Recognition, power, and trust: Epistemic structural account of ideological recognition

Recognition is one of the most elusive and ambivalent concepts in political and social thought. In recent studies of the ambivalence of recognition (Ikäheimo et al., 2021; Lepold, 2019; McQueen, 2015), recognition has the emancipatory aspect: Recognition is a necessary condition for individual freedom by forming a social basis of self-worth, and the struggle for recognition plays the significant role in political movements for emancipation. However, recognition has a dominating aspect: The demand for recognition can be exploited as an instrument for domination, reproducing existing problematic practices and identities. For example, sweatshops induce employees to voluntarily subjugate themselves to harsh working conditions by praising the employees’ self-dedicated character and enhancing their self-worth.

In recent philosophical debates, Axel Honneth has developed the most systematic theory of recognition. He discusses the problem of ambivalence in offering the concept of ideological recognition (Honneth, 2007). Honneth's argument consists of two steps. In the first step, he defines ideological recognition as distinct from misrecognition. Misrecognition occurs when addressees believe their subjective self-image is not consistent with the recognition they receive. They feel misrecognized when their self-worth is inflicted. By contrast, the addressees have “good reason to accept” ideological recognition because they attain a stronger sense of self-worth through the recognition (p. 341). However, “ideological forms of recognition suffer a second-level rationality deficit” as it encourages the addressees’ willing subjection to the dominant social order (p. 346). This suggests that ideological recognition, an issue of “second-level rationality,” is judged independently from the addressees’ subjective perspective. Ideological recognition can be defined as that accepted by the addressees from their subjective point of view, but unjustified from an objective or theoretical point of view.

In the second step, Honneth proposes a substantive standard of ideological recognition, a standard of “how we can draw a distinction between justified and unjustified forms of social recognition” from an objective point of view (p. 340). According to Honneth, recognition is ideological when it maintains the addressees’ self-worth, while the evaluative promise expressed by the recognition cannot be materially instantiated. In the example above, recognizing the self-dedicated employees in the sweatshops is ideological and unjustified, for the sweatshops will not guarantee material and economic conditions for realizing the employees’ dedication to the company (not providing a minimum income level, for instance).

I argue against Honneth's substantive standard, not the conceptual definition of ideological recognition itself. It is because his standard is not broad enough to capture the ideological recognition particularly observed in feminist studies. The issue is that some women demand recognition of their femininity. As long as the received recognition heightens their self-worth, they do not suffer misrecognition. However, desiring recognition for their femininity can subject women to a certain injustice. As the recent debates on epistemic injustice illustrate, recognizing feminine gender identity can result in exclusion from the epistemic community where people in everyday communication cooperate to pool information (Fricker, 2007). Examined from an objective point of view, women can suffer epistemic exclusion through the recognition they demanded for themselves. By Honneth's definition, subjectively demanded recognition of femininity should be called ideological recognition as these women are willingly subjected to the dominant gender order. Nevertheless, Honneth's standard of ideological recognition cannot be applied to the recognition of femininity because there is no lack of material means for women to realize feminine qualities. In other words, his account considers the recognition of women who voluntarily reproduce gender stereotypes as morally justified.

In this respect, Honneth's account faces two problems: normative and descriptive. The former is that his account does not provide a substantive normative standard that can identify the unjustifiable ideological recognition above mentioned. The second problem is that he cannot objectively explain why the addressees of ideological recognition accept it as expressing positive qualities. Honneth (2007) claims that a sufficient account of ideology should uncover its “regulative power” (p. 344). If this is the case, an analysis of power that encourages the internalization of gender ideology is needed.

This article offers an alternative account of ideological recognition called an epistemic structural account. The epistemic structural account overcomes the normative problem of Honneth's account by elaborating on the concept of epistemic trust. This concept provides the basic norm of trust that allows us to identify unjustified ideological recognition through which women suffer epistemic injustice as a result of seeking recognition of their feminine qualities. My account also overcomes the descriptive problem by providing the concept of structural power to analyze the issue of internalization of problematic identities. This analysis illustrates how structural power is exercised to create a situation in which it is favorable for addressees to internalize a quality as the basis of their existential self-worth.

Although there have been fruitful studies of trust in social epistemology and pragmatics (e.g., Brandom, 2019; Jones, 1996) and a recent debate on the structural dimension of power (e.g., Forst, 2015; Heyward, 2018), the concepts of epistemic trust and structural power have not been associated with recognition. I integrate the two concepts into a theory of recognition to develop an account of ideological recognition. My account claims that recognition is ideological when the following two conditions are met. First, the basic norm of epistemic trust is violated. Second, the recognition nevertheless maintains the subjective self-worth of the addressees because they accept the recognition due to the operations of structural power.

Beyond criticizing Honneth's account, the epistemic structural account has several implications for critical theory by examining the interconnectedness between epistemic trust and structural power in recognition practices. First, my account advances the task of criticizing social pathology, a central target in critical theory, to epistemic and structural dimensions.¹ Second, my claim that the theory of recognition should consider structural power in epistemic practices develops the discussion of the relationship between recognition and power.² Third, I analyze domination and oppression through recognition from perspectives of feminism and critical race theory. I focus mainly on ideological recognition in the case of gender, but address the case of racial identity. Fourth, my account provides a practical critique of the ambivalence of recognition because it plays the role not only of normative guidance for emancipation but also of a descriptive disclosure of power dynamics.

The rest of this article proceeds as follows. I first examine the two problems in Honneth's account of ideological recognition (Section 2). To overcome these problems, I then elaborate on two concepts of the epistemic structural account: epistemic trust (Section 3) and structural power (Section 4). From these arguments, I formulate the epistemic structural account of ideological recognition and discuss its theoretical and practical benefits for critical theory (Section 5).

In this section, I examine Honneth's account of ideological recognition and specify its problems by giving examples illustrating the ambivalence of recognition surrounding women. I criticize his account for not developing sufficiently the argument for explaining ideological recognition beyond the subjective perspective of the addressees.

Honneth describes ideological recognition in two steps. In the first step, he defines it by distinguishing between misrecognition and ideological recognition. The former is determined from a subjective point of view, and the latter from an objective or theoretical point of view.

Misrecognition implies that subjects feel that their expectations for recognition are betrayed (Honneth, 1995). In Honneth's analysis, the modern subject shapes these expectations in three forms of recognition—love, legal respect, and social esteem—whereby each subject recognizes valuable qualities, such as personal intimacy, equal rights, and economic achievement. The subject experiences misrecognition if the qualities that accord with one's self-image are not recognized by others. Moreover, misrecognition violates a precondition for freedom by psychologically damaging the subject's “positive relation-to-self” through which one can be sure of the social value of one's identity (Honneth, 1995, pp. 79, 174). Along with the three forms of recognition, the subject forms three types of positive relations-to-self: self-confidence, self-respect, and self-esteem, which are conditions for pursuing one's self-chosen life goals without psychological inhibitions and fears. Therefore, addressees of recognition judge it as acceptable or misrecognition from their subjective point of view, based on whether positive relations-to-self are acquired. When they feel misrecognized, they are motivated to a “struggle for recognition” to recover a damaged relation-to-self (Honneth, 1995, p. 138).

In contrast to misrecognition, ideological recognition is determined independently of the addressees’ perspective. Honneth argues that the value-statements that ideological recognition expresses seem “positive,” “credible,” and “contrastive” in the eyes of the addressees (Honneth, 2007, pp. 337−339). These adjectives indicate that ideological recognition can serve the self-realization of the addressees by heightening their relations-to-self and giving them a “sense of being distinguished” (Honneth, 2007, p. 339). When a value-statement includes discriminatory or banal qualities, the addressees feel misrecognized and have no good reason to accept it.

In the second step, Honneth proposes a substantive standard of ideological recognition by focusing on the gap between “evaluative promise and its material fulfillment” (Honneth, 2007, p. 346). In ideological recognition, the symbolic premise of recognition is not accompanied by material fulfillment, such as institutional and material means to realize certain valuable qualities. He exemplifies these qualities in the entrepreneurship of workers recognized by capitalist corporations that praise entrepreneurship as a quality that all workers should have, and workers enhance their self-worth by trying to acquire it. However, not every worker can institutionally achieve it. The morally unjust nature of ideological recognition, as Honneth claims, lies not in the symbolic value-statement per se but in the way it is institutionally realized. From an objective point of view, the recognition of entrepreneurship is a privilege that only a few can realize, but advertised as if everyone has the chance to achieve it.

I argue against Honneth's substantive standard of ideological recognition because it cannot apply to the recognition of ideological gender identity. His account relies exclusively on the subjective perspective of the addressees as to the justifiability of recognition as long as its value-statements are realized. To challenge Honneth's account, Amy Allen (2010) gives an example of a 5-year-old girl who is dependent on her parents’ authority and accepts gender ideology to receive recognition from her parents (pp. 25−26). Although Allen is on the right track in highlighting the recognition of femininity in the asymmetrical relationship between parent and child, I suggest that the issue of internalizing femininity also applies to adult women who accept individualistic values. To illustrate the problem, consider the following two women. I leave aside ethnic or intersectional cases; however, I discuss in Section 5 the claim that racial identity is also relevant to ideological recognition.

Thus examined, Honneth's account has two shortcomings. First, his account cannot elucidate why it is morally unjustified to recognize Suzanne's identity but morally justified to recognize Brianna's because he does not provide normative arguments. The original question of ideological recognition is to identify recognition accepted from a subjective point of view but unjustified from an objective point of view. Honneth considers ideology in terms of the gap between value and reality. However, the objective point of view in his account involves only the way a value is realized, not the value itself. Ultimately, his account does not contain a normative standard to distinguish between morally justified and unjustified recognition. His theory lacks a normative argument regarding the morally unjustified nature of ideological recognition.⁴

Second, his concept of ideology is too narrow to understand the ambivalent situation of Suzanne descriptively. Suzanne accepted mainstream “girlish” qualities as a positive self-image, rather than rejecting them as restrictive. However, from an objective or theoretical point of view, the recognition she received may disadvantage her in terms of epistemic trust (as discussed in the next section). A descriptive claim is necessary here, explaining why she accepted gender norms and demanded recognition for femininity even if she valued individualism. Honneth's theory of recognition lacks a descriptive argument about how Suzanne reproduced gender stereotypes while not feeling misrecognized.

So far, I have argued that Honneth's account of ideological recognition fails to explain, first, what constitutes the substantive normative standard for delineating between morally justified recognition and unjustified ideological recognition and, second, why problematic qualities are internalized as positive. In the following sections, I provide an alternative, epistemic structural account of ideological recognition that overcomes Honneth's two problems. As for the normative problem, the concept of epistemic trust contains the basic norm of trust that allows us to identify the objective injustice that Honneth's account cannot capture. Regarding the descriptive problem, I explain the internalization of problematic identities by analyzing the operations of structural power.

In this section, I elaborate on the concept of epistemic trust. Epistemic trust refers to a basic interpersonal attitude that underlies the three forms of recognition in Honneth's theory, that is, recognition as a knower who contributes to epistemic cooperation.⁵ This concept provides the basic norm of trust, which includes a substantive standard to delineate between morally justified and unjustified recognition. The recognition in the case of Suzanne was unjustified because it violated the basic norm: She could be mistrusted. I discuss the basic norm of epistemic trust by applying discussions of feminist epistemology and pragmatics of trust (Brandom, 2019; Daukas, 2006) to normative arguments.

What is trust? In a general context, trust is, like recognition, a reactive attitude that involves normative expectations such that the trustee moves directly and favorably toward the trustor (Baier, 1986). Trust is also an affective attitude in that the trustor harbors moral resentment when the trustee betrays these expectations (Jones, 1996). In the epistemic context, trust is a condition of epistemic cooperation, for we have to trust someone who is better placed to know something relevant to our interests, given the limitations of our epistemic capacities.

I consider epistemic trust a basic or primitive form of recognition in which the trustor recognizes the trustee as a knower who has the epistemic status and authority to partake in relevant epistemic practices. We cannot collect and produce knowledge through communication unless we assume that a speaker is trustworthy, that is, she has a certain epistemic status and authority to influence our knowledge. The following discussions of trust overlap with Habermas's formal pragmatics. Habermas (1984) elucidates the condition of intersubjective agreement. However, my account further develops the conditions of discursive practice in terms of intersubjective recognition and normative expectations. The concept of epistemic trust involves a form of recognition that must be expressed before embarking on the acts of reaching understanding.

What, then, is the epistemic status or authority that must be recognized by others when one participates in epistemic practices? To illustrate this point, Daukas (2006) considers the presumption of trustworthiness as necessary for productive epistemic exchange. First, in assigning trustworthiness to a speaker, we must assume that the speaker is “sincere” and “competent” in the domain of the person's testimony (p. 110).⁶ When, for example, I ask a stranger for directions, I presume that she is not lying and is familiar with the area. Second, a successful exchange of testimony requires the presumption that the speaker is neither excessively “diffident” nor “self-confident” (pp. 112−113). If I think the stranger behaved in an overly confident manner, I would not see her as trustworthy even though she is, in fact, not lying and familiar with this area. I trust her as a knower insofar as she has an accurate sense of her epistemic competence. Therefore, epistemic trust contains the basic norm that a speaker should have the default epistemic status of being sincere, competent, and sensible of their own competence. Unless there is reasonable ground to withhold their epistemic trust, we should recognize the speaker as having the default epistemic status.

The recognition of Suzanne, who demanded to be recognized as “girlish,” is unjustified because her identity could preclude her from being recognized as having a default epistemic status. Suzanne made a schoolgirl error to gain the favor of the men around her. However, her behavior and received recognition could disadvantage her in terms of epistemic trust. She may suffer an undue withholding of epistemic trust, such as not being delegated an important business meeting by her colleagues. Such mistreatment amounts to what Fricker (2007) calls “testimonial injustice” (p. 17). Suzanne's epistemic trust could be unjustly undermined because women are traditionally prejudiced by stereotypes, such as being “unintelligent” or “irrational,” a stereotype that has been codified in gender norms. Due to the identity of being “girlish,” the default sincerity, competency, and self-sensibility could be unequally attributed to Suzanne from the outset.

Furthermore, the epistemic trust that conditions participation in epistemic practices is not limited to the default status, for the necessary condition of normal discursive communication involves not only the attribution of epistemic status but also the illocutionary performance of the uttered words. Thus, the basic norm of epistemic trust also regulates performative forces. In this context, Brandom (2019) considers the concept of epistemic trust as recognizing the other as a member of the epistemic “community” whose members share language and conceptual norms (p. 529). Epistemic trust means recognizing one's authority to use a concept properly and to evaluate the correct applications of conceptual norms. By way of illustration, Suzanne, a manager, ordered a worker: “Finish printing by 12 p.m.!” The workers recognized her authority to perform a speech act of order and followed her when they took her to use the conceptual norms correctly. If she, for example, mistook noon for midnight, violating the conceptual norms about time, the workers would have withheld their epistemic trust. Epistemic trust entails the basic norm that we should recognize a speaker as a member of the epistemic community unless there is reasonable ground to withhold trust.

The epistemic authority in the dimension of performative force is relevant because feminine gender identity can distort the performative force of women's speech acts. As Kukla (2014) notes, even if the same words are uttered, gender differences can affect “the social uptake” of the words because social images of women can distort the audience's perception of the speaker's intention (p. 443). This happens particularly when a woman enters a domain typically considered men's. Consider that Suzanne issued an order at a printing factory, where most of the workers were men, in a grammatically correct way. However, her male colleagues could be less compliant with her order. Even though she lived up to the linguistic conventions that would typically mark her speech acts as orders, her male workers perceived her as issuing “requests” instead (p. 446). As the proper response to a request being granted is gratitude and she, as a manager, was not in a position to show gratitude to her workers, she appeared rude, and her workers could become disobedient. In this sense, Suzanne's “girlish” quality can disadvantage her by preventing the equal application of conceptual norms and distorting the performative forces of speech act.

To sum up, epistemic trust is the underlying form of recognition as a knower who has the default epistemic status of being sincere, competent, and sensible about one's competence, as well as the default epistemic authority to use concepts and evaluate the correct application of conceptual norms. The basic norm of epistemic trust is that we must recognize each other as having such a default epistemic status and authority. This norm delineates between justified and unjustified recognition. Even though the addressees accept recognition where it is consistent with their self-image and maintains their relations-to-self, such recognition is unjustified only if it results in a violation of the basic norm of epistemic trust, that is, mistrust.⁷ Those who are mistrusted as not having the default epistemic status and authority also suffer epistemic injustice, in which their testimony is not fairly heard, or they are excluded from the knowledge-producing practice. The basic norm of trust plays the role of a minimal normative standard that determines whether the accepted recognition from the subjective point of view is justified from the objective point of view.

The basic norm of epistemic trust is grounded in the fact that it is a precondition for epistemic cooperation and social freedom. Unless trusted as knowers, we cannot express and create our ideas through the use of language. Feminist movements, for example, have sought to deconstruct stereotypical identities through the redefinition and appropriation of what it means to be a woman or “queer” (Butler, 1997). The politics of “resignification” challenges hegemonic conceptual uses and attempts to create a climate in which the oppressed express their own identities as a part of free self-realization. For the success of such politics, even if subjects create a novel vocabulary that deviates from everyday language use, they must be recognized as knowers who try to apply conceptual norms creatively and produce knowledge. In this sense, being recognized as a knower conditions “positive linguistic expressive freedom” (Brandom, 2019, p. 520). Such freedom is socially complemented because people are dependent on others’ recognition of their epistemic status and authority.

This normative argument elucidates why it is morally unjustified to recognize Suzanne's identity and morally justified to recognize Brianna's. In the case of Suzanne, she was recognized as “girlish,” but because of that recognition she could be mistrusted. Her recognition reproduces gender stereotypes that place women in situations that make them vulnerable to epistemic injustice. Such recognition is unjustified, as well as ideological to the extent that she accepted it in herself. By contrast, Brianna was recognized as “girlish,” but also an “empowered” woman who has the autonomous subjectivity to challenge gender stereotypes. Her self-image was not compatible with the “charming” character that Suzanne desired for herself. Rather, Brianna demanded epistemic trust for participating in knowledge-producing practices for deconstructing gender stereotypes. Thus, the recognition of Brianna's identity is justified because her recognition satisfies the minimal normative standard of epistemic trust.

Moreover, my account can clearly define ideological recognition by classifying misrecognition along the subjective and objective dimensions. The unjust withholding of trust can be objectively observed in relation to epistemic status and authority. In this case, a speaker is attributed a level of credibility that does not match their sincerity, competence, and sense of competence, or a speaker's speech act does not have performative force due to their identity despite the proper applications of conceptual norms. In this sense, I call mistrust objective misrecognition, which occurs even when subjects do not feel misrecognized. Ideological recognition can be understood as objective misrecognition in which the addressees do not feel subjectively misrecognized.

The distinction between subjective and objective misrecognition can further clarify the two cases. Both Suzanne and Brianna could be mistrusted insofar as there remain social prejudices against women. In that instance, however, the two receive different types of misrecognition. When Suzanne was mistrusted, she was not subjectively misrecognized but might receive objective misrecognition in terms of epistemic trust. Subjectively, she may not have been aware of it as injustice because she accepted the “girlish” quality as the basis of self-worth. We should call such recognition, which is not subjectively but objectively misrecognized, ideological recognition. Unlike Suzanne, Brianna demanded recognition for her “empowered” quality; she may feel subjectively misrecognized when her rework of gender stereotypes is rejected by the mainstream. When mistrusted, she suffered both subjective and objective misrecognition in the form of epistemic trust.

Thus far, my account has attempted to delineate between justified and unjustified recognition. However, it is important to note that my account never argues that Suzanne was entirely morally wrong in desiring objectively unjustified recognition or that she was wholly responsible for being mistrusted. Instead, the epistemic structural account explains the ideological nature of the unjustified recognition Suzanne received by focusing on structural power, which can induce voluntary subordination to femininity.

In this section, I propose the concept of structural power to answer the descriptive question: Why do the addressees accept recognition that is unjustified from an objective or theoretical point of view? I outline this possibility and discuss the importance of structural power for the analysis of ideological recognition, and then focus on the two operations of structural power that induce the internalization of social norms.

Structural power refers to the power of social structures—sets of social norms, rules, and laws—to situate a person in particular social roles and positions by constituting patterned actions and social meanings. Wartenberg (1992) discusses structural power in comparison to agential power. Agential power refers to the capacity of a powerful agent to intervene in the conduct of the powerless agent. For example, a high school teacher might exercise power over her students by threatening to lower their grades if they do not work harder. In contrast, the teacher's agential power over students is “constituted” by social structures because the teacher–student relationship in the education system allows her to put pressure on her students by means of a grade (p. 82). Social structures have a higher order power to condition the teacher's agential power by constituting the social role of the teacher and the social meanings of grades. Such social meanings are not only determined by the education system in high school alone but also by other institutions, such as firms that affirm the authority of the teacher's grading by rejecting applications from students with lower grades. Accordingly, the teacher's agential power over her students is constituted and reproduced by complex coordination among the “peripheral social agents” external to the mentioned agents (p. 90).

Recognition is also conditioned by structural power because social structures assign people with certain traits to specific social positions and establish social expectations that someone in a certain social position will behave in ways relevant to that position (Heyward, 2018). Social norms play a major role when recognition is routinized as social practice. For instance, a teacher expects students to work hard and shows disappointment in the case of betrayal. She may recognize students’ achievements by referring to the implicit norms of diligence in the classroom. Likewise, other agents, such as firms, recognize a student's valuable traits in a job interview based on the student's grades or the capacity for teamwork. Social norms condition a patterned perception of evaluative qualities, from which subjects learn how to respond to valuable qualities that matter in a relevant social practice. In this way, social norms situate people in advantageous or disadvantageous social positions and establish social expectations through recognition.

The importance of considering structural power is made clear by examining the agent-centered concept of power that Honneth implicitly premises. He views power exclusively as the agential power of social groups to “control these means of symbolic force” and impose a particular cultural interpretation on other groups (Honneth, 1991, p. 303; 1995, p. 127). For him, power is the ability of powerful agents to impose an identity on powerless actors by controlling cultural resources, leading to their misrecognition. With this agent-centered concept, he discusses power relations only from the subjective point of view of the two agents. Honneth's agent-centered concept reflects recent investigations of power, which is represented in Steven Lukes's definition: “A exercises power over B when A affects B in a manner contrary to B's interests” (Lukes, 2005, p. 30). This concept supposes that Subject A has the intention to dominate B in ways that restrict B's freedom or influence B's interests. Lukes assumes that power is exclusively possessed by agents and is reluctant to think that social structures themselves have power; otherwise, the theory of power cannot identify powerful agents and “fix responsibility for consequences held to flow from the action, or inaction, of certain specifiable agents” (Lukes, 2005, p. 58).

Nonetheless, the agent-centered concept does not help much in answering our question about ideological recognition. Consider the case of Suzanne. She internalized gender stereotypes even though her peers might not have intentionally aimed to impose them. Rather, Suzanne desired to be recognized as “girlish” in acquiring her relations-to-self. Her desire was not necessarily influenced by her colleagues and powerful group alone, but rather by social norms about gender. In these cases, it is too narrow to consider her recognition only in terms of relationships between agents. To analyze the voluntary subordination to gender norms, we should take an objective point of view that examines social structures.

However, structural power by definition unavoidably involves any act of recognition because recognition is always practiced within a social structure. If so, how can we understand the internalization of problematic identities, as in the case of Suzanne? To illustrate this point, I concentrate on the adaptive and constructing operations of structural power. These operations explain the way ideological recognition is accepted from the addressees’ subjective point of view.

I have expounded the concepts of epistemic trust and structural power that are components of my account. By examining the interconnectedness between epistemic trust and structural power, I propose my substantive standard of ideological recognition briefly and discuss its theoretical and practical benefits in expanding the scope of domination and critique.

If we rest on the concepts of epistemic trust and structural power that I have discussed in the previous sections, recognition is ideological when the following two conditions are met. First, the basic norm of epistemic trust is violated. Second, recognition nevertheless maintains the subjective self-worth of the addressees because they accept recognition due to the operations of structural power. Ideological recognition is unjustified because the addressees are mistrusted as knowers in such a way that their recognized identities distort others’ judgment of their epistemic status and authority. Subjectively, the addressees feel suitably recognized and their desire for recognition is satisfied because the operations of structural power create a situation where they positively accept ideological recognition. However, objectively, they are misrecognized in a basic form of recognition, regardless of whether they are clearly aware of the injustice.

The structural epistemic account overcomes the normative and descriptive problems of Honneth's account. For the first problem, Honneth does not actually provide a normative argument to identify ideological recognition. The concept of epistemic trust provides the basic norm of trust and identifies objectively unjustified recognition as mistrust. Regarding the second problem, Honneth cannot descriptively explain why the addressees of ideological recognition accept their identities as having positive qualities. I explain the problem of internalizing problematic identities by analyzing the operations of structural power. Honneth assumes that the internalization of gender norms has been an anachronistic topic. This is partly true, given the impact of feminist philosophy on social movements. However, my account attempts to address the persistent injustices in the structural and epistemic dimensions, which cannot be grasped by a subjective sense of misrecognition or an interpersonal analysis of power relations.

Beyond criticizing Honneth's account, the epistemic structural account has theoretical and practical benefits for critical theory by examining the complementary relationship between epistemic mistrust and the internalization of problematic identities. I will argue that my account not only advances the analysis of social pathology (5.1) and power dynamics in recognition (5.2) but can be utilized to discuss racial identities (5.3) and the task of practical critique (5.4).

In this article, I have argued that recognition theory must consider the ambivalence of recognition because a desire for recognition is sometimes exploited as an instrument for reproducing domination. The central question in this article is how to identify ideological recognition, recognition that is acceptable from the addressees’ subjective point of view but morally unjustified from the objective or theoretical point of view.

Honneth's answer, as I pointed out, fails to offer a substantive normative standard for identifying the recognition of ideological gender identity, thereby ignoring the injustice that the addressees suffer even if they are recognized as consistent with their self-image. His account also fails to descriptively address the structural dimension of power that forces the internalization of gender norms and stereotypes.

The epistemic structural account that I have offered discusses, first, the normative argument of delineating between justified and unjustified recognition by proposing the concept of epistemic trust. It is the underlying form of recognition and includes the basic norm of epistemic trust that allows us to identify objectively unjustified recognition as mistrust. Second, my account described how the addressees accept a problematic identity by proposing the concept of structural power. Structural power can create a situation where it is favorable for the addressees to internalize problematic social norms and identities to maintain their self-worth. From these arguments, I have claimed that recognition is ideological when the following conditions are met. First, the basic norm of epistemic trust is violated. Second, recognition maintains the subjective self-worth of the addressees because they accept the recognition due to the operations of structural power.

Further, I have argued that the epistemic structural account has theoretical and practical benefits for analyzing domination and advancing the task of critique. By examining the interconnectedness between epistemic trust and structural power, my account develops not only the critical analysis of social pathology, but also analyzes the power dynamics in recognition by inquiring into the triadic relationship between social structures, the privileged, and the oppressed. I also tried to show that my account can be applied to analyze the ideological recognition of African American youths and can be used in practical critiques for overcoming epistemic injustice. Such critical tasks, of course, require further theoretical and empirical studies.¹² However, developing an analysis of ideological recognition can advance critical theory for emancipation.