The Epistemic Fata Morgana: Appropriation in the Institutional Context

Abstract

In this article, I identify a conceptually distinct form of epistemic appropriation: the creation and proliferation of the epistemic fata morgana. An epistemic fata morgana is a hermeneutical resource that is hollowed out, stripped of its meaning and political power, and yet, posited as if it were still accessible. This resource is taken up by dominant knowers in a way that preserves only its perception, but not access to it. This process is illustrated by an examination of the resource “sexual harassment” within the university context. The epistemic fata morgana is an important addition to the field of epistemic injustice for it lends itself to highlighting the frustration that is felt by marginalized people, especially within institutional contexts.

An experience of sexism and misogyny becomes more difficult to process because to recognize what is going on can mean giving up a belief in an organization (Ahmed 2021, 66)

In this article, I describe a conceptually distinct form of epistemic appropriation: the creation and proliferation of the epistemic fata morgana. I argue that an epistemic fata morgana is a hermeneutical resource originally created by marginalized people to describe their experiences of oppression, which is taken up by dominant groups in a way that harms the marginalized group who created the resource but conceals this harm. In this article, I define hermeneutical resources as follows: hermeneutical resources are resources that are used for making sense of and evaluating our experiences, they can be things such as language, concepts, and standards with which we evaluate our own experiences (Pohlhaus 2012, 718). Furthermore, these resources do not exist independently; they are dependent on the people who use them to make sense of their experiences.

The presence of the epistemic fata morgana is a distinct form of epistemic appropriation—epistemic appropriation being defined by Davis (2018) as wrongful epistemic uptake that obscures meaning or origin of the epistemic resource—because in inter-communal uptake, the resource has been appropriated in a way that hollows it out, changes it, but masks the fact that this has happened. This resource thus appears to be available for marginalized people to describe their experiences—and maybe even to use it for (political) action—but, in fact, the resource is inaccessible for this purpose. In this sense, it is a fata morgana, it is—like a mirage of an oasis in the desert for the starving traveler—something that is perceived, but not really there.

The epistemic fata morgana is an important addition to the field of epistemic injustice, or, more specifically, epistemic appropriation, for it lends itself to highlighting and explaining the frustration that is felt by marginalized people, especially within institutional contexts. Although an organization might portray itself as progressive, this does not necessarily have to mean that it is a hospitable environment for marginalized people that come into contact with it. This sentiment, in the context of the university, has been discussed by many, most notably by Sara Ahmed (Ahmed 2012, 2021; see also Saxe 2022). However, this has not been discussed explicitly in the field of epistemic injustice.

In order to illustrate how an epistemic fata morgana comes into being, I elaborate upon four characteristics of the fata morgana. I do this by showing them in one example: sexual harassment in the university context. Universities pledge themselves to fighting sexual harassment, but it seems to be increasingly difficult to speak of it and “do something about it” within the context (Ahmed 2021, 2012).¹ I therefore argue that “sexual harassment” can be an epistemic fata morgana. I elaborate on this when discussing all four characteristics and applying them to this example, giving an overview of the genesis of such an epistemic fata morgana. I limit my scope to the university context, even though the dynamics outlined in this research could be applied to other (institutional) contexts. I do so for two reasons: first, as a woman in academia, I have experienced this phenomenon, which inspired me to write about it. Secondly, the institutional, specifically university, context is quite explicit in its presumed norms and values. There are mission statements, diversity policies, and complaint procedures that acknowledge (structural) injustices and commit the university to fighting these injustices, which makes it possible to see and illuminate the mechanisms at work more clearly.

The four characteristics of the phenomenon are mentioned briefly here and explained in the following four sections of the article. First, an epistemic fata morgana is a hermeneutical resource that is originally created by marginalized people to make sense of their own experience, but shared inter-communally, meaning that outside of the community of origin it has traveled (Lewis 2013; Salem 2018). Secondly, this hermeneutical resource describes something about which there is consensus in the dominant discourse that it is bad or undesirable (e.g., racism, sexism). The third characteristic is the fundamental incompatibility of the hermeneutical resource with the epistemological system of the dominant group. The fourth and last characteristic is that, while truly taking up the hermeneutical resource is not compatible with the dominant group’s epistemological system, being perceived as taking up the resource is compatible.

An epistemic fata morgana is a hermeneutical resource that displays these four characteristics. In the case of sexual harassment, this means that effectively utilizing the term becomes increasingly difficult for those who wish to use it within the university context for any kind of liberatory aim. Although the term is institutionalized within complaint procedures, these procedures are often emotionally draining and ineffective (Ahmed 2021). Furthermore, using the term can have adverse consequences because its use is often deemed inappropriate. This can have negative emotional and social consequences for those who choose to utilize it. There are also distinct epistemic consequences: testimonial smothering—or self-silencing—and oppressive double binds (Bayruns Garcia 2019; Berenstain 2016; Dotson 2011). These two consequences are further discussed in the last section of this paper.

1. First characteristic: marginalization and inter-communal uptake

The first characteristic of an epistemic fata morgana is that said hermeneutical resource was conceived of by marginalized people to describe their own experiences and later shared inter-communally.² There are three components to this, namely: creation of the resource, distribution of the resource, and the environment in which it is used. Hermeneutical resources are created in any kind of community, dominant or marginalized, since they are resources that are used to make sense of experiences. However, the focus of this article is specifically on hermeneutical resources that are created by marginalized groups to describe or understand their own experiences, most notably the oppression they face. These specific resources often remedy a kind of hermeneutical injustice.

Hermeneutical injustice is defined by Miranda Fricker as “the injustice of having some significant area of one’s social experience obscured from collective understanding owing to a structural identity prejudice in the collective hermeneutical resource” (Fricker 2007, 155). Hermeneutical injustice, in this definition, is faced by marginalized knowers when they lack the resources to understand and/or communicate their own experiences. This definition has often been understood to suggest that hermeneutical resources can be created and then evenly distributed throughout dominant and marginalized communities. There is, in this view, a conceptual gap, a lack in resources.

Over the years, this initial definition of hermeneutical injustice has been expanded upon, as there seem to be different kinds of hermeneutical injustice, specifically when due attention is paid to the way in which hermeneutical resources—once they come into existence—are distributed and used (Dotson 2011, 2014, 2012; Pohlhaus 2012). For example, Fricker (2016) herself outlines maximal, minimal, and midway cases of hermeneutical injustice. In maximal cases, such as the case of Carmita Wood that Fricker uses in her 2007 book, an individual is “not in a position to make proper sense of her own experience even to herself” (6). In a minimal case, an individual can make perfect sense of their own experience, but is unable to communicate it to others, as presumably others do not possess the same hermeneutical resources. In this minimal case the problem is distribution, not conception. Fricker argues that midway cases are those that exist in between these maximal and minimal cases of hermeneutical injustice, where certain communities share hermeneutical practices and resources that are “not shared across further social space” (8).

Trystan Goetze (2018) builds upon this clarification by Fricker. They outline two kinds of harms that result from hermeneutical injustice, cognitive and communicative harms (78). Individuals suffer cognitive harms when they are unable to make sense of their own experiences, like in Fricker’s maximal case of hermeneutical injustice. They suffer communicative harms if they can understand their own experiences, but not effectively communicate this to others. Goetze argues that these harms are instances of the same harm, namely that “the subject has some distinctive and important social experience that at some crucial moment lacks intelligibility” (79). In this article I am interested in epistemic appropriation, which takes place when hermeneutical resources are distributed. These harms can be both communicative and cognitive.

The distribution of hermeneutical resources and the kinds of hermeneutical injustice that can result from this has been discussed often in the years following Fricker’s 2007 book, notably by incorporating the ways in which this dissemination is influenced by prevailing power relations (Pohlhaus 2012; Dotson 2012). Davis (2018), for example, argues that societal power relations influence the way in which epistemic resources are created, distributed, and taken up, meaning that there is a chance that these resources will be used in ways that benefit the powerful in a society. This could easily result in hermeneutical resources being conceived in marginalized communities, but not taken up in dominant ones, because it does not benefit dominant communities. Furthermore, resources can be widely used in marginalized communities before they reach dominant communities. A conceptual gap can therefore be not a lacuna within a marginalized community, but a lack of inter-communal uptake, appearing as a conceptual gap to dominant knowers. For example: in the classic sexual harassment example, women of color and feminist women conceived of the resource long before it entered the mainstream, and was picked up by other groups, such as men (Berenstain 2020; Fricker 2016, 2007). Sometimes, then, marginalized, resistant, knowers develop their own hermeneutical resources to make sense of the oppression they experience at the hands of the dominant communities (Pohlhaus 2012). These resources are thus developed intra-communally but can be shared inter-communally.³ The process of inter-communal uptake is influenced by prevailing power relations between marginalized and dominant knowers. Hermeneutical resources might be developed and shared, but dominant communities might refuse to take up these resources, to use them, because these very resources highlight their wrongdoing (Dotson 2012; Pohlhaus 2012). The refusal to use alternative hermeneutical resources, however, is not the only way in which hermeneutical uptake is subject to prevailing power relations. Hermeneutical uptake can also go awry when there is not a refusal to take up a resource, but appropriation of the resource.

Epistemic appropriation, as noted, can be defined as epistemic uptake that is wrongful, in the sense that it obscures or changes part of the resource (e.g., its origin, its meaning) (Davis 2018). There are multiple kinds of epistemic appropriation, most notably resource-preserving and resource-obscuring epistemic appropriation (Podosky 2023). Resource-preserving epistemic appropriation occurs when epistemic resources are detached from the community that created them and can be misdirected: used in ways that benefit the powerful, sometimes by crediting this resource not to marginalized knowers, but dominant ones, concealing the resource’s origin (Davis 2018, 713). An example of this is research that is conducted by academics in elite universities on disadvantaged schools. It is the participation of the research subjects that creates the knowledge, but it is credited to the researchers (detachment). This prestigious research then benefits the researchers and the university, but not the people responsible for the knowledge that is created (misdirection) (713). Most often these research subjects get next to nothing in return for their participation.

Another form of epistemic appropriation is resource-obscuring epistemic appropriation. As opposed to resource-preserving epistemic appropriation, which does not change the resource that is appropriated, resource-obscuring appropriation changes the conceptions attached to a resource.⁴ In this case, when a hermeneutical resource reaches a dominant community, this leads to this resource and the conceptions originally attached to it being obscured and changed. Therefore, the people who developed it “are not able to convey understanding across social space—especially to those whom such understanding is needed” (Podosky 2023, 145). The resource gains differing conceptions, hindering inter-communal communication. When the resource is appropriated by a dominant community, the resource is “undermined as a means of gaining understanding of the world, and as a means of communicating understanding across social space” (149). This means, in the event of resource-obscuring appropriation of a resource that is used to describe oppression faced by marginalized knowers, that this resource cannot be effectively used to communicate this oppression to dominant knowers.⁵

An example of resource-obscuring appropriation might be one of the things that has happened to the term “woke.”⁶ Originally used by Black Americans to signal awareness of social justice issues, white communities have taken up the term, and the associations with the term have changed (Podosky 2023, 151). Now, instead of referring to awareness of social issues, the term is often used to mean general “awareness,” of “anti-feminist warnings, celebrity gossip, the dangers of soft drink and razors, and contempt for progressives” (151). The use of the term, in this context, has shifted completely and seems unrelated, or even opposite, to its previous use. The original conception is thus obscured, the term is used differently in marginalized and dominant communities. There is a problem then, in the inter-communal uptake of resources that maintains a deficiency in resources (154). This deficiency is not the product of the absence of a certain resource, a lacuna. It is an “attempt to fill in a lacuna that has gone awry” (154).

As the distribution of a hermeneutical resource is important, so is the environment in which it is used. It is possible for everybody to have access to hermeneutical resources, but for these resources not to be used, as this depends on the context. Crerar (2016) illustrates this when discussing the menstrual taboo. Even though most people possess the resources to meaningfully discuss menstruation, there are environments in which doing so is not possible. These environments are expressively unfree (205). The environment in which individuals communicate matters, as institutional or certain social contexts, can be either free or unfree.

In the introduction to this article, I suggested that “sexual harassment” has become an epistemic fata morgana in the university context. I contend that the first characteristic—development within marginalized communities resulting in inter-communal uptake—is clearly visible. The concept was first conceived by feminists to describe their experiences, so it originated in a marginalized community, or communities (Berenstain 2020; Fricker 2007, 2016). Furthermore, the resource has most definitely been shared inter-communally, as it is a widely used term now. Within the specific context of the university, there are policies that aim to fight sexual harassment and other forms of gender-based harassment and violence (Ahmed 2021). In this sense, it has not just been shared inter-communally—it has been taken up to the extent that it has become institutionalized.

In this section, I have highlighted the first characteristic of an epistemic fata morgana: the resource is created by marginalized communities to describe, interpret their own experiences, and is shared inter-communally. I have outlined what can go wrong in inter-communal uptake, including refusal to take up the resource, obscuring the source of the resource but keeping the initial associations, and lastly, changing the original conceptions attached to a resource, obscuring it. Since the epistemic fata morgana is taken up by dominant communities, but something goes wrong in this uptake, it qualifies as epistemic appropriation. However, neither resource-preserving nor resource-obscuring appropriation describe the phenomenon fully, as it incorporates elements of both. The epistemic fata morgana is similar to resource-preserving epistemic appropriation in the sense that the resource seems to not have been obscured—it appears to be available; sexual harassment, at least to an extent, seems to keep the same meaning.⁷ However, this resource cannot be used for inter-communal communication anymore, making it similar to resource-obscuring appropriation. It seems to be available, it is visible, but it is not available for use, making it distinct from the other forms of epistemic appropriation outlined here. I contend this is due to the second characteristic: the phenomenon described by the hermeneutical resource is undesirable, bad, or otherwise important to fight.

2. Second characteristic: undesirability

The second characteristic is that the phenomenon described by the hermeneutical resource that becomes an epistemic fata morgana is deemed as negative, bad, or otherwise undesirable in dominant discourse. A way to gauge this is to see to what extent these discourses are entrenched within institutions. If there is a consensus about undesirability, the phenomena described by hermeneutical resources that have the potential of becoming epistemic fata morganas might be illegal, or otherwise sanctioned. For example, racism, discrimination on the basis of race, is illegal under Dutch law (Government of the Netherlands n.d.). This is a good indication of the fact that racism is regarded as undesirable. Therefore, combating racism is often seen as somehow virtuous, it reflects well upon the person or institution or person doing it.⁸ Fighting an -ism, or at least, being perceived as fighting this -ism, benefits an institution when this -ism is seen as undesirable.

Unfortunately, the institutionalization of a hermeneutical resource does not automatically lead to social progress, nor does the problem immediately disappear. An example of this is outlined by Falbo (2022), writing about the criminal case of Chanel Miller and Brock Turner, where Turner raped Miller, but despite overwhelming evidence, his sentencing was relatively light (349). The rape happened at a university party and Turner was apprehended by witnesses and arrested immediately. There was little to no doubt that he indeed assaulted Miller; instead the conversation focused on whether he should face the consequences (Miller 2019). In this situation, the act Turner committed was deemed undesirable and thus institutionalized—it was a crime—but nonetheless he was punished very lightly. This might have happened because there were competing hermeneutical resources to describe Turner, namely “rapist” and “golden boy.” Falbo argues that these frames competed, and “golden boy” won out. “Rapist” was seen as an inappropriate way to describe Turner and this frame, this resource, was not utilized effectively in this case, because it could not be applied to Turner (350–51).⁹ Therefore, even though the resource (rapist) was institutionalized (and penalized), the resource was not fully accessible, either because of an overabundance of competing hermeneutical resource, or, as I would argue, because it is an epistemic fata morgana. In acknowledging that a hermeneutical resource can be institutionalized as undesirable, but not used within this context, it is possible to separate two elements of hermeneutical resources: perception and access. A resource can be perceived, institutionalized, but be limited in use, in access, like the resource “rapist” in the Turner case.

It is important here to clarify what I mean by access and perception. A way to do this is by discussing the difference between manifest and operative concepts, as outlined by Haslanger (2006). The manifest concept is “the concept I take myself to be applying,” where the operative concept is “the concept that best captures the distinction as I draw it in practice” (99). The manifest concept is what the concept is supposed to capture, the operative concept is what it actually captures. Within the context of sexual harassment in the university, the manifest concept is the definition of sexual harassment as institutionalized by the university, in policy, the perception, and the operative concept is the understanding that is held by students and employees, the concept that is accessed. Katharine Jenkins (2017) takes this distinction and applies it within the field of hermeneutical injustice. Jenkins argues that rape and domestic abuse myths constitute hermeneutical injustice. Because of rape and domestic abuse myths, victims of it seem to know the manifest concept, the official definition of rape, of domestic abuse, but fail to apply this to their own situations, because the operative concept is influenced by these myths. Here, there is a case in which hermeneutical resources are widely available, but they are not accessible for use, because they seem to be distorted. There is a problem in application.

Something similar is stated by Mona Simion (2019), who argues that hermeneutical injustice can be a case of failure in concept application, not necessarily concept possession (178). In Jenkins’s case, victims of rape and domestic abuse know what these things are, they possess the hermeneutical resources, but they cannot apply them to their own situations. This might then, returning to Goetze (2018), mean that they suffer a cognitive harm, as they cannot understand their own situation as rape, or domestic abuse.

In this article, I am not necessarily focused on whether victims of sexual harassment are able to apply this concept to their own situation, but more so the fact that they suffer communicative harms because of this: even if they know they were sexually harassed, they possess this hermeneutical resource, something is stopping them from applying it, even though the environment in which they communicate seems to be free, as there are extensive policies against sexual harassment.¹⁰ There is something happening here: the hermeneutical resource is institutionalized, it is visible within the university policy, but accessing, using it within this context is very difficult.

The practical situation of accessing “sexual harassment,” utilizing the resource within the university context, is difficult. Reporting sexual harassment often involves filing an official complaint against the institution or the person who harassed you. This places you at the beginning of a terribly exhausting bureaucratic process. To illustrate: Sara Ahmed (2021) interviewed people at all points of complaint procedures and found that complaint procedures were often experienced differently by those going through them: they were experienced as opposite to how they are portrayed by the university: difficult, inhospitable, vague, not healing, open, accessible. The experience was draining, complicated, and the procedures were inaccessible to them, and it was hard to find out what the actual procedure was. Complaint procedures often take place behind closed doors, which means that the visible part of the procedure is solely the portrayal by the university, an institution that portrays itself as caring about social safety. This perception of the university as a progressive institution that cares turned out to be just that: perception (Ahmed 2021). One of Ahmed’s interviewees even “described the new procedures [complaint procedures] and policies as ‘window dressing’,” allowing the university to portray itself as “progressive” with regards to sexual harassment policies, without having to take action to face the structural problems within the institution. Relating this to Haslanger’s manifest and operative concepts, then, it seems to be the case that the manifest concept of sexual harassment in the university differs from the operative concept, as it becomes clear that practically very few and only certain cases of sexual harassment receive the consequences that are seemingly attached to it. It is then the case that some people can see this gap between the two concepts; they can see that the hermeneutical resource that they hold, the resource that is portrayed, and the resource that is “put into practice” all differ. The hermeneutical resource is perceived within the institution, as this posited as a progressive institution, but it is not easily accessed in this form. In order to elaborate on this, I return to Ahmed who conceptualizes certain speech acts within this context.

Ahmed argues that the university’s apparent commitment to diversity comes about by speech acts, by speaking it into existence (Ahmed 2012, 59). The line of thinking might go: diversity is good, and should therefore be furthered, sexual harassment is bad and should therefore be fought. We, the university, commit to this publicly. This is then a speech act, an act that ought to bring about the effect they name. However, I have highlighted that institutionalizing these hermeneutical resources does not necessarily mean access to them. From this, then, it can be assumed that mere spoken commitment does not necessarily have to mean that diversity will be furthered, or sexual harassment will be fought.¹¹

The acknowledgment that spoken commitment does not necessarily lead to action allows us to question the performativity of speech acts. Speech acts can also be non-performative: they can not produce the effect they name. Therefore, verbally committing the institution to fighting sexual harassment can not commit the university to fighting harassment. A commitment to eradicating sexual harassment made by a university official can be considered non-performative when it does not work to bring about this effect. Notably, this is different from a simple failure of the speech act to reproduce the effect it is naming: the failure is the goal of the speech act. A performative speech act can fail in bringing about its goal, but remain performative. Non-performative speech acts have the goal of not bringing about the effect they name. Ahmed states that “such speech acts [commitment to diversity] are taken up as if they are performatives (as if they have brought about the effects they name), such that the names come to stand in for the effects. As a result, naming can be a way of not bringing something into effect” (Ahmed 2012, 117, emphasis mine).

Regarding sexual harassment, by institutionalizing this hermeneutical resource, it can be used for non-performative speech acts in the sense that its institutionalization and uttering gives the university enough license to not have to make it a performative speech act. For a speech act to be performative, it must be iterative, it has to be repeated in its performance. Simply put, commitment to fighting sexual harassment must be repeated, which would include actual accessible complaint procedures and preventative measures. This is not the case for non-performatives, since then “the past that accumulates [not fighting sexual harassment] overrides this futurity [future commitment], as what the institution is committed to, by sheer force of habit” (126–27). Fighting sexual harassment is not repeated, it is not performed. What the university is actually performing is sexism and by uttering non-performative speech acts, they can continue.¹²

Moving into the language of epistemic injustice, combining these elements one might say that hermeneutical resources created by marginalized people, circulated inter-communally, can be taken up within institutional contexts—when dominant discourse has deemed the phenomenon described by this hermeneutical resource undesirable—in a way that furthers only perception, not access. So, the resource can be seen, perceived—within institutions and policies—but not effectively accessed. This further illustrates that this is indeed a distinctive form of epistemic appropriation, for the resource is not changed in a way that leaves it unrecognizable, because it can be perceived—as would happen in resource-preserving epistemic appropriation—but at the same time, it is obscured. The uptake of this hermeneutical resource obscures its access, making it so that the resource cannot be used for its original purpose—describing marginalization and oppression and combating this. In this, structural problems within the institution might also be obscured, as the visibility of the resource can be wielded to validate the institution’s commitment to fighting the said phenomenon. Furthermore, this serves to invalidate any complaints about inaccessibility, because the resource is posited as if it were accessible, effectively silencing any opposition.

Moreover, even uttering the term with the purpose of accessing it can have adverse consequences. To an extent, because it has been institutionalized as undesirable, sexual harassment as a term has been overinflated. It has become something that is perceived as an ultimate wrong, making it rare, as extreme cases seem to be the only ones to adequately make it through painful and convoluted complaint procedures (Ahmed 2021). It is therefore not perceived as something that routinely happens in (patriarchal) institutions. Consequently, it is difficult to say one has been sexually harassed by, for example, a coworker, because the resource has been posited in such a way that only the most textbook cases of sexual harassment are considered to fit this definition—this manifest concept differs from the operative concept—but this is not acknowledged. Thus, the conceptions attached to the term change in the sense that it becomes an overexaggerated version of itself, a simulacrum, or a caricature. The sexual harassment routinely experienced by those who are marginalized within the institution diverges from the sexual harassment described by the university, but because of this very institutionalization that covers up the gap between visibility and access, this cannot be identified by those who do not experience it. Therefore, those who do identify this suffer from communicative harms, even in a situation where the hermeneutical resource is seemingly institutionalized in its original form.

Sexual harassment, in this sense, fits the first two characteristics of the epistemic fata morgana. The resource was developed by marginalized people and taken up inter-communally, deemed as undesirable in a way that has even institutionalized the resource and assigns consequences to it (such as being fired and other possible workplace consequences). However, there is a disconnect between visibility of the resource and access to it. For institutional consequences to be assigned, a complaint procedure is necessary. If this procedure is emotionally draining and dubious, utilizing the resource becomes increasingly difficult. In this sense, sexual harassment as a hermeneutical resource is clearly perceived in the institutional context, but not accessible.

3. Third characteristic: epistemological incompatibility

The third characteristic of an epistemic fata morgana is that of a fundamental incompatibility of the hermeneutical resource in question with the dominant epistemological system within the receiving community in inter-communal uptake. Epistemological system refers to assumptions and foundations for systems of thought. Epistemological systems can be individual—every person has their own system—but they are also shared intra-communally, as most members of a community will share certain truths about their life and their community. Epistemological systems exist in any kind of community, dominant or marginalized, or both. Since the focus of this article is on epistemic appropriation in the institutional context, my emphasis is on communally shared epistemological systems.

Epistemological systems are comprised of dominant, within a community, shared epistemic resources; shared concepts, ideas, truths (Dotson 2014). These resources can themselves be inadequate for explaining the world and making sense of one’s experiences, as the foundational truths within such a system might not be true. However, because of this, most epistemological systems are resilient. As an epistemological system forms the foundations for certain truths to be held, this epistemological system itself becomes truth (131). To change this requires a reflection on the inadequacies of an epistemological system that posits itself as adequate, as that is its very foundation. A process to change an epistemological system therefore requires “recognizing one’s instituted social imaginaries and altering them” (131). In other words, it is necessary for one to first acknowledge that epistemological systems exist, that there are certain social imaginaries that are fundamental to this, and then to recognize that these imaginaries, these resources, can be inadequate. This is an increasingly arduous task, as “the primary difficulty … is the work required to ever acknowledge its existence, which is made difficult due to epistemological resilience itself” (132).

Any information that challenges an epistemological system requires, then, an acknowledgment of the existence and resilience of this epistemological system and a shift in thinking, to be properly received. The information needs to first be acknowledged as being conflicting to be received properly. However, this is difficult because of the very resilience of epistemological systems. Hermeneutical resources that enter an epistemological system are evaluated in the context of this system. If these resources conflict with the foundations of such a system, with what is presumed to be fundamentally true and unquestionable within such a system, they might not be taken up, and so not acknowledged. Or, they might be taken up in a way that changes the hermeneutical resource, that changes the conceptions associated with this resource, in a way that conforms with the epistemological system that they enter. This happens in any kind of community, as epistemological systems appear to be a basic part of the way in which individuals and communities receive and process information. However, my focus here is on epistemological systems that are shared within dominant communities, communities that are more powerful both socially and epistemologically.

Relating this to epistemic appropriation, we see this mechanism at work: conflicting information (in the form of a hermeneutical resource) enters a “hostile” epistemological system, a system that is not equipped to fully appreciate these resources, and the conceptions associated to this resource change in the process. This happens specifically because these hermeneutical resources aim to describe something that questions the very adequacy of the epistemological system in place. If the epistemological system is resilient in maintaining its adequacy, it will incorporate, in inter-communal uptake, new hermeneutical resources in a way that fits with the epistemological system. Returning to the example of “woke” (discussed above as a form of resource-obscuring epistemic appropriation), the epistemological system in place in the white, dominant community is not able to fully take up the term “woke” because it would force them to acknowledge the inadequacies of this dominant system, which is founded upon assumptions of white supremacy. They cannot fully understand this resource because to fully understand it, they would have to confront the inadequacies of their worldview and epistemological foundations. This is not to say that no change is possible—epistemological systems are resilient, but not unchangeable.¹³

In this “the issue is, the dominant group members fail to appreciate exactly what the resource is, and to what extent it is meaningful” (Podosky 2023, 154). It is their ignorance to the intended meaning of the resource, and their ignorance to their ignorance—their meta-ignorance—which leads to them obscuring the resource in acquiring it (160).¹⁴ Their ignorance is foundational, determined by their epistemological system, which is influenced by social conditions, by their position in society, leading them to not be able to accurately grasp the psychological associations and significance of the hermeneutical resource. Resource-obscuring epistemic appropriation is then a kind of wrongful uptake that is facilitated by ignorance, owing to social and epistemic power relations. Within the wrong—in this case, ignorant—epistemological system, hermeneutical framework, a resource cannot be taken up in a way that does not obscure its meaning (Podosky 2023; Mills 2017).

However, as I have illustrated previously, what happens to an epistemic fata morgana is different from resource-obscuring epistemic appropriation. The meaning of the resource does not change, or, at least, does not seem to change. Even though the resource is not accessible—uttering it, utilizing it within the institution for its original purpose becomes very difficult—it is perceived.¹⁵ “Sexual harassment” still seems to mean the same thing, to point to the same phenomenon, whereas “woke” is removed from its context. Nonetheless, neither resource can be effectively utilized for its original purpose.

Yet, if the resource and the current conceptions attached to it are fundamentally incompatible with the epistemological system in place within the institution, why then, does the resource seem to stay the same? Regarding sexual harassment, meaningfully combating this within the university would require structural change. It would require better complaint procedures, and mostly, it would require the university to admit it has a structural problem, and change this—not acknowledge it in order to not be required to make meaningful changes.¹⁶ But truly acknowledging this structural issue is not in line with the epistemological system of the university, where the foundational assumptions are that the university is good, respectable, caring, and the shared hermeneutical resources, influenced by societal power relations, serve to keep up this idea.¹⁷ Acknowledging the structural issues becomes increasingly difficult, because having structural problems does not fit within the dominant, resilient epistemological system. However, the resource is still incorporated within the institution in a way that at least lets it be perceived, meaning mostly unchanged, even while carrying an acknowledgment of a structural problem. Why then, is this the case? In this, I argue we should turn to the next characteristic: although the resource is fundamentally incompatible with the dominant epistemological system, being perceived as taking up the resource is not.

4. Fourth characteristic: perceived compatibility

The fourth and last characteristic is that the hermeneutical resource that becomes an epistemic fata morgana is perceived as being compatible for uptake but is actually fundamentally incompatible with the epistemological system of the community it is taken up by. The argument is as follows: if the dominant epistemological system does not allow for a full acknowledgment of structural issues, the meaning of hermeneutical resources that point out these structural issues will be distorted—which might change their meaning entirely—or they might be rejected. However, if the fundamental assumption that there are no structural issues is based upon a notion of benevolence, these hermeneutical resources will be distorted in a way that fits with the dominant epistemological system because being good is compatible with this system. They will be taken up in a way that seems to preserve their original meaning. Yet, because the system is resistant to change, truly acknowledging structural problems will not be the result of this uptake. It will not result in real access to this hermeneutical resource, it becomes an epistemic fata morgana.

This becomes clearer in the context of the university. In the second section I discussed non-performative speech acts. I would argue that commitments to fighting sexual harassment in the university are non-performatives, because fighting sexual harassment requires admitting or acknowledging structural problems, which does not happen satisfactorily. Often these acknowledgments do not result in any real change, as reporting sexual harassment remains difficult and inaccessible. However, the university does not have to fight sexual harassment; these non-performatives are not completely necessary, silence would suffice. But universities do not stay silent because the sexual harassment is seen as undesirable, and the university is supposed to be good, progressive, and caring (Saxe 2022; Ahmed 2012, 2021). This leads the university to a need to make statements, and these statements are non-performatives because structural change and acknowledging the need for this are not compatible with the dominant epistemological system. Therefore, being perceived as fighting sexual harassment is compatible, because the university sees itself as good, but actually fighting sexual harassment is not compatible, because the university sees itself as good, and therefore doesn’t see itself as having a problem with sexual harassment. The hermeneutical resource is integrated within the dominant epistemological system in a way that does not truly acknowledge structural issues because this is incompatible with the system in place. However, the idea of integrating this resource is compatible, which results in the epistemic fata morgana.¹⁸

With these four characteristics, it is now clear why an epistemic fata morgana is a unique form of epistemic appropriation. It is a form of epistemic appropriation because hermeneutical resources are taken up inter-communally by dominant groups in a way that harms the marginalized creators and users of the resources. However, it is neither resource-preserving nor resource-obscuring epistemic appropriation, because while the meanings and associations of the resource shift, they are not necessarily obscured. Their perception is fundamental. This happens because of the second and fourth characteristics, because the phenomenon described by the hermeneutical resource in question is perceived as undesirable, and therefore, fighting against it is perceived as somehow virtuous.

5. The epistemic fata morgana and its consequences

I have alluded, in this paper, to some of the consequences of the prevalence of epistemic fata morganas, such as defensive attitudes and adverse consequences at work. However, in this section I elaborate more on the specifically epistemic consequences of the fata morgana, and I end by reflecting on what the epistemic fata morgana means for the academic field.

I define an epistemic fata morgana as a hermeneutical resource that was originally developed by marginalized knowers to describe and make sense of their own experience and taken up inter-communally in a way that, because of the (in)compatible epistemological system of the receiving community, upholds the resource’s visibility but restrains access to it. There are four specific characteristics of a hermeneutical resource turned fata morgana: the resource is developed by marginalized people to describe their experiences, the phenomenon described by the resource is deemed undesirable in dominant discourse, the epistemological system of the dominant community is fundamentally incompatible with the hermeneutical resource but—characteristic four—being perceived as taking up the resource is fundamentally compatible with the epistemological system. The existence of fata morganas in the university context, such as “sexual harassment,” can negatively influence those who produced the term in the first place or most need access to it. There are, I believe, at least two epistemological consequences: testimonial smothering and oppressive double binds.

Testimonial smothering is self-silencing. It is a specific subset of testimonial oppression where the speaker perceives “one’s immediate audience as unwilling or unable to gain the appropriate uptake of proffered testimony” (Dotson 2011, 244). The speaker does not think their audience will be capable of appreciating their testimony, which means that they will refrain from speaking. It might manifest in a queer person changing the way they speak in a group of people they perceive as heterosexual, it might manifest in them not sharing their experiences, for fear of being met with homophobia or not being believed. In our specific harassment case, we might imagine marginalized persons not coming forward with their experience, because they know, or they can judge that their audience (the university) is not competent to understand their testimony, suffering a communicative harm. Or, as a cognitive harm, they might start to doubt themselves, because what they experience and categorize as sexual harassment does not fit the institutional definition, or, more precisely, what they know they have to prove in order to be believed.¹⁹ They are capable of seeing (through) the fata morgana, they know that the inclusion of sexual harassment complaint policies is “window dressing,” to be perceived but not to be accessible. This is both an epistemic harm—their epistemic agency is violated—but also a very material harm because it might damage them emotionally, financially, and socially.

Another, related, consequence is knowers being placed in oppressive double binds (Bayruns Garcia 2019; Berenstain 2016; Hirji 2021). Hirji (2021) writes that double binds are “choice situations in which a member of an oppressed group is forced to choose between cooperating with and resisting some oppressive norm” (667–68). Because they are in this situation and “the way their own prudential good is bound up with their ability to resist oppression, they end up to some degree reinforcing their own oppression no matter what they do” (668). Berenstain (2016) outlines the double bind of epistemic exploitation, when marginalized people are required to educate their oppressors, who, because of identity prejudices, will not believe them anyway, locking them in these oppressive double binds (572). Bayruns Garcia (2019) outlines a similar phenomenon, where marginalized knowers are disadvantaged in terms of expression styles, as those who are members of non-dominant identity groups can be penalized for utilizing certain expression styles, styles that would allow them to maximize understanding in their audience. Therefore, they are in a double bind: they have to choose between using certain expression styles to maximize understanding, but be penalized for it, or not using these expression styles but not being able to maximize understanding (255). In the sexual harassment case, there is a similar double bind. When one does not come forward with one’s own experiences, it is harmful, and after the fact they might be blamed for not coming forward. However, when one does come forward, there is a chance the process will lead to nothing and end up hurting the knower significantly. There is no good choice in this scenario, which again leads to epistemic silencing. It is possible for a person to come forward and speak about their experiences, but this comes at a cost, as there is no way to get out of this double bind. It is only possible to go through it, but this might be unsuccessful. In this then, those who are aware of this might choose not to speak up, which, again, violates their epistemic agency.

These are just two of the possible epistemic harms—both cognitive and communicative—of the epistemic fata morgana. Related to this, by way of conclusion, I would now like to return to what this means for the field of epistemic injustice. I think the acknowledgment of another distinctive form of epistemic appropriation shows that an open conceptual structure of epistemic injustice, oppression, and appropriation is absolutely necessary. Dotson (2012) argued for this when reflecting on Miranda Fricker’s work, stating that “a catchall theory of epistemic injustice is an unrealistic expectation” (41). Theories of epistemic injustice are just that, theories, and closing conceptual structures can further forms of epistemic injustice, as epistemic oppression is pervasive—similarly to non-epistemic, social oppression. The epistemic fata morgana emerging as another form of epistemic appropriation shows this pervasiveness, and in no way means that the conceptual structure can be closed with this addition. Furthermore, this phenomenon specifically does force those in the field of epistemic injustice to extend their conceptual structures. We cannot assume that hermeneutical resources can be developed, and all will be well—this is clearly visible with both resource-preserving and resource-obscuring epistemic appropriation. The addition of the epistemic fata morgana also shows that, even if a hermeneutical resource is developed, shared, and lands in a community that seems to be relatively well-meaning, this does not mean its uptake will be without its problems; there is a real risk of it becoming an epistemic fata morgana. While this might be somewhat disheartening—there is much work to do—it does highlight, again, the importance of structural changes and the importance of expanding or changing one’s epistemological system.