that slavery is wrong from others. Similar to standpoint epistemologies, taken together, Haslanger suggests a kind of epistemic humility: ‘We should listen to those directly affected by the practices in question because they are likely to have better access to morally relevant facts’ (Haslanger 2017b, 166). ¹⁶

In light of this nonideal moral epistemology, Haslanger ultimately defends social movements as a paradigm case of contentious politics, thereby disrupting the very social practices we inhabit: ‘not by offering reasons, nor by rational discussion, but by queering our language, playing with meanings, and monkey-wrenching or otherwise shifting the material conditions that support our tutored dispositions’ (Haslanger 2017a, 10). To summarize, drawing on Elizabeth Anderson’s pragmatist view of moral progress, Haslanger favours radical moral changes: ‘A crucial step in disrupting ideology is to create experiential breaks that allow for (and often depend on) the creation of new and potentially emancipatory concepts and other tools for thinking, feeling, and acting’ (Haslanger 2017a, 11).

Haslanger’s ambiguity and nonideal moral knowledge

Although I largely agree with her proposal, there are unresolved problems with Haslanger’s view. First, it is unclear whether this introduction of nonideal moral epistemology actually helps her theory of ideology. If the aim of ideology is to turn the actuality of voluntary servitude into the appearance of non-servitude and thus prevent the oppressed from recognizing the oppression, is it then viable to appeal to some obvious moral truths to critique ideology? Granted, we need not have a complete theory of justice for social critique. Nonetheless, the idea of moral truths can be controversial insofar as there may be a reasonable but substantial degree of disagreement concerning what counts as ideological, especially if such moral truths are not about gross injustices, such as slavery, genocide and rape. Haslanger’s suggestion of epistemic humility may not help either because the very point of ideology is that the oppressed do not necessarily perceive the moral wrong better than others do. Attempting to resolve the problem of ideology by referencing standpoint epistemology (listening to ‘those directly affected by the practices in question’) only pushes back the initial question of ideology if there is no privileged epistemic position for ideology critique in the first place. Even if one has to avoid a moral scepticism and have normative resources for ideology critique (Haslanger 2020, 116), such a privileged epistemic position, if any, should be achieved, rather than presumed.

Even if those who are oppressed are likely to have some epistemic advantages over the oppressor, there is no a priori guarantee of epistemic privilege. Given this danger of circularity, the epistemic position of the oppressed by itself does not secure the epistemic privilege. When Charles Mills formulates the notion of white ignorance as a group-based cognitive handicap, note that he is quick to add that ‘white ignorance is not indefeasible’ (Mills 2017b, 58). As he writes:

The ‘white’ in ‘white ignorance’ does not mean that it has to be confined to white people. Indeed … it will often be shared by nonwhites to a greater or lesser extent because of the power relations and patterns of ideological hegemony involved (This is a familiar point from the Marxist and feminist traditions – working-class conservatives, ‘male-identified’ women, endorsing right-wing and sexist ideologies against their interests.). (Mills 2007, 22)

For Haslanger’s moral epistemology to work, the scope of ideology should be very narrow. As a result, in relying on some putative moral truths, it is difficult to see how Haslanger’s position ultimately differs from Shelby’s cognitivist view. In her earlier paper, ‘But Mom, Crop-Tops are Cute! Social Knowledge, Social Structure, and Ideology Critique’, Haslanger focused on this problem of moral epistemology. Consider the following dialogue. When the Daughter says ‘crop-tops are cute; and I don’t want be a dork’, the Parents reply that ‘Crop-tops are too revealing…. Crop-tops are not cute’ (Haslanger 2012, 408). One may suggest that the daughter perceives the crop-tops to be cute, whereas their parents do not have the same perception. But this disagreement over fashion norms is not resolved by stating that it is simply a matter of subjective and aesthetic tastes. Rather, for the Parents, who are constrained by parental norms, the issue is critiquing the very social framework that constitutes the meaning of cute/dork. Note that the Parents and Daughter belong to different social milieus with different social knowledge. One suggestion is to expose the Daughter to an alternative social milieu other than her own, thereby broadening her perspective and destabilizing it. Yet, this may not be sufficient. Drawing on the common ground between different social positions, Haslanger’s proposes:

To say that a critique is genuine … is not to say that it is the final word; rather, it is to say that a response is called for…. For Parents to have a critique of Daughter’s choices, they should offer more than a flat denial of her claim relative to their milieu; it is their responsibility to seek common ground from which Daughter can assess their critique. If Parents can find common ground with Daughter and their claim that crop-tops are not cute is true relative to that common ground, then because Daughter shares that ground, she must address Parent’s concern; hopefully, the two sides will continue to engage until they reach a mutually acceptable common ground. (Haslanger 2012, 425)

Although it is unobjectionable to think that genuine ideology critique rests on some kind of common ground, it is far from clear whether we can achieve ideology critique by forming a common milieu. Especially when common ground is symmetrical between conflicting positions, there should be further criteria for favouring certain normative stances at some point. Even though Halsanger suggests formal criteria for a superior common ground, such as the absence of coercion or violence (Haslanger 2012, 426), this is too narrow to be a substantial guide. Note that the point of ideology is to make de facto coercion or violence de jure consent. In the case of crop-tops, their cuteness, as dictated by relevant social groups, is the very point of discrepancy and incommensurability. Which view is emancipatory, rather than illusory? The Daughter may argue that crop-tops can be empowering, rather than oppressive. Even if the Parents are right, how can they convince the Daughter? It seems that Daughter and Parents have fundamentally irreconcilable interests. In this case, we are back to the issue of nonideal moral epistemology.

This appeal to the nonideal moral epistemology is also reflected in the fact that Haslanger remains ambiguous about the normative features of ideology. When critiquing Shelby’s cognitivist account of ideology, recall that Haslanger defines ideology as a subset of cultural technēs. Notice that in Haslanger’s view, ideology is deemed as descriptive, normatively neutral and less pejorative because culture as such is mostly inescapable and not necessarily subordinating. Although she believes that a sociopolitical dimension of ideology critique is indispensable, ideology as cultural technēs need not be necessarily morally problematic or just. As she admits, ‘we cannot live together without ideologies to guide us. Although some ideologies are pernicious and partly constitute unjust social structures, improved ideologies are crucial in order to achieve social justice’ (Haslanger 2014, 386). ¹⁷ Clearly, she wants to ‘avoid normative overreach’ (Haslanger 2017b, 192) because a moral critique may be too abstract or inefficacious.

Simultaneously, she is explicit in stating that ideology is ‘a cultural technē gone wrong’ (Haslanger 2017b, 159). Haslanger writes: ‘Some technēs … sustain, good and just forms of coordination. Ideologies, however, are technēs that produce or sustain injustice by guiding us to enact unjust practices. If racism is an ideology, in this sense, then it partly constitutes social practices that give people reason to act in racist ways’ (Haslanger 2017a, 16). This is because ‘human coordination is, in principle, valuable; we also know that not all forms of human coordination are morally acceptable’ (Haslanger 2017c, 46). A cultural technē becomes ideological by virtue of its unjust outcomes, which are to be evaluated according to some theory of justice and morality (Haslanger 2017c, 39).

At this point, Haslanger moves back and forth between these two positions. ¹⁸ Whether a cultural technē as a set of social meanings is ideological or not may depend on the context. To determine whether cultural technē is ideological, in any case, Haslanger must appeal to her version of nonideal moral epistemology (‘a mutually acceptable common ground’), which is dubious. Although she convincingly argues against Shelby’s cognitivist view of racist ideology, it is unclear how Haslanger’s account serves to better connects the cognitive dimension and the functional dimension of ideology, simply by adopting different terms, such as epistemic dimension and normative dimensions (or reasoning and valuing) of ideology.

Radicalizing pragmatist critique

As noted in the previous section, an important aspect of my critical remarks on Haslanger is that she curiously steps back from her key insights on ideology as cultural schemes when introducing nonideal moral knowledge. Although it is nonideal, Haslanger holds on to the notion of moral knowledge, which is supposed to illuminate those who are in the grip of ideology. However, in my perspective, she is better off if she abandons the view that having nonideal moral knowledge would help us to overcome oppressive ideologies. Although intentional agents have moral ideas and usually act in accordance with them, because of the influence of various – sometimes competing – ideological social practices, they oftentimes cannot act accordingly, despite knowing what is moral. Given the incentive structures behind the social norms-based approach I defend, the availability of nonideal moral knowledge may be necessary but insufficient. This is because nonideal moral knowledge and enabling individuals to deviate from harmful social norms are vastly different. Therefore, one cannot simply refer to nonideal moral knowledge to disrupt ideology. In other words, although the discursive superstructure has moral components, the functional base involves social and psychological motives, desires and needs (Jost et al. 2013, 233). As an intersection between individual minds and social structures, the social norms approach I advocate is concerned with effectively communicating and signalling beneficial social norms rather than sticking to the idea of moral knowledge.

It may be noteworthy that my view is notably pluralistic and pragmatic, compared with Haslanger’s version of the standpoint epistemology in encompassing various ways of learning experiences. For instance, although it is true that we learn from social movements, there are multiple methods of learning new social norms and unlearning harmful ones. When social movements transform harmful social practices, this is not only a matter of creating new epistemic possibilities and moral knowledge. Rather, social movements change people’s incentive structures, as well as the perception of ordinary life related to problematic ideologies. My social norms-based approach is not vulnerable to the problem of nonideal moral knowledge insofar as social norms are not just a matter of knowing what is just or not but also having psychosocial habits of perceiving and feeling norms’ functionality.

Although I find Haslanger’s materialist view of ideology as ideological formation unobjectionable, it is my claim that her insight should be pushed more. Especially, Haslanger’s pragmatic conception of ideology as cultural technēs should be developed further, rather than simply appealing to alleged moral truths. In a similar vein of pragmatism, I reject absolutist and monist approaches in ideology critique in favour of contextual and pluralist approaches. The pragmatist view is also fallible and open-ended. To the extent that what matters in ideology is our ordinary social interaction, we should be cautious about the supposed privilege and authority of critical theorists who may be too paternalistic and uncharitable towards their subjects. As far as it is implausible to suppose that we live under the spell of ideology as an inescapable and total illusion, we must give due respect for the basic epistemic agency of the oppressed. ¹⁹ Moreover, even disengaged philosophers and social critics are not immune from ideology. A relevant point of critical theory would be that despite the gap between the standpoint of third-person observers/critics and first-person participants, these standpoints should be properly mediated in the task of social critique. Although critique and its object are kept separate in traditional theory, ideology critique in critical theory should be reflexive, immanent and self-referential (Geuss 1981; Ng 2015). If we are not merely passive victims of ideology, then there must be some already-existing resources in the social actors’ ordinary experiences and self-understanding to critique ideology. However, this should not commit to the status quo. Instead, what we need is a critique without authoritarianism. In critiquing ideology, we require ‘the practice of critique without presupposing a privileged epistemic position and a break with ordinary practices of justification’ (Celikates 2006, 35). ²⁰ When a single ideology cannot be thoroughly hegemonic and may involve differing presuppositions, ideology critique should be immanent, revealing inconsistencies, disharmonies and paradoxes within a particular ideology, without appealing to an external standpoint. ²¹

Of course, ideology critique in Shelby or Haslanger’s manners is a valuable task. Yet, we should bear in mind that ideology critique is always already embedded in various sociocultural mechanisms. Although I think that Haslanger’s view of racism as ideological formation vaguely captures this point, she does not provide a more detailed account. Whether it be fixing cognitive failings (Shelby) or relying on moral epistemology (Haslanger), knowing epistemic or moral truth does not by itself overcome ideology. Ideologies are not simply negative and constraining but also positive and enabling. In this regard, I emphasize the ambivalence of ideology as such and its instrumentality.

In my view, Haslanger seems to underestimate the paradox that ideologies are supposed to be simultaneously normative and functional (Jaeggi 2008; Prinz and Rossi 2017). As Rahel Jaeggi writes, when we say that ideologies suffer from a practical contradiction, it is ‘characterized by the fact that the obstacles or crises that are part of it are normatively problematic in both senses: something does not work (well), and the way it works is not good’ (Jaeggi 2008, 78). This partly explains morally bad ideologies tend to be functionally bad as well. Although no society can be maintained by brute violence, we should note that there is no strict correspondence between the normative and the functional in ideologies. For instance, some ideologies can be moral but may still not work well. Conversely, other ideologies may be functional while being immoral. While all ideologies are more or less unstable and contestable, the point is that the normative (‘the way it works is not good’) and the functional component (‘something does not work well’) are closely interconnected in the case of successful ideologies.

Insofar as ideologies are functional as well as normative from the broader perspective of the social, morality may be only one good among other valuable goods. Therefore, ideology critique can be normatively important, but it should not be normativistic and moralistic, such that it appeals to some abstract ought as external and permanent normative criteria. ²² In this regard, Haslanger’s mode of ideology critique is not an exception. A moralistic understanding of ideology is not feasible if it ignores the ways in which ideologies may not include immoral elements while supporting domination and oppression. Alternatively, ideologies can be valuable because certain ideologies can be mobilized to promote morally admirable aims because, as Haslanger herself passingly admits, ‘Improved ideologies are crucial in order to achieve social justice’ (Haslanger 2014, 386).

Given this variability of ideologies, rather than focusing on ideology critique by reference to external epistemic or moral truths, my point is that we need both analysis and critique as an immanent critique of ideology. ²³ To discern what ideologies can be harmful or valuable in particular situations, ideology critics should be attentive to the social realities while carefully assessing their normative merits. When someone says that ‘inequalities, even material inequalities, give rise to flawed ideological beliefs’ (Stanley 2015, 216), it may be only trivially true ever since Aristotle. To capture the processes of socialization and transmission of ideology, however, what is at stake is understanding the causal mechanisms of ideology, rather than simply drawing on moral intuitions about it. ²⁴

Ideology as harmful social norms

As far as I consider my work to be a friendly critique of Shelby and Haslanger’s theory of ideology, I have attempted to expand and radicalize their insights such as functional and nonepistemic aspects of ideology, cultural schemes and social practices. In particular, the idea of ideology as cultural schemes elucidates implicit, opaque, informal and unintended aspects of social practices. Although my view of ideology as a harmful social norm is similar to Haslanger’s view of ideology in that regard, unlike her, the view of harm does not assume that any particular moral theory actually functions as the solution to the problem of ideology. Moreover, perceiving ideology as social norms allows us to see the functional (nonmoral) aspects of ideology, which requires the realist manners of critique.

In the remainder of this article, I will illustrate the importance of understanding the causal mechanisms of ideology in reference to social norms. It is worth noticing that I understand that the theory of ideology and ideology critique are closely intertwined. This aspect is another crucial feature of the realist and pragmatic view of ideology critique. As is clear in the critical theory since Max Horkheimer and Theodor Adorno, critical social theory requires empirical social sciences as well as normative political theory. For instance, although we can learn from many contemporary philosophers, realist ideology critique requires interdisciplinary studies, such as those involving social psychology, sociology or the rational choice theory as well. ²⁵

If we make a distinction between the moralist critiques (Shelby and Haslanger) and social critiques I have advocated, the latter cannot be merely an epistemic project because understanding concrete social mechanisms embodied in various social practices is indispensable for social critique. Ideologies tend to be stable and recalcitrant because ideological formations are multifaceted clusters of informal convention, norms and practices. Insofar as ideologies are connected to people’s minds and habits, understanding psychological mechanisms concerning implicit biases, norm acquisition and norm execution is also crucial. Therefore, intervening in harmful social norms requires the detailed analysis of how targeted social norms in particular societies specifically function to coordinate their social lives.

In my view, the ambivalence of ideology would be better and more simply understood when ideology is considered in terms of harmful social norms. In doing so, I will argue that Haslanger’s nonideal moral epistemology should be connected to strategic and pragmatic considerations of collective action concerning individual and group interests. Consider her examples of traffic laws and norms (Haslanger 2017c, 34). Driving on the right side of the road is one kind of traffic coordination but that is not the only and true answer. What matters here is not its epistemic element but its functional aspect. Social norms are related to a kind of intersubjective and pragmatic problem-solving process, which is itself embedded in the process of learning and social evolution. In this sense, social norms are ordinary, conventional and changeable. Akin to Haslanger’s cultural technēs, Bicchieri also discusses schemata and scripts (Bicchieri 2006, 93–94). They function to cognitively support social arrangements and related social norms. There may be beneficial and harmful kinds of social norms, but the function of social norms on our social life is not different from that of cultural technēs. Culture and social norms not only constrain but also enable social agents, such that we can interpret and understand what is valuable and salient in the social world. This is possible via determining what kind of options and preferences are available and appropriate to agents. ²⁶

Although ideology is a part of various constituents of social norms, ideology focuses on the cognitive and justificatory elements of social norms. As noted above, however, this does not mean that critiquing ideology is a matter of rational revision of distorting representations of the social world. Although it can be argued that the relation between social norms and cultural technēs is unclear in this respect, it should be noted that social norms always involve various values, technologies and norms, which are irreducible to beliefs or statements. This is why adopting the framework of ideology as harmful social norms does not necessarily take us back to a cognitive approach. ²⁷ In accommodating functional and motivational dimensions, conceiving of ideology as a whole interlocking cluster of dynamic social norms with relevant social expectations is a much more comprehensive and explanatorily superior hypothesis. As I argued against Shelby’s postulate of rationality, it seems plausible that ideology is a kind of device that solves a collective action problem wherein agents rationally opt to participate in the patterns of social action that contradict their actual interests. Insofar as the reproduction of oppressive social structures is not necessarily dependent upon collectively false and irrational beliefs, what is at stake is not rational persuasion; rather, the point is to effectively intervene on the underlying social mechanisms to change the norms of ideology. ²⁸ Insofar as ideologies’ function implies their own degree of stability over time, they can function in diverse ways, whether it be legitimating, nonmoral or unintentional.

Importantly, ideologies can be effective only when they are deemed sufficiently credible and reliable enough to their addressees. Although this suggests that brute coercion without any moral validity would not be likely to succeed, even people’s acceptance of certain harmful social practices may be merely prudential, rather than based on good normative considerations, oftentimes driven by the hopelessness that there cannot be any viable alternative other than the status quo. In this regard, the value of social movements is not limited to creating new epistemic possibilities. Although Haslanger’s proposal of social movements as such is not objectionable, she does not properly address that social movements are usually costly and even risky. Hence, the simplistic dichotomy between ideology and social movement is debatable. Even though we are not perfectly self-interested agents, it should be emphasized that we are also both largely instrumentally rational and significantly constrained by others’ expectations and actions. ²⁹ After claiming that we should not mistake a collective action problem for ideology, Joseph Heath writes:

The most significant thing about these collective action problems, from the standpoint of critical theory, is that agents often have a hard time getting out of them, even if they realize that they are engaging in collectively self-defeating behavior. The reason is that the mere recognition that the outcome is suboptimal does not change the incentives that each individual has to act in a way that contributes to it…. It is only if everyone stops that I will begin to see a difference. But I have no control over what everyone else does… in order to change the interaction, everyone has to stop doing what they have been doing. (Heath 2000, 366)

Certain social patterns of action can persist because they are stuck in a suboptimal equilibrium. Even when individuals are perfectly rational in a sense, they can be collectively irrational. In the case of labour movements, a revolution is unlikely to arise because it is overly risky for workers, despite its emancipatory promise. Merely knowing people’s real interests does not mean that those interests immediately become incentives to change existing behaviours. Why is having nonideal moral epistemology insufficient? This is because what matter in ideologies are various incentive structures that sustain injustice. ³⁰ For instance, Heath makes a case with respect to the feminist claim that women should be free from the dependence on the concept of beauty and the beauty industry. The problem is that endorsing this feminist lesson may not directly motivate women’s personal conduct. A collective action problem helps explain this point:

Beauty has an inherently competitive structure… When a 50-year old woman gets a face-lift that makes her look 40, the action can be described in one of two ways. In a sense, she has made herself look younger. But in another sense, all she has done is make all the other 50-year old women in the population look a little older. These women may then be motivated to get a face-lift just to retain position. (Heath 2000, 369)

This practice is maintained not because a face-lift is advantageous but rather because not doing so is disadvantageous. Beauty as a positional good is not the only example. Rather, it is important to see that conforming and defying harmful norms should be understood in light of interests and needs. Again, one may make a distinction between individual interest and group interest. The point is that individuals can oftentimes benefit from going along with oppressive social norms even though those norms would collectively disadvantage those individuals as a group and the individuals in question are well aware of this fact.

If this is the case, it is better to consider ideology as a social norm with its own stability. A primary function of ideology as a sort of convention is to coordinate social agents by reference to shared understandings and social expectations. Given this coordinating function, it is understandable why mere epistemic revision is insufficient to motivate people: a one-sided deviation from a given equilibrium can be very costly. Here, it is notable that the cognitivist assumption of ideology goes hand in hand with the individualist view of social agency. The task of ideology critique should be collective, not individual. When racist practices are in place, a few individual anti-racists can be easily punished and disregarded. Racism can be claimed to be prudentially rational from the perspective of economic reason. For instance, consider ‘the case of a Jewish employer who, while personally not antagonistic toward Arabs, refuses to hire Arab personnel because all businesses in the community have been threatened by boycott should they employ Arabs’ (Goldberg 1990, 383). Ideology is embedded in strategic situations where the consequence of one’s actions depends on others’ expectations. Insofar as ideologies are seen as self-maintaining and persistent conventions, they are ‘solutions to coordination games with multiple equilibria’ (Guala 2016, 44). When an ideological equilibrium is not desirable, it is important to convince people that there can be multiple equilibria other than the current one. However, something more is required as well. The questions of order, safety and trust are as decisive as the legitimacy and morality of certain social norms. ³¹ This is why we need a realist and contextualist standpoint concerning ideology critique (Prinz and Rossi 2017). ³²

Here, Bicchieri’s example of female genital mutilation (FGM) illustrates why ideology is not merely dissolved by invoking its illegitimacy, immorality or even outcomes like health risks. The practice of FGM cannot be understood without seeing the contextual value of female purity, chastity and family honour since social norms are ‘embedded in a thick web of values, beliefs, and other norms that form shared cognitive schemata’ (Bicchieri 2017, 121). In turn, due to a correlation between high levels of poverty and inequality, valuation of virginity as a desirable trait of a young and modest bride is also related to the vital need to find a rich husband. Surely, discussion and deliberation may convince people that justice should be prioritized over honour since ‘protection of one’s child (in the sense of avoiding harming him or her) carries more weight than the need for purity, as attained through female genital cutting’ (Bicchieri 2017, 79). Even if the value of honour is maintained, people can be convinced that ‘honor is not necessarily best protected by female genital cutting’ (Bicchieri 2017, 126). However, when it is a pervasive practice, although FGM is a costly way to control fidelity, many are motivated to imitate others and continually engage in such a practice.

The practice of FGM persists as far as there are no known possible alternatives and, as in the collective action problem, being the first mover is very costly. Bicchieri explains the persistence of harmful social norms nobody likes by referring to pluralistic ignorance: ‘Individuals may not believe that the particular practice is positive or even desirable, but a lack of transparent communication and the observation that the practice is widespread makes them conclude that it is widely supported. Here, people might have good reasons to change, but they do not actively voice them’ (Bicchieri 2017, 122). ³³ According to this, even if we know that the purity norm behind FGM is wrong and unjust, deviation from the practice requires more than knowing its disadvantages. Initiating deviation is costly due to the pluralistic ignorance.

What is striking is that despite its historical persistence, the practice of FGM was rather rapidly abandoned. As Bicchieri states, ‘the observation of a few, highly schema-discrepant instances can induce a sudden, dramatic revision of the associated schema’ (Bicchieri 2017, 137). Abandoning a social schema is possible because of a trendsetter who is a relevant role model in the reference group. It may also occur due to continuous exposure to an alternative culture with different social norms. To this end, a norm shift requires advertising a new frame that ‘linked traditional values of honor and purity to the idea that uncut girls are complete and pure’ (Bicchieri 2017, 139–40). Yet, we need assurance as well as proper knowledge. In addition to various educational campaigns about the harmful health effects of FGM, there should be some effective measures to carefully coordinate a norm shift, thereby reducing first mover costs. ³⁴

Taking this into consideration, the task of ideology critique should accompany a proper account of the moral and social psychology of ideology. Ideology is a kind of motivated cognition, according to John Jost and his colleagues, since ‘ideologies possess both a discursive (socially constructed) superstructure and a functional (or motivational) base or substructure’ (Jost et al. 2013, 233). Even though Shelby and Haslanger examine the discursive superstructure of ideology as a bundle of beliefs, attitudes and values, both end up underestimating the functional substructure of ideology, such as social and psychological motives, desires and needs. For instance, in the political realm, what Dan Kahan calls ‘identity protective cognition’ may be one example of motivated reasoning. ³⁵ This functional structure is worth emphasizing since we are inescapably dependent upon relevant others. On this view, ideology is a motivation of system justification:

The tendency to defend, bolster, and justify the status quo is motivationally compelling because it satisfies epistemic needs to attain certainty, order, and structure; existential needs to maintain safety and security and to minimize danger and threat; and relational needs to affiliate with others and to acquire a sense of belongingness and shared reality. (Jost et al. 2013, 236)

This does not mean that ideologies inherently justify systems and support the status quo. Rather, ideologies can be also liberal or challenging to the extent that people who endorse them are open to new experiences, cognitive complexity and tolerance of uncertainty. Haslanger’s nonideal moral epistemology is committed to this as well. As noted above in our discussion of FGM, the crucial point is to adequately motivate people and address their epistemic, existential and relational needs in specific situations. Returning to the issue of racism as ideology, although I disagree with Dummett that racism is inherently irrational, his following statement can be better understood in light of ideology as social norm: ‘People adopt racist attitudes when they find themselves in a social milieu in which it is acceptable or encouraged to display such attitudes, and in which maintaining them is advantageous’ (Dummett 2004, 33). If my view is right, ‘By giving persons incentives to act in ways counter to oppressive social norms, they will choose to do so, and perhaps choose to do so in better ways than could have been devised from those farther removed from the problem (Cudd 2006, 230)’.

Conclusion

So far, I have critically examined Shelby and Haslanger’s respective views of ideology. Despite their merits, although Shelby’s view of ideology is too cognitive, Haslanger’s appeal to nonideal moral epistemology limits her insight that ideology serves as cultural technēs. In short, contrary to Shelby and Haslanger’s view, ideology is neither simply irrational nor merely immoral. However, both Shelby and Haslanger primarily perceive ideology as a kind of epistemic flaws in a broad sense. In either approach, the functional aspect of ideology is not taken seriously enough. In light of this, I have also argued that thinking of ideology as a social norm is more comprehensive and less cognitivist in accommodating motivational and functional aspects of ideology critique. Ideology critique can be successful when we consider the functional and motivational points, in addition to the epistemic ones. Such a view is more amenable to potential social changes and learning.