Determining the scope of epistemic injustice within psychiatry

ABSTRACT

In this article, we delve in debates around the usefulness of the notion of epistemic injustice in psychiatry to show that the concept has been misportrayed in the literature. We suggest that epistemic injustice should revolve around phenomenology and regard first and foremost the failure of mental health professionals to acquire and utilize information that service users are experts in, i.e. first-person testimony pertaining to what it is like to be them. We use this conceptualization to demonstrate the unique benefits that the concept of epistemic injustice can bring to psychiatric practice, and we illustrate these through specific case studies, focusing particularly on autism.

KEYWORDS:

• Epistemic injustice
• philosophy of psychiatry
• natural expertise
• patient testimony
• phenomenological psychiatry
• phenomenological feedback

1. Introduction

The epistemic injustice concept (Fricker, 2007; Kidd et al., 2017) has taken the world of philosophy of psychiatry by storm. Indicatively, an “epistemic injustice psychiatry” search on Google Scholar returns 6,140 results for papers published since just 2022, while the 2023 annual meeting of the Association for the Advancement of Philosophy and Psychiatry featured 22 presentations, 11 of which were directly concerned with epistemic injustice. Given the concept’s short history of just over 15 years, debates on its fundamentals are still ongoing. Notably, Kidd et al. (2023), authors pivotal in the concept’s development, have pointed out that it is often employed without adequate attention to its recent, more sophisticated articulations, and that the “epistemic injustice” term is inappropriately used as a catch-all device to capture several unrelated epistemic vices. More provocatively, it has been argued that using the concept in psychiatry could lead to negative changes in the way psychiatrists administer care, with the normative consequence that psychiatrists should resist any attempts to change their practice based on the concept (Kious et al., 2023a). Philosophers of psychiatry also underline the need for additional clarity regarding epistemic injustice, the manifold of its categories, as well as its twin concept, epistemic justice. Perhaps even more importantly, they seek accounts guiding the integration of the epistemic injustice framework within clinical approaches. Arguably, the conceptual anatomy of an injustice is, after all, as good as its capacity to facilitate the undoing of that very injustice within human systems of practice.

Here, we aim to contribute to the clarification of the concept of epistemic injustice, particularly of its testimonial strand, and to the determination of its scope within psychiatry. Recent works have brought forward incidents within psychiatric practice, which we claim are wrongly conceived as exemplary cases of epistemic injustice. Consequently, within debates on the very usefulness of the concept, parties argue whether epistemic injustice does harm or good to psychiatric conduct over cases that give a misleading sense of what the concept is, or should be, about. Behind such misattributions, we argue, lurk all-too-generic conceptions of what the concept signifies. To undo epistemic injustice within psychiatry is often taken to mean something along the lines of “listening to the marginalized voices of service users”. However, the questions of why and when these voices ought to be listened to, as well as how they are to inform mental health practice, have been little explored, as Kidd, Spencer, and Carel (2023) point out. Concrete answers to these questions are important in adequately defining the epistemic injustice concept, portraying it properly in debate, and examining whether it indeed confers unique benefits to mental health practice not covered by preexisting concepts.

The arguments we advance are structured as follows. In Section 2, we provide a brief overview of epistemic injustice’s main variations, underlining conceptual blind spots that, we contend, have given rise to confusing entanglements in recent debates. We also highlight research directions for clarifying the concept and for exploiting it more powerfully in clinical practice. In Section 3, we examine cases from the literature that have been presented as paradigmatic exhibits of testimonial epistemic injustice and discuss the recent major debate on the suitability of the concept for psychiatric practice. In Section 4, we address the questions of when and why mental health professionals most ought to listen to service users in their practice. We propose a reconceptualization of testimonial epistemic injustice that prioritizes what we term phenomenological epistemic injustice. We do not maintain that phenomenological epistemic injustice is the only form of testimonial epistemic injustice, but we do hold that it should be considered its most important and characteristic subtype. The concept of phenomenological epistemic injustice recognizes that service users are first and foremost experts in what it is like to be them. As such, it underscores the necessity of integrating phenomenological testimony in psychiatric practice. We also contend that this conceptualization is instrumental in highlighting the unique benefits that the testimonial epistemic injustice concept can procure in psychiatric practice, thus providing answers to recent critical debates. In Section 5, we revisit the cases discussed in Section 3 to demonstrate that exemplifying testimonial epistemic injustice through these cases gives a misleading sense of the primary locus of the concept. We then identify this locus via a brief discussion of cases from the history of psychiatry, paying increased attention to that of autism.

2. Epistemic injustice: testimonial, hermeneutical, and beyond

The epistemic injustice concept was coined by Fricker (2007) in her titular landmark monograph. There, Fricker defines epistemic injustice as a wrong done to someone in their capacity as a knower, unintentionally and by way of tacitly operating prejudices. Fricker delineates two basic categories of epistemic injustice: testimonial and hermeneutical. Testimonial injustice occurs when a party’s credibility is unduly reduced within an epistemic process. For example, a woman’s testimony in court may be dismissed owing to the jury’s unconscious gender stereotypes concerning female honesty. Testimonial epistemic injustice was, at least in its first articulations, framed as chiefly agential, i.e., requiring the prejudices of specific agents to take effect. On the contrary, hermeneutical injustice is necessarily a structural phenomenon linked to processes of knowledge production and dissemination. It obtains when the interpretative practices and conceptual resources available to an epistemic process are inadequate to capture the experiences of a social group; “when collective hermeneutical resources … have a lacuna where the name of a distinctive social experience should be” (Fricker, 2007, p. 150). To illustrate hermeneutical epistemic injustice, Fricker provides the example of a woman who underwent post-natal depression in an era when psychiatry had not yet individuated and labeled this kind of experience. As a result, Fricker (ibid., p. 151) argues, individuals in positions such as said woman’s can be left feeling “cognitively handicapped”, e.g., in their inability to meaningfully identify what is going on in them, as well as “deeply troubled, confused, and isolated, not to mention vulnerable to continued harassment”. Fricker’s foundational work has been significantly built upon in recent decades, both by providing alternative accounts of epistemic injustice and by constructing additional sub-categories of her two main kinds. Notably, Coady (2017) has offered the concept of distributive epistemic injustice, which involves the misdistribution of epistemic goods (e.g., trustworthiness), and Mason (2011) has differentiated hermeneutical epistemic injustice into different stripes according to origin: global lack of resources vs. failure or denial of responsible authorities to incorporate such resources into the epistemic canon of relevant psychiatric practices.

In this paper, our focus will be primarily on testimonial epistemic injustice. This is for two reasons. First, we engage with clinically relevant epistemic injustice, and testimonial epistemic injustice is the primary concern within the day-to-day trenches of the psychiatric clinic. Hermeneutical epistemic injustice is very much relevant to psychiatry in general, as it draws professionals’ attention to the fact that their arsenal of tools for making meaning out of clients’ experiences is scarcely complete, thus stimulating recording and researching experiences that call for novel frameworks to accommodate them. Nonetheless, developing interpretative practices and conceptual resources to capture yet elusive experiences is a painstaking philosophico-medical endeavor that operates on a different timescale than even the most meticulous psychiatric diagnostic and treatment process. We should, however, add certain nuances to this assertion. Ritunnano (2022) has convincingly argued that hermeneutical epistemic injustice in psychiatry can occasionally take, besides a structural, also an agential form, in at least the case of service users diagnosed with psychosis. Combating this strand of hermeneutical epistemic injustice does not require developing novel interpretative practices and conceptual resources, but making sure clinicians are aware of and appropriately employ them. Therefore, this kind of hermeneutical epistemic injustice is indeed very relevant to day-to-day clinical practice. Furthermore, within psychiatry, hermeneutical and testimonial epistemic injustice may at times become intertwined. This implies that meaningful engagement with a service user’s testimony may require psychiatrists to become acquainted with unfamiliar concepts or hermeneutical resources. In the case of autism, for example, it has been argued that psychiatrists need to familiarize themselves with the phenomenology of the condition in order to properly comprehend autistic testimony (Pantazakos, 2019). Conversely, excluding testimonial participation of service users may conceptually impoverish psychiatric processes. We comment on this issue more extensively in Section 5.¹ The second reason for presently sidestepping hermeneutical epistemic injustice is that part of this paper’s purpose is to show that several recent accounts of epistemic injustice, and debates on its merit, have followed misleading applications of the concept. We also aim to provide a robust circumscription of epistemic injustice to guard against future misapplications. As it happens, these accounts and these debates engage primarily with testimonial epistemic injustice, and by extension so will we.

Recently, Kidd, Spencer, and Carel (2023) pointed out two major issues with the current employment of the concept. The first issue is that the “epistemic injustice” term is used far too broadly, signifying “anything bad epistemically” (ibid, p. 4), rather than specifically picking out harms relating to one’s capacity as a knower. Such overzealous uses of the concept, we add, not only miss what is distinctive about it, but can also obfuscate its usefulness in psychiatric practice, making it seemingly the case that what epistemic justice encapsulates is covered by already existing guidelines of good psychiatric practice. The second issue is that the epistemic injustice literature often deploys either under-articulated versions of the concept or reverts to the original Frickerian framework, ignoring rightful criticisms and amendments to it. Both tendencies hamper conceptual progress on epistemic injustice and under-exploit its potential, resulting in a lack of sophistication in analyses pertinent to the structure of the concept and its application in psychiatric practice.

Beyond cautioning against these phenomena, Kidd, Spencer, and Carel lay out directions for future research into epistemic injustice. Three are relevant to the present conversation. Firstly, they ask for conceptual and argumentative clarity regarding ameliorative projects of bringing about epistemic justice: “we must do more than just urge others to ‘listen’ to the ‘voices’ of the marginalized: there must be a clear idea of which voices ought to be listened to, and of what they are in fact saying, and why what they are saying matters, and how it could be translated into practical action” (ibid., p. 16). Secondly, they draw attention to the authority of first-person knowledge in properly understanding the richness of the experience of psychiatric conditions. Moreover, they argue, researchers and clinicians should find ways to situate the authority of mental health professionals besides, and not antithetically to, the authority of service users. This is complemented by, thirdly, understanding that psychiatric conditions often effect phenomenological changes, altering the experiential world of the subject who undergoes them. Thus, mental health practitioners should refer to service users themselves to comprehend what their world is like and what works best in its terms, rather than presupposing that their world is shared with those in neurologically typical conditions.

In what follows, and in heartfelt agreement with Kidd, Spencer, and Carel’s directions, we will attempt to draw the boundaries of testimonial epistemic injustice by tackling some of the key questions they put forward. Our approach centers on phenomenology, specifically on the idea that service users are first and foremost experts on their own phenomenal experience. We will introduce the concept of phenomenological epistemic injustice to delineate a distinct subtype of testimonial epistemic injustice, critical for elucidating specific instances of injustice that shape the landscape of psychiatric practice.

3. Purportedly paradigmatic cases of testimonial epistemic injustice

Let us now discuss cases from the literature that are meant to showcase testimonial epistemic injustice. Bueter (2019) has drawn attention to a species of pre-emptive testimonial epistemic injustice, effected by knowers being outright excluded from an epistemic decision-making process, making it impossible for them to provide testimony.² According to Bueter, this kind of injustice is exemplified by the widespread exclusion of service users from psychiatric classification. Opposing this, Bueter holds that taxonomic decision-making in psychiatry is thoroughly value laden. Therefore, the insights and values of the immediately impacted party, service users themselves, should feature in the process, inter alia to help “detect such value influences and provide input on stigmatizing diagnostic criteria” (Bueter, 2019, p. 1070).

Bueter presents three cases of diagnostic criteria laden with value judgments. Firstly, attention deficit hyperactivity disorder (ADHD) is defined by criteria such as “fails to give close attention or makes mistakes in schoolwork, at work, or during other activities” and “often leaves seat in situations where remaining seating is expected (e.g., in the classroom)” (APA, 2013, 59f). Bueter draws on Hawthorne (2010) to argue that these behaviors are considered symptomatic of a disorder only within specific socio-cultural contexts that uphold values such as rigid professional standards. Secondly, female sexual dysfunction criteria also involve behaviors rendered problematic only within societies that adhere to specific gendered norms. Thirdly, Bueter highlights service user exclusion from the diagnostic criteria of schizophrenia. She references studies on the subjective experience of the disorder to underscore that it often involves a disrupted sense of self, something which cannot be adequately captured through third-person observations or biological markers. Consequently, neglecting to involve individuals diagnosed with schizophrenia in the development of diagnostic criteria leads to the oversight of symptoms. In each of the cases above, Bueter concludes that the refusal to include service users in diagnostic deliberation has deprived the diagnostic process of information of fundamental importance that they possess. Often, this exclusion is to their detriment, constituting a form of preemptive testimonial epistemic injustice.

Another set of cases purportedly illustrating testimonial epistemic injustice can be found in the works of Sanati and Kyratsous (2015), and Crichton et al. (2017), all of which are discussed by Kious et al. (2023a) in their efforts to discredit the value of the epistemic injustice concept for psychiatrists. Sanati and Kyratsous present the case of J. N., a woman in her mid-20s admitted to a psychiatric ward, presenting with an acute onset of psychotic symptoms. Following a period of brief remission after treatment, she began to exhibit extreme distress, calling out to her partner to return by her side when he was absent, yet refusing to interact with him when he was present at the ward. When probed about the issue, J. N. would become frustrated or even angry at the medical team, also accusing nurses of being involved with her partner and her partner of intending to make her appear insane. Initially, the psychiatric team framed this as a development of delusional jealousy, only to soon discover, confirmed by J. N.’s partner himself, that he had indeed left her for another woman. Sanati and Kyratsous also present the case of M. G., a man in his late 20s diagnosed with schizoaffective disorder. After threatening to attack another person, M. G. was taken to a psychiatric hospital by the police. There, he was found to be acutely psychotic, expressing several irrational beliefs. Inter alia, M. G. claimed that the person he was threatening to attack had abused a close relative of his and was initially met with disbelief. Nonetheless, further examination revealed that his proclamation was factually true. Crichton et al. (2017) provide a similar case: that of a young man in a psychiatric ward in Munich claiming to be a relative of an ex-Soviet leader. His consultant considered this belief an illusion of grandiosity, only to later discover it was true.

Sanati and Kyratsous view theeir incidents as examples of testimonial epistemic injustice, as both J. N. and M. G. provided genuine testimony that was initially dismissed as fictitious. This dismissal occurred within the context of their status as psychiatric service users, leading the authors to argue that they were unfairly assigned a general, all-encompassing state of irrationality, and thus that their beliefs were deemed bizarre and incomprehensible, all of which is consistent with the social representations of their conditions. J. N. and M. G. were therefore wronged in their capacity as knowers, this owed to an acting stereotype. Moreover, the authors consider such cases not as isolated incidents but reflective of a systematic process that deprives service users of agency and testimonial credibility based on prejudices associated with mental disorder (Lewis & Appleby, 1988; Scheff, 1975). This results in the loss of crucial information regarding a service user’s history and life reality, as well as in feelings of disempowerment and frustration within the clinical context, all of which have a negative impact on their clinical progress. Crichton et al. do not explicitly argue why their case constitutes instances of testimonial epistemic injustice, however it can be self-evidently ascribed to the same category as those of Sanati and Kyratsous.

In the next section, we will argue that these cases, often taken as paradigmatic examples of perpetrated testimonial epistemic injustice, misrepresent the concept. Before this, let us review the criticism of the above papers by Kious et al. (2023a), which will be instrumental in demonstrating that they also adhere to false assumptions about the concept. Kious et al. begin by acknowledging that testimonial epistemic discrimination based on identity characteristics such as gender, race, ethnicity is wrong, as there is no evidence to suggest that, for example, racial minorities provide less credible testimonies. However, they argue that, unlike in the case of these identities, being a psychiatric service user does have epistemic significance. Kious et al. go no further in explaining just what the epistemic significance of being a psychiatric service user is but, as they discuss testimonial epistemic injustice, it is implied that the testimonial capacity of psychiatric service users is often affected by the conditions they are in, and thereby their credibility is, in at least some regards, rightly deflated. This is demonstrated in their reading of, for example, J. N.’s case. Despite her being correct about her partner intending to leave her, Kious et al. argue that attending psychiatrists would have done well to exercise caution in believing her initial proclamations. J. N., that is, was psychotic when she made these claims, which were in their own right antecedently improbable to be true, much like the Munich man’s statements regarding his blood relation to the ex-Soviet leader. Therefore, this hesitation and initial refusal to believe J. N. do not constitute acts of epistemic injustice. In all probability, it was rather “just a mistake” (Kious et al., 2023a, p. 3) rather than a moral wrong.³

Kious et al.’s overarching point is that incidents that have been brought to the fore as paradigmatic exhibits of testimonial epistemic injustice in psychiatry are not cases of injustice at all, thus undermining the concept’s credibility. They mobilize their criticism to assert that the concept of epistemic injustice is at best obsolete and at worst actively harmful for psychiatric practice. In support of this position, they advance three theses. First, there is no evidence that cases of testimonial epistemic injustice, even if such injustice is indeed perpetrated, are representative of clinical encounters in psychiatry. Second, existing guidelines for good psychiatric practice already urge psychiatrists to be interested in what service users have to say. Besides, psychiatrists generally regard service users’ testimony as the very foundation of psychiatric information gathering.⁴ Third, changing clinical practices in response to claims that psychiatry is epistemically unjust would risk harming service users by replacing healthy skepticism and critical thinking, elements indispensable for psychiatry, with a pursuit of ill-conceived social justice. Often, Kious et al. (ibid., p. 4) maintain, great benefits are to be had in resisting the conclusions service users push for, making them see things they would otherwise be blind to through “gentle nudges, applied gradually and over time”.

Running against epistemic injustice’s prominence in the philosophy of psychiatry, the paper by Kious et al. prompted several immediate responses. Aftab (2023), for instance, argues that adhering to the concept of epistemic justice does not require psychiatrists to believe everything a service user says or otherwise give up good epistemic standards; the concept only calls for giving epistemic credence where it is due, steering clear of unjust biases due to prejudice. Kidd, Spencer, and Harris (2023) are aligned with Aftab on this point, adding that, contrary to the protestations of Kious et al., there is an abundance of evidence showing that epistemic injustice is systematically perpetrated within psychiatric practice. This, they argue, indicates that the available psychiatric tools are not well-crafted to address this problem. Moreover, simply “being interested” in what service users have to say is not enough to solve the issue of testimonial epistemic injustice; a sexist person may be interested in what women have to say and still deny them appropriate forms of credibility, testimonial uptake, and epistemic authority. This could happen, for example, if said sexist took interest in women’s words just to select some as evidence for their already entrenched sexist views. Similarly, a clinician might be interested in the content of a service user’s discourse plainly to gauge the nature of a perceived delusional pathology, i.e., without crediting them as a reliable testifier. The debate on these fronts rages on, with Kious et al. (2023b) having already published a response to some of the responses.

Before moving on, it is important to emphasize that the above cases’ capacity to demonstrate the testimonial epistemic injustice concept has gone largely unquestioned. We are not saying that there is unanimous agreement on that testimonial epistemic injustice has indeed been perpetrated in all these cases; as we just saw, quite the opposite is true. The ensuing disagreement, however, revolves around whether there is something morally wrong within these cases, rather than their suitability as paradigmatic illustrations of testimonial epistemic injustice – which is mostly just assumed. Problematizing this assumption, in the next section we argue that most of the cases just discussed are not characteristic of testimonial epistemic injustice properly construed. Thus, the merit of the concept is not weakened even if everything in these cases is morally stellar. In our view, this misportrayal is indicative of a wide misuse of the concept in the literature and implies that ongoing debates based on the above cases will result in misguided conclusions. To argue our point, we will first ground testimonial epistemic injustice in phenomenology, shifting the weight of the concept toward service user testimony on their phenomenal experience. We will then utilize this framework to identify the primary locus of testimonial epistemic injustice in psychiatry. We will conclude by demonstrating how the debate we surveyed in this section can make significant headway toward resolution under the guidance of testimonial epistemic injustice phenomenologically understood.

4. Testimonial epistemic injustice and phenomenology

We will now address two of the central questions Kidd, Spencer, and Carel (2023) put forward, namely why and when (when they speak about what) service users should be listened to in psychiatric practice. We begin by drawing an informational distinction. There is voluminous information of all kinds that pertains to us: our height and weight; gender and ethnic background; education; sexuality; biological markers; medical and family history; and so forth. Within this body of data, we can draw a distinction between third-person information, accessible from a third person perspective, and first-person information, the access to which requires lived experience. To provide some examples of third-person information, others can generally find out who our relatives are by referring to the relevant documents of the registry office; whether our partner is unfaithful by hiring a private detective; and whether we have bowel cancer by reviewing the results of a flexible sigmoidoscopy. In such cases, the epistemic value of our own input in accurately providing relevant information is at best equal to that of appropriate third-party sources. It may even be minimal, as in the case of bowel cancer, as most of us can report blood in our stool but cannot with any reliability self-diagnose bowel cancer.

On the other side of the distinction lies first-person information, which concerns the information we have, in principle, privileged access to: information about ourselves that no one besides us possesses and can convey as accurately as we can. First-person information regards mostly, though not solely (see next paragraph), what philosophers, primarily phenomenologists, refer to as “what it is like to be us”. For instance, imagine that I am sitting in a conference room, watching a presentation. My experiential world in that moment, ranging from the objects that my gaze is directed to and the angles I view them from to my cognitive and emotional states, is something I can know as well and as fully as no one else, simply because I am the only one in that specific position with the range of experiences that such a situation triggers.⁵ Numerous recent philosophical and clinical accounts advocate for the need to integrate first-person perspectives in psychiatry (Friesen et al., 2021; MacLeod, 2019; Stinson, 2019; Tekin, 2022; Zerbe et al., 2020). This integration is instrumental, inter alia, for an accurate understanding of certain psychiatric categories (Stanghellini et al., 2021; Stinson, 2019). According to such accounts, the phenomenological expertise of service users is unique, thus conferring on them epistemic authority on what it is like to be them. Therefore, and insofar as psychiatry is about helping service users get by better, i.e., improving what it is like to be them, this information is invaluable, and psychiatrists should prioritize it as primary information in research and in the clinic.

Two important remarks on the proffered distinction are in order. First, there are cases where first-person access is not about what it is like to be us, and yet also confers epistemic authority. Another way to put this point is to say that it is possible that I am the ultimate epistemic authority on information about me that is in principle derivable from third-person observations, but not actually. For example, suppose that I am the last surviving member of a tribe that kept no records of the genealogy of any of its members, and that all knowers of my family history except me are long dead. In this instance, I indeed have more epistemic authority on my family history than any other source does. In cases like this, the kind of information involved falls under the second leg of the distinction – first person-information, or information about me that I am privy to – while not concerning what it is like to be me. The difference between this kind of information and information that pertains to what it is like to be me is that the former can arise only circumstantially, as in the above example, while the latter is by its nature best available to me. We will return to this matter to show how confusion about it may have contributed to some of the misunderstandings within the debate on epistemic injustice in psychiatry.

Second, one could object that immediate access to what it is like to be me does not confer epistemic authority over a third-person perspective because I might not have the conceptual resources to express it. Seemingly, a trained specialist may be able to surmise some information on what it is like to be me, which might not be available even to me. For example, if my psychiatrist is present in the same conference room as I and observes me frantically tapping my foot on the floor before my talk, they may soundly conclude that I am anxious. However, this criticism does not suffice to contradict that one has privileged access to what it is like to be oneself. Firstly, the hypothetical psychiatrist’s knowledge of my situation still does not capture the experiential wealth of what it is like to be me in that conference room in its entirety, but only in a piecemeal way. Secondly, internal states produce varying behavioral manifestations in different people. Thereby, extrapolations about internal states based on external cues are, in general, not as reliable a source of knowledge regarding such states as phenomenological experience. Thirdly, my experiential state is not tracked by the knowledge that anxiety is tacitly operating in the background; this is a diagnostic statement relating to unconscious and/or neurobiological processes in effect within me, and these are, by definition, not part of my phenomenology. What it is like to be me regards the, so to speak, “raw” phenomenal material, is separate from my conceptual capacity to express it (Colombetti, 2009), and thereby the subject of my own expertise by experience.

Having made these clarificatory remarks, it is now time to specify the role and importance of phenomenology within the concept of testimonial epistemic injustice. In line with those who underscore the crucial link between phenomenology and psychiatry, we argue that psychiatry, in its aiming to improve service users’ well-being, is primarily concerned with what it is like to be one. Psychiatrists endeavor to battle, e.g., depressive and anxiety disorders, in order, plainly put, to help service users better cope and live more fulfilling lives. Moreover, psychiatry is far from identifying concrete causal relationships between states of the body and the brain and phenomenal states of the mind, both in its own regard and compared to other strands of medicine. For instance, draining an appendix abscess is far more warranted to alleviate pain than prescribing antidepressants is to alleviate depressive moods. While similar criticisms have been advanced about other medical specializations and specific disorders (Carel, 2012), phenomenological underdetermination is particularly acute in psychiatry. This is best evidenced by the fact that most disorders listed in the Diagnostic and Statistical Manual of Mental Disorders may not even be diagnosed on neurochemical or otherwise biological markers alone. In turn, this elevates service user testimony to a special status within psychiatric practice, making it, as Kious et al. (2023a, p. 3) maintain, “the very foundation of psychiatric information gathering”. This includes, besides the service user’s various records such as family and medical history, chiefly information about what it is like to be them, especially regarding pre- and post-treatment intervention comparisons. In the absence of third person-accessible markers, successful psychiatric treatment necessitates the establishment of a dialectical process between the attending team and service user to facilitate to corrective maneuvers. Consider, for example, prescribing most psychiatric medications, such as anxiolytics. Psychiatrists may have a general idea of what works from statistics, but it is only via adjusting the kind and dosage of medication to the specific treated person by way of integrating their first-person testimony that the optimal result may reliably come about.

The preceding answers Kidd, Spencer, and Carel’s questions of why psychiatrists should listen to service users as well as when. Psychiatrists should not listen to users in the name of some arbitrary concept of justice and participation for participation’s shake. They should listen to them because psychiatrists are chiefly concerned with how users get by, and users are unparalleled knowers of how they get by. Moreover, psychiatrists should not listen to service users about everything; they should listen to them when users speak about topics they know – and what it is like to be them is a topic they know like no one else. It follows that seeking to acquire and exploit phenomenological service user testimony are among a psychiatrist’s most important epistemic duties. Conversely, failing to make efforts to acquire such testimony, ignoring it when available, and/or underutilizing it in intervention constitute acts of testimonial epistemic injustice. Ignoring an ultimate authority’s input in a process that regards them is, by definition, as strong a case of epistemic injustice as they come. We term this kind of injustice phenomenological epistemic injustice: doing a wrong to someone in their capacity as a knower of what it is like to be them.

We contend that testimonial epistemic injustice in psychiatry should predominantly be conceptualized as phenomenological epistemic injustice. By this, we mean two things. Firstly, phenomenological epistemic injustice should be regarded as intimately tied to psychiatry’s main mission – helping people cope – and thus as potentially the most catastrophic form of epistemic injustice. As just argued, when phenomenological epistemic injustice occurs, it obstructs access to a pool of evidence that is crucial for guiding psychiatrists in aiding service users to cope better, evidence that can only be sourced from the users themselves. Secondly, as we will substantiate in the next section, there is evidence to suggest that phenomenological epistemic injustice has likely been widely perpetrated in psychiatry, leading to significantly adverse outcomes for service users.

However, and this is a delicate yet important point, we do not suggest that testimonial epistemic injustice in psychiatry should exclusively regard phenomenological epistemic injustice; we do not, in other words, intend to equate the two concepts. If a service user’s epistemic credibility is unfairly deflated when they testify about something other than what it is like to be them – for example, about who their relatives are – then this too constitutes testimonial epistemic injustice, a kind which the phenomenological epistemic injustice concept does not cover. Our earlier discussion concerning the capacity of one to be a unique expert in third-person information about themselves is relevant here, as we can express the point just made by saying that psychiatric testimonial epistemic injustice that regards third-person information can well be perpetrated. Service users can be, and doubtlessly have been, unfairly dismissed when testifying about something other than what it is like to be them. Thus, what we propose is, so to speak, a “shifting of the conceptual weight” of testimonial epistemic injustice so that it primarily addresses neglects pertinent to phenomenological testimony. Practically, this entails prioritizing phenomenological epistemic injustice by highlighting relevant cases to exemplify testimonial epistemic injustice in psychiatry and drawing psychiatrists’ attention to this form of injustice, while still accommodating other forms of epistemic injustice in research and educational programs.

5. Phenomenological epistemic injustice in action

We will now flesh out the concept of phenomenological epistemic injustice, and its significance within debates concerning testimonial epistemic injustice, by discussing individual and larger-scale cases from the history of psychiatry. We will begin by returning to the alleged cases of testimonial epistemic injustice from Section 3. We argue here that these cases are doubly problematic. Firstly, some are plausibly not even instances of testimonial epistemic injustice, phenomenological or other. Secondly, and independently of the first point, they concern information about service users that users are not natural experts in, i.e., information about them that is reliably accessible in the third person. Therefore, they fail to illustrate the concept of testimonial epistemic injustice in psychiatry, and any general conclusion about testimonial epistemic injustice drawn from them alone is misguided.

Let us firstly take the case of J. N., who alleged that her partner was abandoning her. Sanati and Kyratsous (2015) write that “at the same point that the psychiatric team was able to formulate her beliefs as a case of delusional jealousy, they realized that her partner had left her for another woman. This was in fact confirmed by him”. Now, it is not clear what “realized” entails here but, unfortunately, it is quite crucial. Does it mean that the psychiatric team stumbled upon the accurate information completely by chance just after framing J. N.’s beliefs as delusional and dismissing them? Or did they adopt a “trust, but verify” attitude, whereby the team was hesitant to believe J. N., leaned toward a delusional framing, and so referred to a third-party source, namely J. N.’s partner himself, to establish the facts? We maintain that the psychiatrists’ stance is warranted in at least the second scenario.

While one might argue that adopting a “trust but verify” approach falls short of unconditional trust and consequently reduces the service user’s epistemic credence, we contend that this reduction is justified in the present context. This is so owing to three reasons. Firstly, we think that, regardless of whether one is a psychiatric patient, a degree of skepticism is appropriate in cases involving suspected infidelity. Secondly, J. N. had only very recently exhibited psychotic symptoms when she made the assertion that her partner was leaving her, and thus it was plausible that she was still entertaining delusional beliefs – recovery from psychosis and delusion does not happen overnight but is often slow and gradual (Eisenstadt et al., 2012). Thirdly, just before expressing the belief in question – that her partner was leaving her – J. N. had made an obviously delusional statement on the same topic, accusing her nurses of being sexually involved with her partner. When two similar beliefs are expressed by an individual very recently in a psychotic episode, and one of these beliefs is delusional, it is reasonable to consider the other belief quite likely to be false as well. These considerations show that J. N.’s epistemic credence was reduced neither in general nor due to her past psychotic experiences, which are the hallmarks of testimonial epistemic injustice. Instead, her credence was specifically reduced in a context closely associated with her psychotic episode and concerning a topic she was already delusional about. We thereby conclude that a “trust, but verify” attitude was warranted in J. N.’s case, and that she was not the victim of epistemic injustice in at least the scenario where her attending doctors actively tried to confirm her claims instead of outright dismissing them. Although the argumentative force of J. N.’s assertion turning out to be true is not lost on us, we also invite the reader to consider whether they would find an accusation of committing testimonial epistemic injustice toward J. N.’s doctors plausible had it not turned out to be true. Ultimately, the veracity of the claim should not influence the assessment of the incident, as the debate is precisely about the appropriate stance when such veracity is unknown.

The cases of M. G., who accused another person of abusing their relative, and of the Munich man, who claimed that he was the relative of an ex-Soviet leader, are even harder to judge, as even more pieces of the puzzle are missing. We are neither informed about how their claims were cross-referenced by the attending teams, nor about the Munich man’s diagnosis and general mental state. What we do know is, firstly, that the truth about their statements came to the surface, and secondly, that the sought-after information was third-person information. Given that M. G. and the Munich man were in a state that also excused a “trust but verify” attitude on behalf of their psychiatrists and that appropriate third-person sources were consulted, e.g., a court case documenting the alleged abuse or a family witness in the case of M. G., and official registry documents in the case of the Munich man, one may soundly conclude that these cases do not constitute cases of testimonial epistemic injustice either. Of course, assuming that these conditions actually obtained is problematic but so is, we hold, assuming that they did not obtain. Therefore, these cases are, minimally, tentative instances of testimonial epistemic injustice.

Returning to the main argument of this paper, and entirely independently of whether there was something morally suscept within each these cases, we contend that they are not suitable examples for exemplifying the testimonial epistemic injustice concept. The information involved in all the above cases is exclusively third-person information and does not pertain to what it is like to be a service user. Thus, if we are correct in maintaining that testimonial epistemic injustice should be primarily understood as phenomenological epistemic injustice, then J. N., M. G., and the Munich man are incidents inadequate for encapsulating the essence of the testimonial epistemic injustice concept. As such, even if Kious et al. (2023a) are right about discarding these cases as incidents of testimonial epistemic injustice, they have not done much to scathe the concept’s usefulness for psychiatry. Moreover, even if they are correct in asserting that everything that went wrong in these cases is already covered by existing psychiatric guidelines, they have not demonstrated that the testimonial epistemic injustice concept does not confer unique benefits to psychiatric practice.

It is now time to highlight where we believe current debates on epistemic injustice, such as the Kious et al. (2023a) vs the rest, have gone astray. Kious et al. are skeptical about what epistemic injustice has to offer, and Kidd, Spencer, and Harris (2023) accuse them of failing to see this because of operating within too narrow an understanding of what epistemic injustice is. We think Kious et al. have the exact opposite problem: they adopt an overly wide conception of testimonial epistemic injustice, which unwittingly encompasses all kinds of information, and mainly third-person information, as in the previous three examples. Grounding testimonial epistemic injustice in phenomenology and translating it primarily as phenomenological epistemic injustice solidifies the distinct benefits the concept can confer upon psychiatry, benefits not fully covered by existing good psychiatric conduct guidelines. Moreover, the conceptualization proffered helps alleviate concerns about the loss of healthy skepticism and critical thinking in psychiatric treatment. Through the proposed conceptualization of testimonial epistemic injustice, psychiatrists are “allowed” to doubt and be apprehensive about believing whatever they want, provided that they consult appropriate sources to settle arising matters. This with the added provisos that the appropriate sources of how service users are doing are service users themselves, and that how service users get by is given the importance that it is due.

A reasonable question may at this point be raised as to whether phenomenological epistemic injustice is typically perpetrated in psychiatric practice, especially in comparison with testimonial epistemic injustice that is not phenomenological. We have already granted that third-person testimonial epistemic injustice is possible in psychiatry; the issue of whether it is frequent outruns the scope of the present paper. What we will now argue is that there is abundant evidence that phenomenological epistemic injustice has indeed run rampant in mental health practice, expounding on relevant case studies, particularly emphasizing autism.

Recall Bueter’s discussion of ADHD and schizophrenia. The case of ADHD is, by our conceptualization, not a case of testimonial epistemic injustice, preemptive or otherwise, insofar as the information of interest is that some of the current symptoms of ADHD are rendered pathological in certain social contexts. This is not information that people diagnosed with ADHD have privileged access to; certain sociologists of medicine do, irrespective of whether they are ADHD service users. On the contrary, once the conversation comes to the stigmatization of pathologizing such symptoms, especially regarding the impact of such pathologization on the individual, service user input is very much relevant, and their exclusion from relevant deliberations becomes an act of phenomenological epistemic injustice. The case of schizophrenia is even more straightforward, quite clearly a case of preemptive phenomenological injustice. The information of interest here regards subjective symptoms of schizophrenia, such as the loss of self, which are important in diagnosis and potentially in treatment, and which can be detected by no other source than people with schizophrenia themselves.

Before we proceed with more examples of phenomenological epistemic injustice, we want to take a brief conceptual detour through an issue we alluded to in section 2: the intertwining of testimonial with hermeneutical epistemic injustice. Consider the case of schizophrenia discussed previously. The exclusion of testimony from individuals diagnosed with the condition does not just rob the diagnostic process of crucial symptoms. It also impedes the development of concepts necessary to capture such testimony in all its wealth, as it deprives the diagnostic process of the “raw material” – the service users’ phenomenal experience – that signals the need for the development of such concepts. Conversely, adequately registering the testimony of service users may require psychiatrists to familiarize themselves with phenomenological concepts typically obscure in mainstream psychiatry, e.g., the loss of self, or even endeavor to develop new ones when confronted with an experience that no available concepts seem adequate to capture. In psychiatry, this issue is most pronounced in attempts to do phenomenological epistemic justice, as this requires effective communication and interpretation of what it is like to be one in a psychiatric condition – a task which, as one anonymous reviewer put it, may be “devilishly tricky” to accomplish. Carel and Kidd (2014) argue this point regarding the context of illness broadly conceived, where hermeneutical epistemic injustices arise from the inadequate understanding of illness by those who have not experienced it. Despite having kept testimonial and hermeneutical epistemic injustice relatively apart so far, we recognize the significance of the issue, not least in emphasizing the importance of rigorous phenomenological inquiry within psychiatric processes.

Let us move on to our focal point of autism in relation to phenomenological epistemic injustice. According to the sovereign, medical model, autism is the result of neurological and psychological disfunctions rooted within the individual. Within clinical practice, autism is operationalized as a neurodevelopmental disorder, characterized by deficits with communication and sociability, restricted and repetitive behaviors, and/or multimodal sensory sensitivities (APA, 2013). Treatment patterns data for the U.S. show that Applied Behavior Analysis (ABA) is by far the most dominant treatment for autism, followed distantly by pharmacological interventions (Shoaib et al., 2021; Xu et al., 2019) As a behavioral approach, ABA understands autism as a collection of problematic behaviors. In turn, the method’s purpose is to eradicate such behaviors, as well as to establish ones thought to be more adaptive. ABA is self-described as “using scientific principles and procedures discovered through basic and applied research to improve socially significant behavior to a meaningful degree” (Association of Professional Behavior Analysts, n.d.).

Critical viewpoints characterize ABA as a process “in which the autistic child is rewarded for engaging in activities that make him more normal” (Wilkenfeld & McCarthy, 2020, p. 37). Various approaches to autism that most self-advocates espouse, e.g., the neurodiversity movement or critical autism studies, run against both the medical model and ABA, opting for non-pathologizing descriptions of autism. Historically, ABA has carved a well-known trajectory of abuse and violation of human rights, employing tools which have since become widely condemned as illegitimate, such as shouting, slapping, and delivering electric shocks to meet set goals (Bowman & Baker, 2014). Currently, the many iterations of ABA have largely distanced themselves from such techniques, and mostly from aversive methods in general.⁶ Yet, ABA’s many critics purport that it remains at core a method of “reprogramming” autistic individuals in a way that conforms with social norms. Further, they argue, this process is entirely out of step with the needs of autistic individuals themselves. ABA, critics maintain, often forces autistic individuals to leave perfectly harmless behaviors of, e.g., self-regulation, behind (Kapp et al., 2019; Milton & Sims, 2016), as well as to take on new ones that are meaningless for them, thus failing to improve their quality of life. For instance, many autistic adults maintain that eradication of stimming (self-stimulatory behaviors) deprives them of an invaluable tool for self-regulation (Rudy, 2018), while forced acquisition of typical social behaviors such as eye contact are, even after integration, experienced as thoroughly distressing (Trevisan et al., 2017).

Criticisms of ABA often fall under the wider rubric of the neurodiversity movement (Arnaud & Gagné-Julien, 2023; Pantazakos, 2019; Pantazakos & Vanaken, 2023). The concept of neurodiversity, conceived of in the mid-1990s and propagated in parallel with the development of online communications, is minimally defined by its opposition to the medical model of disability (Dwyer, 2022). The neurodiversity movement opposes the pathologization of mental disabilities, including autism, and promotes acceptance of diversity in how human minds and brains operate. The movement understands “neurominorities” as morally equal but marginalized groups (Catala et al., 2021). It conceptualizes disability in terms of a social model, whereby disability occurs as a mismatch between individual and society and is not located plainly in problems within the individual. In this vein, the movement espouses the term “autistic person” instead of “person with autism” to pinpoint autism as a non-problematic feature of the self; instead of having autism, one is autistic, this constituting a core part of their identity. By extension, neurodiversity advocates oppose the forced normalization of autistic people and, insofar as ABA serves this purpose, they oppose ABA as well. In the place of normalization, neurodiversity exponents offer a call to abolish prejudices, stigma, and gain better understanding of autistic people in public health, education, and society at large. Since autism entails nothing inherently pathological, the argument is, society should move to accommodate autistic individuals and their needs, rather than (solely) the other way around.

Let us move to phenomenological testimony provided by autistic people, which will prove critical in connecting this conversation to ours. In her article, Quiet Hands, Bascom (2011), an autistic self-advocate, contends that that she underwent abuse during a behavior modification program, which aimed at abolishing her, by normative standards, purposeless behavior of rubbing her hands and moving them around. As a result, Bascom claims, important means of accessing the world and processing information were taken away from her:

When I was six years old, people who were much bigger than me with loud echoing voices held my hands down in textures that hurt worse than my broken wrist while I cried and begged and pleaded and screamed.

… I’ve been told I have a manual fixation. My hands are one of the few places on my body that I usually recognize as my own, can feel, and can occasionally control … My hands know things the rest of me doesn’t. They type words, sentences, stories, worlds that I didn’t know I thought. … They tell me what I think, what I know, what I remember. … My hands are an automatic feedback loop, touching and feeling simultaneously. I think I understand the whole world when I rub my fingertips together.

… [I]f you grab my hands, if you grab the hands of a developmentally disabled person, if you teach quiet hands, if you work on eliminating “autistic symptoms” and “self-stimulatory behaviors,” if you take away our voice, if you …

Then I …

I …

Her article ends without conclusion, signifying her resulting loss of voice. Kapp et al. (2019) present an account of autistic views on stimming, which shares Bascom’s spirit. According to Kapp et al., occurrent themes from their phenomenological analysis of autistic adults’ testimony corroborate the view that stimming amounts to an effective management of sensory and cognitive overwhelming that autistics often face when they find themselves in a noisy environment. Luke, a participant, said of his controlled self-stimulation (ibid., p. 1786):

It helps you talk to yourself at a rhythmical pace, so when I’m doing this I can sort of think in the rhythm that I’m moving my hand … which is very helpful because it means like when you’ve got your internal monologue it doesn’t all come in at once and you find yourself sort of shouting at yourself in your head to get everything done.

Sally, another participant said:

And I started kind of incorporating it more in my life, and it actually managed to help me stave off some panic attacks. For example, I never used to wave my hands that much, but I’ve started doing it more and it actually helps, like if I’m in a crowded elevator or something.

Rose, an autistic adult, commented on ABA’s targeting stimming for elimination:

[T]hey basically condition them like Pavlov’s dogs to stop stimming … to me it was abuse, because stopping those children stimming when they’re trying to calm themselves down or cope with a situation, because even if they manage all the environment around them, there might be situations that they find stressful, and if they haven’t got the ability to calm them down, then they could be relying on other people for the rest of their lives or have a breakdown …

In another paper by Trevisan et al. (2017), autistic individuals relay similar experiences regarding what it is like for them to make eye contact. Here are some indicative quotes:

Making eye contact feels sort of like the first breath one takes under water using scuba gear, where there’s this moment of panic as your body says, “No, no, you’ll drown!”

Eye contact is sort of like getting shocked or something; it is very unpleasant and almost hurts, and if I am forced to do it for any length of time I get increasingly panicked.

I’d say it always causes a lot of discomfort and stress, psychological and physical. (tremors, stuttering, sometimes headaches)

[Eye contact] … saps my energy [from] my core like being hypnotized by a cold-blooded energy sucking vampire.

Trevisan et al. note that “improving” eye contact behaviors in autistic children has long been the goal of intervention, this fueled by the assumption that children must look at the behavioral instructor and attend to them before learning can take place. ABA, they add, has been protractedly utilized to achieve this end, despite yielding small to moderate improvements in joint attention and other eye gaze behaviors. Behavioral approaches have also frequently sought to enforce practices like “quiet hands” and to eradicate stimming (Sandoval-Norton et al., 2019).

Providing a comprehensive evaluation of ABA and adjudicating the raging debates around it is obviously beyond the purview of this paper. However, the testimonies we presented, taken in the light of our analysis of phenomenological injustice, make clear that at least many autistic individuals do not get by better following the application of techniques such as the above. In fact, their lives are sometimes substantially worsened. Thus, insofar as these techniques aim to improve the quality of life of autistics, they mostly fail. This failure is a result of ignoring phenomenological testimony on behalf of service users themselves: what it is like to be them, as seen by them. In other words, mental health professionals have here failed at what we outlined above as one of their chief epistemic duties: to seek, acquire, and try to understand phenomenological testimony on behalf of the service users when aiming to improve what it is like to be them. They have done a wrong to autistic people in their capacity as knowers of what it is like to be them, and no existing guidelines of good practice have been able to protect such service users. As such, we are here presented with a case of phenomenological epistemic injustice par excellence.

This bolsters our argument that the perpetration of such injustice is a real and protracted phenomenon in the history of psychiatry, worthy of being addressed by a novel framework. Our position is that, for reasons outlined previously, it is cases such as this that should be employed to demonstrate the concept of testimonial epistemic injustice in psychiatry, and not cases like the ones of Section 3. Notice, finally, that, here too, phenomenological epistemic injustice may have been exacerbated by hermeneutical epistemic injustice and vice versa. For example, part of the insistence on the above interventions could be attributed to hermeneutical challenges in translating autistic experiences of imposing eye contact and limiting stimming in a way accessible and comprehensible to non-autistic practitioners. This problem may become even more acute in the case of service users who do not use verbal language as a means of communication.⁷

The tendency discussed above, of ignoring autistics’ own understanding of themselves and their needs, can be seen to extend beyond behavioral intervention and to constitute a problem for clinical research in general (Pantazakos, 2023). The 2018 portfolio analysis by the Office of Autism Research Coordination on behalf the Interagency Autism Coordination Committee (IAAC), the primary advisory body to the U.S. government on issues regarding autism, found that just 6% of both private and public funds went toward researching the quality of services for autistic people, while a mere 2% was utilized in researching the needs for autistic adults (Interagency Autism Coordinating Committee, 2018). In a statement relating to a previous report by the same agency and showing similar numbers, Ne’eman, then president of the Autistic Self Advocacy Network, declared that such figures “show a shocking lack of interest in aligning scientific investment in autism research to the priorities of the most important stakeholders: autistic people ourselves” (Autism Self Advocacy Network, 2016). Notably, the IAAC itself has reacted to this state of affairs, recommending a paradigm shift in research orientation with the aim of improving services across the lifespan and addressing the daily needs of autistic people (Interagency Autism Coordinating Committee, 2017). Our analysis in this paper suggests that such a paradigm shift can be cast as bringing about phenomenological epistemic justice within the autism clinic, and that it can materialize by actively seeking and utilizing autistic phenomenological testimony in such settings. Our proposal aligns with others (Drożdżowicz’s, 2021; Guenther, 2017) advocating for fulfilling a vital epistemic duty in psychiatry by focusing on justice that privileges the firsthand experience of service users.

6. Conclusion

Our purpose in this paper was to clarify the concept of testimonial epistemic injustice and denote its scope within psychiatric practice. We started by discussing examples of purportedly paradigmatic cases of testimonial epistemic injustice that have been brought to the fore as part of the wider debate on the very usefulness of the concept for psychiatrists. We argued that some such incidents are plausibly not cases of testimonial epistemic injustice, but also that this is immaterial for the value of the concept in general, as such cases misrepresent what testimonial epistemic injustice is chiefly about. We offered that the concept should first and foremost regard areas of knowledge in which service users are natural experts in, i.e., phenomenological knowledge that conveys what it is like to be them. We underlined the importance of such knowledge in psychiatry, and we defined phenomenological epistemic injustice as what transpires when an epistemic process fails to acquire, ignores, or underutilizes such knowledge.

We believe that framing testimonial epistemic injustice thus assuages worries raised by Kious et al. (2023a, 2023b) regarding whether the concept confers a unique benefit to psychiatric practice. Our account demonstrates a crucial, distinct, and yet not included in relevant guidelines way in which the testimonial epistemic injustice concept should inform psychiatric practice: urging psychiatrists to attain and appropriately exploit specifically phenomenological feedback within their clinical encounters. Paying attention to the subjective phenomenology of mental illness, as well as articulating and integrating therapeutic goals factoring in this phenomenology, is one of the most critical areas where psychiatrists can do testimonial epistemic justice to their service users. Lastly, against Kious et al.’s contentions that it is doubtful whether epistemic injustice regularly obtains within psychiatry, we demonstrated a few cases where phenomenological epistemic injustice has been consistently perpetrated in recent decades.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. We thank an anonymous reviewer for bringing this issue to our attention.

2. Bueter’s work can be subsumed under the wider rubric of Wanderer’s (2017) structural testimonial epistemic injustice, which pinpoints not the agential, but the structural-institutional barriers to the participation of rightful knowers in epistemic activities.

3. It is admittedly difficult to see how J. N.’s psychiatrists’ course of action could simultaneously be considered both reasonable based on the initial probability of her statements being true and “a mistake”.

4. This seems to us to potentially contradict Kious et al.’s earlier thesis about the epistemic significance of mental illness. We will return to this point in the next section.

5. The question here is neither about the comparative worth of first-person vs. third person information in general, nor about the capacity of the latter to capture the former. These questions have been debated in the literatures on introspection and consciousness and lie beyond the scope of this paper (for views on these topics, see respectively Block, 2011; Phillips, 2016; Prinz, 2004; Schwitzgebel, 2008). What we claim is simply that the subject having an experience is the primary one to have access to what it is like to be them. Even if others can acquire all knowledge there is to be acquired about our experience once we fully describe it, it is still only us that could provide this description about our internal states in the first place, therefore we still have privileged access to this information in a way that no one else does.

6. Though not without exception, see Brown (2014).

7. Spencer (2023) has proposed to broaden the concept of epistemic injustice to apply where the communications of non-verbal people are disregarded or overlooked due to identity prejudice.