Misgendering as epistemic injustice: A queer STS approach

Misgendering is perceived as the use of incorrect pronouns and gender categories when addressing Transgender and Gender Non-Conforming (TGNC) people. This common habit is widely observed in reports, surveys and assessments, where the pressures to comply with a binary understanding of gender are high and alternative options for self-identification are not frequently offered. The present study reads misgendering as a manifestation of epistemic injustice, and uses resources from Science, Technology and Society (STS) Studies in order to highlight the importance of situated perspectives on the matter. After being analysed as a commonplace microaggression, misgendering is conceptualised as an act of structural hermeneutical marginalisation that is not usually intentional but product of society’s lack of sensibility towards gender diversity. Finally, a reparative approach against misgendering is offered through the Gender Fair Language model, which involves relational and situated contributions in order to prevent already marginalised people and experiences from being further excluded.

Transgender and gender non-conforming people (TGNC) have a gender identity that does not coincide with the sex they were assigned at birth (American Psychological Association, 2015;Mizock & Lundquist, 2016). It is an umbrella term that encompasses people who pursue bodily transitions in order to present themselves in a binary way (trans men or trans women), people who transition only socially, and people who prefer not to reach a self-identification inside the binary. In this spectrum of experiences and subjectivities (Hyde et al., 2018;Magliozzi et al., 2016), there is a higher possibility of being addressed, named and characterised in impertinent ways. This act of inadequate addressing has been called misgendering.
Misgendering is a relatively new concept that has been described by literature (mostly Anglophone) as the use of improper pronouns and grammatical genders to refer to people who belong to the TGNC umbrella (McLemore,1 Departamento de Ciencia, Tecnología y Sociedad, Instituto de Filosofía, Consejo Superior de Investigaciones Científicas, Madrid. Email: konstantinos.argyriou@cchs.csic.es -ORCID ID: https://orcid.org/0000-0002-0578-7960 Las Torres de Lucca. Revista internacional de filosofía política ISSN-e: 2255-3827 https://dx.doi.org/ltdl.76464 DOSSIER 2016). It is hard, if not epistemically false, to describe the phenomenon in a generalised manner, since each TGNC person has a unique appearance, self-expression, and life trajectory. Moreover, as shall be described below, there is a wide variety of manifestations of misgendering, which makes both the concept and the associated practice highly context-and culture-dependent (Hochdorn et al., 2016).
Everyday life of TGNC people is filled with examples of invalidations of the kind, as misgendering is generalised and persistent. By not complying with traditional aesthetic and social gender norms, a non-binary person addressing a public institution may face discrimination and stigma, or even trigger reactions of distrust, suspicion or violence, both covert and overt. Misgendering can also escalate to denying access to certain gender-segregated spaces, such as restrooms, and is associated to minority stress, that is, to stress associated with exposure to discrimination for pertaining to a minority group (McLemore, 2014).
Importantly, though, misgendering is a phenomenon where language plays a central role. By ticking the boxes male or female, TGNC people are denied the possibility to be properly recognised, because linguistic and social categories of gender do not properly apply to them. Misgendering is therefore an issue of structural exclusion. 2 Admittedly, the greater the power of linguistic conventions in a specific context, the harsher the effects will be on TGNC people's sense of self. In contexts such as public administration, the workplace or a health care provider's office, TGNC people face unique types of discrimination (Freeman & Stewart, 2018;McKinnon, 2019). Language use can and should be addressed at a structural level, since it is a locus of epistemic inequality.
The growing tendency inside transgender communities to opt out of strict binary identifications is still met with particular suspicion in today's societies -at least in those of Western gender tradition. Under these circumstances, people with alternative gender identities and gender expressions are highly likely to face exclusion and discrimination (Serano, 2013). Even when exclusion is not completely manifest, it takes the form of microaggressions, subtle expressions of verbal and emotional violence that may even go unnoticed (Freeman & Stewart, 2018;Nadal, 2018;Lilienfeld, 2017).
Epistemic injustice is a relevant notion, both to (mis)recognition and to microaggressions, that can frame research on gender diversity under novel milieu. Academic works on epistemic injustice have largely used racial discrimination and female harassment as focal examples in order to strengthen and expand their theorisations (see Dotson, 2012;Fricker, 2007;Medina, 2012). Albeit meticulously, matters of gender diversity have been examined less (see Aultman, 2016;McKinnon, 2019). For instance, Lauren Freeman (2018;Freeman & Stewart, 2018 has covered epistemic injustice regarding gender identity in clinical practice, though without making sufficient remarks over the diverse subjectivities and experiences self-identity encloses. The present paper wishes to exhibit the current debates on how to provide justice for people who do not define themselves in a binary manner. By attaching discrimination in terms of gender identity and gender expression to epistemic injustice, it is hoped that researchers and activists will be provided with a helpful analytical and methodological tool. Using Miranda Fricker's notion of epistemic injustice (2007) and Eve Kosofsky Sedgwick's notion of reparative reading (Hanson, 2011;Kosofsky Sedgwick, 1997;Walker, 2006), I shall describe a process of epistemic reparation that begins in language and spans to several other mechanisms of structural, societal, and cultural transformation. For that purpose, I review methods of addressing misgendering and of introducing inclusive methodologies of gender self-definition. A queer Science, Technology and Society (STS) methodology (Latham, 2016(Latham, , 2017 is employed for such purpose. STS approaches focus on situated knowledge, intersectionality and critical ways of understanding sociocultural structures (Adrian et al., 2018;Campbell, 2009;Haraway, 2016;Pérez Sedeño, 2019).
The paper is divided into four main sections: first, the explanation of the mechanisms that surround misgendering; second, the reading of misgendering as a form of hermeneutical injustice; third, the forms of misgendering observed in the completion of surveys, questionnaires and official documents; and fourth, the analysis of the Gender-Fair Language Model, introduced by Sczesny et al (2015,2016), as an effort to bring epistemic justice into the realm of gender classification -at least until the practice of demanding that information is rendered obsolete. In the final discussion, issues of power dynamics and situatedness are reintroduced through a comprehensive lens.

Misgendering in a normative world
Three issues directly connected to misgendering are misrecognition, misidentification and misrepresentation. The first refers to denying a TGNC person their right to self-determination and self-presentation (Mizock & Lundquist, 2016); the second, directly associated to misrecognition, to rendering this right illegitimate (Butler & Athanasiou, 2013); lastly, misrepresentation is the description of TGNC people as deviant and dangerous within justice systems, mass media and social life in general (Juang, 2006). The three mechanisms may stem from a variety of statements: some are deliberately harmful, others well intentioned but still prejudicial, while others seemingly innocuous but eventually impertinent (Lilienfeld, 2017). This variety of intentions calls for a careful examination of each particular context.
In cisheteronormative societies, that is, in societies where normativity is dictated by compulsory heterosexuality and cisgenderism, the mechanisms of recognition, legitimisation and representation are based upon binary gender 2 Structural exclusion is understood here as an embedded linguistic order that goes beyond social agents and institutions. assignment, cisnormativity -the belief that people have their sex assigned at birth and gender identity aligned by default-, and other similar ideals rooted in prejudice (Aultman, 2019, Lombardi, 2007Serano, 2016). In this constellation, genderqueer, nonconforming, intersex or agender people are practically nonexistent and pushed to the margins of intelligibility.
When an identity is misclassified, or someone's sense of self is not being properly recognised, the psychological implications that result from it can range from disruptive to traumatic (McLemore, 2014). Devalued identity is rendered illegitimate, thus unintelligible. Although frequency of misgendering does not seem to interact with negative affect, the sense of being stigmatised while being misgendered does (McLemore, 2016). To satisfy one's need for coherence is essential for their well-being, even in cases where gender seems to fluctuate more, as is the case of gender fluid and genderqueer identities. The concepts of current gender identity and gender nonconformity cover this issue, while securing that oscillations are not perceived as vices or plays.
Accordingly, non-conforming and genderqueer individuals are misgendered at a more frequent rate than transgender men and transgender women (McLemore, 2014). The ambiguity, fluidity and transgression associated with these alternative gender expressions is seen as a threat to normative intelligibility mechanisms (Butler & Athanasiou, 2013). Normativity, by the same token, is not solely understood as an external system of oppression, belonging exclusively to the cisheteronormative imagery. On the contrary, it permeates in the transgender community as well. Austin Johnson (2016) theorises transnormativity as a pervasive assumption that the only valid form of transgender identity resides in committing oneself to a full bodily transition (with hormones, gender confirmation surgery and aesthetic interventions), and to embracing traditional masculinity and femininity in terms of self-expression.
The strong social desire to fixate trans as a category with clear definitions reveals an underlying notion that cis and trans people have different characteristics -although overlaps as to what constitutes masculinity and femininity are significant between the two groups (Johnson, 2016). Used as one word, transpeople becomes a noun, and marks a difference from trans people, where trans functions as an adjective. In the first case, transpeople seem to be a group or community that is so compact, that they acquire the value of a cohort (as in an ethnic group). On the contrary, trans as an adjective emphasises a priority, a characteristic, an added value to the noun people, thus not providing cis people with any overt legitimation. Definitions of trans subjectivities are condemned to be "perpetually provisional" (Aguirre-Sánchez-Beato, 2018, p. 300); however, this is not necessarily a negative issue. It becomes problematic only because the cisgender mechanisms that control gender expression equate "provisional" with "unintelligible". The term TGNC is employed here partly in order not to perpetuate this dialectic of epistemic marginalisation.
Instead of focusing on linguistic categories, Annemarie Mol connects normativity to the materiality of bodies: Epistemological normativity is prescriptive: it tells how to know properly. The normativity of ethnographic descriptions is of a different kind. It suggests what must be taken into account when it comes to appreciating practices. If reality doesn't precede practices but is a part of them, it cannot itself be the standard by which practices are assessed. But "mere pragmatism" is no longer a good enough legitimization either, because each event, however pragmatically inspired, turns some "body" (some disease, some patient) into a lived realityand thereby evacuates the reality of another. (2002, p. 6) The prescriptive or indicative character of normativity, either external or internalised, either cisheteronormativity or transnormativity, is what renders it unbearable for those who fail to reproduce it "correctly". Bodies, as well as experiences and identities, are condemned into a constant quest of convincing society that they can fit pre-established categories, known as the deception frame (Billard, 2019;Fricker & Jenkins, 2017). This frame is double edged, since it inflicts damage both when TGNC people fail to pass as cisgender, and when they do achieve it. When they fail, they deceive because they seem "not good enough", so they expose themselves to corrective comments. Alternatively, when they supposedly succeed, they incite fear and danger, since they seem to commit fraud. The frame is actually triple-edged, since whichever middle ground between failure and passing is also subject to vigilance.
In the normative framework described above, deception expectations augment the possibility of committing microaggressions. Literature on microaggressions has been critical in better understanding misgendering. Initially used in the field of Black and Decolonial studies (Pierce, 1970;Sue, 2010), the term microaggression progressively expanded towards other collectives in the last five decades (see, for instance, Nadal, 2018 for LGB groups; Yang & Carroll, 2018 for women). Sue's (2010) almost classic division into three types of microaggressions, microassaults, microinsults and microinvalidations could possibly parallel the notions of misidentification, misrecognition and misrepresentation presented above, respectively; although empirical evidence is required to defend this hypothesis. Paradoxically, according to the author, the greater ambiguity of both microinsults and microinvalidations makes their effects more harmful to recipients than those of microassaults. 3 Misgendering might fit under those two subcategories, despite the fact that neither microinsults nor microinvalidations reflect the precise outcome of the practice on recipients. McLemore's (2014McLemore's ( , 2016 understanding of misgendering as comprising of verbal, behavioural and environmental elements coincides with Nadal's (2018) understanding 3 Microassaults are derogations, acts of conscious avoidance and overt discrimination, since they are the most direct of the three. Microinsults consist of rude and demeaning comments, while microinvalidations are excluding attitudes that permeate behaviour. of microaggressions. Bringing the two terms closer helps to conceptualise traumatic experiences through a situated lens that is of particular interest. For instance, retraumatisation is salient in both misgendering and microaggression. Triggering situations of the present activate past memories and create a vicious circle of epistemic entrapment, where TGNC people relive the same pattern of epistemic marginalisation.
The term microaggression has also received critiques as to whether it properly unveils concealed violence or instead serves to further conceal it. Pointing to its oxymoronic nature, Scott Lilienfeld comments that "the term ʽmicroaggression' conflates the outcome of a behavior with its intent" (2017, p. 161). He also suggests that microassaults and acts of severe prejudice be dropped from the "microaggression umbrella", for aspects of this paradox could affect educational programs that use the notion in order to provision social justice. Freeman & Stewart (2018), on the contrary, provide arguments in favour of its use. For them, it achieves: a) manifesting the unique dynamic nature of those aggressions, b) capturing their potentially unintended character, c) separating them from deliberate transphobic acts, and d) conforming to the already existent literature which endorses the term. For what ties it to misgendering, the term is maintained throughout the text.

Misgendering as a form of epistemic injustice
Misgendering can be seen as a form of epistemic injustice, as TGNC people are wrong-ed precisely in their capacity as knowers of their own gender identity and expression (Fricker, 2007). Epistemic violence in the case of misgendering comes from the failure of cisgender hearers to recognise the injustices made against TGNC speakers, when the latter enunciate that their gender identity is not correctly or sufficiently treated (Dotson, 2011;Newbigging & Ridley, 2018). Miranda Fricker's (2006 theorisation of epistemic injustice highlights the importance of understanding power as a dyadic and highly context-dependent relation. Her definition of social power entails sharing a context of common meanings, which allows for effective control. It is not obligatory for the sharing concept to be truthful; what it needs to be is imaginatively coordinated. Like all social power, identity power is not solely agential, but can be structural as well. Operations of social power that involve the categories of man and woman, for instance, rely on collective gender stereotypes. To exercise the identity power in terms of gender, however, does not only mean for men to actively patronise women; power can also be inflicted upon women in an unconscious or passive way, through silencing. The same can be sustained for the binary cisgender/transgender: while cis people may actively exclude trans people from the public sphere as abnormal, pathological, or deceptive (as explained through Billard, 2019), they can also exclude them by rendering them invisible in subtler, everyday practices. Kevin McLemore (2016) found that the closer to the offender the TGNC person is located socially, the more impactful felt stigma will be. In other words, being misgendered by family members and close friends can be subtler but its impact is significantly higher for the receiver. This also attests to the insinuated character of microaggressions. Fricker (2007) distinguishes two types of epistemic injustice: testimonial and hermeneutical. The first type questions a person's capacity as a knowledge producer -Fricker mentions Harper Lee's Tom Robinson case in To Kill a Mockingbird as a representative example. In testimonial injustice, identity power is casted through affecting the hearer's credibility judgement with identity prejudice. Even though testimonial injustice is not always a product of systematic prejudice, incidental cases are not less unethical by being called "incidental". If injustice is diachronic, incidental testimonial injustice can be as unfair as systematic testimonial injustice. Advocacy for testimonial sensibility promises to improve hearers' credibility but may not challenge its prominence as such (Dotson, 2012).
In the second type, a person is hermeneutically marginalised through being deprived of verbalising or rendering intelligible their experience to other people. That means lacking the conceptual or linguistic resources for processing their experiences, being unable to formulate them in order for others to acknowledge them, or even both. Being unable to make one's experience intelligible is double-edged: it involves not knowing how to put a name to it, and facing other people's lack of comprehension. On the other hand, others are not simply receptors of this misunderstanding, but actively contribute to its formulation: "certain material advantages will generate the envisaged epistemological advantage-if you have material power, then you will tend to have an influence in those practices by which social meanings are generated" (Fricker, 2007, p. 147). Hermeneutical marginalisation implies power dynamics of privilege and subordination.
Fricker uses identity prejudice as the key mediator between testimonial and hermeneutical injustice, in order to explain how a certain social group's credibility (e.g. that of TGNC people) can be rendered elusive. Compounded by testimonial injustice, hermeneutical one may fail to be recognised. In other words, when society does not perceive a TGNC person as credible or intelligible in terms of gender, it is more likely to assume that what fails is the person's capacity as knowledge producer, than to name the systematic oppression that has outcast TGNC people's identities. Structural marginalisations are harder to articulate, thus easier to deny.
Misgendering can be understood both as a testimonial injustice and as a hermeneutical one. In the first case, TGNC people are judged for their capacity as knowers of their own gender. Using Xeph Kalma's case as an illustrative example, Freeman & Stewart (2019) introduce testimonial betrayal through a series of self identity microaggres-sions in the emergency room. In terms of hermeneutical marginalisation, gender diversity is not represented in the mainstream collective hermeneutical resources of classifying gender. Therefore, it is not within a cognitive landscape that cisgender society does not comprehend gender diversity, but within a collective code that systematically operates to refuse gender diversity, drawing on Kristie Dotson (2012).
While Fricker (2006) distinguishes once again between structural and incidental (one-off) hermeneutical marginalisation and injustice, between unjust constitutive constructions of identities and unjust casual constructions, Dotson (2011) makes a similar division between instances of silencing and practices of silencing. In Dotson's account, reliable ignorance is what makes practices of silencing potentially harmful, because it prohibits the epistemic agent from reaching the truth. When harm is actually done, then ignorance becomes pernicious.
On the contrary, José Medina (2012) considers that to separate communicative from epistemic analyses of injustice creates a false dichotomy. Credibility and intelligibility are inexorably intertwined, thus in silencing, contextual and structural agency are not distinguishable, or even if they are, we should not be confronted with it. Medina's argument that hermeneutical gaps do not affect all members of a social group at the same level and that counter-interpretations of intelligibility are highly context-dependent is his main point of departure from Fricker's all-encompassing theorisation, even though she also highlights the importance of context. He, however, considers Fricker's approach "insufficiently pluralized" (2012., p. 206).
The particularity of communicative dynamics is a relevant point, are we to approach TGNC people's materiality through a closer look. In transgender studies, this emphasis on the plurality of experience is understood through the proliferation of umbrella terms (transgender, trans*, 4 TGNC, genderqueer, among others), as well as through the recent shift towards new materialities in order to go beyond the essentialist-constructivist debate (Aultman, 2019;Halberstam, 2018). As has been already mentioned, TGNC subjectivities and experiences are highly diverse, thus very hard to enclose in rigid terms. Presupposing a clear-cut transgender identity negatively impacts the quality of resources and services destined for TGNC people and their struggles for legal and social recognition. As Rita Felski states (2006, p. 572): "Not all social subjects, after all, have equal freedom to play with and subvert the signs of gender, even as many do not perceive such play as a necessary condition of their freedom". In addition, by standardising transgender as a linguistic and social category, the risk of forgetting that narratives shape selves, but also that selves shape narratives (McAdams & McLean, 2013), is augmented. Any attempt to address gender diversity through the framework of epistemic injustice should therefore care to incorporate a flexible and situated understanding of TGNC (Latham, 2016).
Understanding of misgendering as a form of epistemic injustice is not new. Rachel McKinnon (2017) makes similar remarks on the mispronouncing of a trans woman, although instead of framing the act as misgendering, she uses the word gaslighting, alluding to the 1944 film that introduced the notion. I, however, would consider gaslighting a more subtle mechanism of manipulation and devaluation, and misgendering a direct, everyday, proclaimed, painful reminder of society's unbending ties to cisnormativity. Misgendering is almost instantly detectable in people's discourse, a clear sign of misrecognition, while gaslighting appeals to doubt.
Additionally, McKinnon reads the exclusionary practice in question through the lens of credibility deficit, therefore as testimonial injustice. A TGNC person claiming to self-define their gender in a society where this definition is a natural given challenges their value as testifiers. I, on the other hand, prefer to understand misgendering as a kind of hermeneutical marginalisation, since TGNC people are deprived of the social resources that legitimise gender and are forced to the margins by having their claims labelled nonsensical. The structural dimension of misgendering is seen in McKinnon's observation that often, cisgender people tend to justify their discriminatory behaviour pointing to their identity as allies: "I am a trans ally, but...". In order to resolve the tension, below I will try to appoint to Medina's intertwining between testimonial and hermeneutical injustice, applying misgendering to surveys and reports.

Misgendering in surveys, documents and questionnaires
Misgendering is not only an issue in social interactions, but also in more institutional parameters of everyday life. The main distinction that differentiates the experiences in each context is that passing and the deception frame operate on a visual basis, and affect recognition in a direct manner. On the contrary, to address a healthcare service, to talk to a healthcare provider, or even to fill an application form for them, implies a series of power dynamics which are not reduced to the visual domain. By failing to embrace TGNC people's self-determination, healthcare providers commit self-identity microaggressions (Freeman & Stewart, 2018. By emphasising their expert knowledge, they might discredit or devalue the lived experience of TGNC people, roll their eyes, interrupt or show ignorant behaviours. These interactions are not heavily documented, unless research studies or institutional archives give prominence to them. Ironically, these are usually their structural equivalents, as well. Misgendering, deadnaming and microaggressions can be reflected in reports by healthcare providers, protocols and other documents used for administrative 4 Although increasingly falling into misuse, the asterisk was initially introduced in an attempt to expand trans identities beyond the binary, as the asterisk in search machines is used as a prefix that multiplies results by incorporating all possible suffixes. In terms of transitioning, that may be translated into refusing (binary) finishing lines or linear procedures. purposes. 5 In Western culture, sex, gender, and sexuality are common denominators in form fulfilment. Most medical forms maintain this categorisation in order to "correspondingly" classify assessments, treatments, interventions and statistical data (Freeman, 2018) and shape this information as ahistorical (Spade, 2015).
Currently, biomedical survey data are gauged only by social behaviours of males and females, and only as agents of maleness and femaleness (Magliozzi et al., 2016). In addition, processes such as bodily transitions are often hidden behind identity categories, which may render societal markers restrictive. For instance, a trans person undergoing hormone therapy might identify differently at the beginning of the process and after months of systematic administration. Especially for gender minorities and people who address healthcare services, data acquisition through official demographic statistics has important policy effects. In that sense, important research gaps and omissions should be addressed, since TGNC identities are often excluded from health-related studies as statistically irrelevant (Ryan, 2019).
Nonprobability samplings through the Internet in national populations and needs assessment studies in regional levels are two methodological tools that have been widely used in previous research in order to respond to trans specific issues, making it particularly hard to generalise results to a wider population (Ryan, 2019). This problematic management of variations in trans experience in quantitative surveys makes it harder to generalise results, as some health conditions might be overrepresented, while others conflated, undervalued or silenced (Freeman, 2018).
In view of this, I argue that misgendering in quantitative surveys is not inflicted by a specific agent, that being the creator(s) of the survey, or at least not only, but rather is the result of a collective hermeneutical convention whereby gender shall be represented in a common binary manner. This symbolic lacuna creates a systematic case of hermeneutical marginalisation, to which subjects may learn to adapt through false-reporting and other evasive, compliant and unreliable mechanisms, while also residing in disbelief (Freeman, 2018;McKinnon, 2019). Moreover, although I seem to be moving through many different linguistic environments, such as the statistical, the official, the administrative, the scientific and the lay one, I do not wish to carelessly extrapolate the same terminologies or language uses to different contexts. I do wish to demonstrate how conventions homogenise gender nomenclature into one solid system of perfect alignment.
Reporting a gender that does not correspond to the current sense of self is not only uncomfortable for the person; it may actually prove risky, especially if health needs are at stake (Bauer et al., 2017). In many cases, distinctions between sex and gender are not even made at a conceptual level (Aguirre-Sánchez-Beato, 2018), thus subjects are often given only two opposite and exclusionary choices enclosing both sex and gender: male and female. More often, surveys and questionnaires do not actually make use of this segregation; since it is considered a fundamental social category, sex is the preliminary question offered for answer. What transgender people answer in the sex question, however, is widely unknown; some might fill the gap reporting their sex assigned at birth, while others with their felt gender identity. In some cases, both or neither could also be ticked (for more information on the matter, see Eliason, 2014). The level and progress of each transition, the subjective meaning each person attributes to transition, cultural and contextual factors, as well as the level of adherence to the binary gender systems, risk being lost in search of homogeneity and normative answers. Consequently, not counting gender diversity can be detrimental, but counting it bad can prove even worse (Ryan, 2019). Some individuals that have undergone a complete process of medical reassignment do not appear to confront the same discrimination as other TGNC people do, independent of whether they self define as transsexual or not. 6 These individuals might not feel the same distress when fulfilling an official document, given that they align with the binary conceptions of gender. Given the coercive power of hermeneutical marginalisation (Fricker, 2006), in this case, it could be sustained that transsexual people opt out of the present discussion. The authoritative binary mechanism of identification does not affect the "constitutive construction" of their identity, since they may have had their personal documents modified already, and they belong to a certain category "more safely" (Fricker, 2006, p. 106).
Many demographic questions do not yet include gender-sensitive questions that distinguish between sex assigned at birth, current gender identity, and sexuality (Eliason, 2014). By performatively repeating the act of filling out forms and questionnaires, several intersex people are faced with the traumatic implications of their external assignment; thus they are also asked to reinforce the binary sex classification that they were also obliged to accept through the initial assignment (Bauer et al., 2017). This structural hermeneutical marginalisation restricts intersex people's resources and threatens their sense of self.
Moreover, attempts to include supposedly "statistically insignificant" questions to questionnaires are met with suspicion. Arguments range from simple ignorance to unnecessary complexity, and create a vicious circle where "dominant epistemic agents are calling for a guarantor for the need for new epistemic resources, but they disqualify marginalized knowers, the very persons whose experienced world reveals the inadequacy of current epistemic resources" (Pohlhaus, 2011, p. 729). The problem lies, therefore, on what reports of TGNC people assessors can and do depend upon, given that transition is a moving process, rather than a homogenous experience. 5 Deadnaming is the use of the antiquated birth name of a TGNC person in order to address them, without their consent. Ansara & Hegarty (2014) associate it with what they call rectroactive misgendering, when healthcare professionals disregard present identifications in order to focus on past realities of TGNC people. 6 Transsexual is a term that, despite its derogatory connotations, has prevailed in literature when describing full bodily transitions, while transgender is a term that does not necessarily designate hormonal or surgical interventions.
By asking TGNC people to collaborate in their own misclassification, we punish them for being misfits and for not having the hermeneutical privilege to identify their gender in the forms society considers correct, while, simultaneously, we deprive them of gaining access to useful resources. Dean Spade (2015) reminds how misclassification perpetuates gender segregation by "securing" services, institutions and spaces, and alerts that, when particularities are ignored, assimilationism and inclusion politics only benefit privileged people. Reparative strategies such as language inclusion should address structural injustices but also prevent further marginalisation and further exposure to microaggressions.

The Gender Fair Language model (GFL)
In this section, I borrow Rachel McKinnon's idea that grammatical policing is strategic (2017, p. 287), 7 in order to promote a reparative pronunciation that goes against misgendering, if not in everyday language use, at least in situated social research. To correct research terminology is not only about reviewing our ways of acknowledging our research participants, but about understanding that language can be a site of doing justice in the same way we use it to harm: through repetition, imitation and pressure (Sczesny et al., 2015). Situatedness is related here to being held accountable for our articulations of what we conceive as sociolinguistic reality (Haraway, 2016;Lykke, 2010). Epistemic responsibility, at least under a critical feminist standpoint view, emphasises the inevitability of parciality and locatability in the production of knowledge, although that does not mean that knowledge can be reduced to a mere reflection of contextual values (Campbell, 2009).
The Gender Fair Language model (GFL), introduced by Sczesny et al. (2016) as a measure in order to abolish asymmetries in everyday and institutional language, represents an example of how a systematised attempt to introduce more inclusive linguistic forms can lead to the unmasking of hidden biases. For instance, the use of more inclusive gender forms has revealed how we often misinterpret male forms as a sign of neutrality, and how effortlessly we associate certain nouns or concepts with maleness, and others with femaleness (for example, in occupations). The use of he as generic is not a particularity of the English language; it can rather be used in order to account for generalised androcentric biases in various Indoeuropean languages, which consider men the default subject and let them benefit from language's disparities (Sczesny et al., 2015).
The ways GFL is interpreted by users needs careful examination. Though it might be expected that those who benefit from it are more accustomed to it, findings show generally mixed attitudes. That leads Sczesny et al. (2016) to suggest that a consistent finding is that speakers do not understand masculine forms as referring to both genders equally but that they interpret them in a male-biased way. This underscores the importance of implementing GFL in everyday language and of using it consistently, so that speakers take up this usage in their own texts and utterances. (2016, p. 6) Incorporating alternative gender forms in everyday discourse, however, does not resolve linguistic inequalities. If institutions themselves do not make an effort to officialise or materialise these forms, in public documents, surveys and speech, lay discourses are insufficient bearers of a paradigmatic shift. This paradox can be observed in the authors' finding that women have a tendency to accept male forms in job search. The generalised insinuation that female terms are socially devalued or exceptional, drives a significant amount of women into denouncing their use. The hierarchy behind not only gendered terminology itself, but also behind the use of it, may be proof of the idea that a top-down change in attitudes needs to mirror the described bottom-up motion. Furthermore, this attitudinal change need not seem merely apparent, but substantial.
Similarly, an affective component behind society's resistance to incorporate alternative gender forms and gender fair language is ascertained. That component seems to be unwillingness, in the sense of indolence, and is attached to a cognitive refusal to accept transition as a legitimate social process (Aultman, 2019). Since language is viewed as a robust system of signifiers resistant to change, GFL associated to nonbinary and genderqueer identities is thereby viewed as an intruder that encourages linguistic and epistemic instability. A different understanding of genderqueer individuals, not as language intruders but rather as language revisionists, might be beneficial: "Because nonbinary identities do not, strictly speaking, rebel against the gender binary, the capaciousness of nonbinary as a descriptor might help redefine what counts as the limits of visibility" (Aultman, 2019, p. 10). The boundary renegotiation behind this argument is consistent with what has been described earlier regarding misrecognition and unintelligibility.
J. Michael Ryan (2019) examines three main solutions to the reporting issue. The first approach is the cross classification paradigm, with a two-step measure that helps corroborate whether sex and gender coincide. The second one relies exclusively on the respondent's identification, since it asks directly if the person is trans or not. The third one, the trans response, offers trans as a third option aside from male and female, risking misidentification of transsexual individuals. Bauer et al. (2017) propose an extended, multidimensional measure to counter the two-step option, which better captures the particularities of gender diversity but may also result tiresome to people who do not have so much detail to offer in their self-identification. Both studies revolve against choices such as it's more complicated or other, as these generate unnecessary inconsistencies and misclassifications.
A series of guidelines developed by Bouman et al. (2017) focus on the importance of avoiding unnecessary conflations between sexual orientation and gender identity, using inclusive instead of obsolete and stigmatising terms, and paying particular attention to translation and adaptation practices, since they frequently perpetuate a hierarchical, Western and exclusionary tradition of falsely universalist discourses. Among the guidelines, and even though they are only applicable to the English-speaking community, there are concrete recommendations purported to eradicate discriminatory language and to renovate biomedical and psychosocial academic discourses with inclusive terms. For instance, instead of using rigid categories with negative and psychopathological connotations, such as transvestitism, newer and more flexible vocabulary is suggested, promoting gender diversity instead (for example, gender creativity, body variation, current gender identity). Such guidelines have a recommendatory attitude and are not obligatorily imposed. They consist of enumerated suggestions for practice and action, with a politically correct innuendo. Nonetheless, they do not cover for every single potentially unjust communicative setting.
As Medina reminds (2012), what feminist and queer theorists have proven is that what passes as obviousness and transparency concerning the categories of gender, often masks a severe lack of awareness. In view of the aforementioned guidelines, one of the main premises of a GFL project is reconciliation with ignorance, and also with error. A new space may be furnished for progressively substituting ignorance with humble acknowledgment of blind points. Under this view, mistakes in addressing gender diverse people can be seen as not necessarily offensive, but rather as an opportunity for correction. In Sedgwick's account: to read from a reparative position is to surrender the knowing, anxious paranoid determination that no horror, however apparently unthinkable, shall ever come to the reader as new: to a reparatively positioned reader, it can seem realistic and necessary to experience surprise. Because there can be terrible surprises, however, there can also be good ones. (1997, p. 24) This framing does not hope to dismiss epistemic responsibility under the guise of surprise. For reparation to take place, people need to eventually assume that, even benevolently or unconsciously, they reproduce language as sexist, as cisgenderist, as trans-exclusionary (following Sczesny et al., 2015). Fricker herself acknowledges that hermeneutical justice is corrective by definition (2007, p. 169). Its a posteriori value intermingles with the renunciation of privilege. When linguistically and epistemically privileged people stubbornly ignore the injustices they commit, it is not simply a sign of bad luck, as Fricker seems to suggest, but rather an integral part of a pattern of how injustice works. Hermeneutical responsibility requires openness, acceptability of difference, and relationality (Medina, 2012(Medina, , 2013. To acquire certain epistemic resources, such as the new terms and options for self definition described above, means to break the dominant codes of knowledge construction. This is an important step towards achieving epistemic justice, though deeply embedded social stigma is harder to repair. The reparative procedure continues with the progressive adaptation of these new resources by the cisheteronormative society. Once familiarised, the mechanisms and institutions that have previously served to exclude these self definitions are faced with their responsibilities. Restoration is therefore twofold, highlighting the privileged groups' (here, cisgender society's) responsibilities in hermeneutical silencing and the marginalised groups' (here, TGNC people's) value as survivors (Newbigging & Ridley, 2018).
Furthermore, in order for restoration to be substantial and to avoid collateral epistemic trespassing, GFL shall benefit from taking into account the constant changes in terminology and felt experiences, and from learning to ask TGNC people first. Embodied narratives of subordination are valuable resources, not in the additive sense, but as signs of the transformative power of lived experience. Drawing on Bristol & Rossano (2020), institutions should solicit epistemic permissions from TGNC people, since they are the ones who have the discernment and authority to decide which is the most adequate way to be asked for their gender identity. In each particular circumstance and depending on what aspects of gender identity, gender expression or other personal characteristics are assessed, questions and formats should be customised.
Here, situatedness might damage generalisation, since each TGNC community may use different terms and each language is structured in different ways in terms of grammatical gender, as has been explained. Standardised questionnaires for extensive, transnational use would be hard to implement. Furthermore, there is no established consensus concerning the best response format in self-reporting of gender identity. Ansara & Hegarty (2014), for instance, consider that the tick-box options are generally restrictive, since they offer a pre-emptively fabricated conception of sex and gender. In that sense, open ended formats are preferable, unless the closed ended formats allow for multi-step identification (measuring sex assigned at birth, cis or trans status, current gender identity, etc.).
However, the presence of widely agreed-upon guidelines such as those of the APA (2015), which explicitly specify the need for inclusive language, aids the process of customising. Therefore, even if terms vary culturally and widespread uniformity is a rather complicated objective, the logic of inclusion and the call for user-friendly tools that do justice are feasible principles. Wholly, consensus on methods necessitates a willingness to promote the TGNC community's inner representations. Avoiding TGNC people's othering and epistemic violence through ignorance and superficial allyship are main issues to be addressed.
Lastly, since transnormative assumptions have both an external and a internalised part, epistemic justice also involves alliances in order for TGNC people to cast aside such assumptions. Fricker & Jenkins (2017) call for stepping-stone concepts that bridge the need for fixing manifest concepts on one hand, and the provision of operative concepts, on the other. These provisional and dialogical meeting points between cisgender society and TGNC agency reduce misgendering by allowing new conceptual spaces and resources -mirroring Medina's (2013) insights to lucidity's capacity to enrich social cognition and overcome blind spots.

Discussion
The main argument I followed throughout the present text is that, by remaining within specific dynamics that block alternative forms of acknowledging communicative difficulties, we contribute and coperpetuate hermeneutical injustice (Medina, 2012). We become complicit in epistemic marginalisation by responding to it deficiently or by failing to respond at all. In Dotson's words (2011), apart from a narcissistic wound for our communicative dependence, this suggests a failure of communicative reciprocity. Transgender and gender nonconforming people are particularly vulnerable to this reciprocity deficiency, since their narratives remain widely unintelligible.
Nevertheless, framing misrecognition as injustice is not novel. For instance, Mason-Schrock et al. (2004) mention that 1990s local support groups in middle-class US transgender communities would use this strategy in order to recruit new members. What is novel, instead, is the expansion of terminology on subtle types of injustice in the last few decades. As has been demonstrated with discussion of microaggressions, several mechanisms of misapprehension (misidentification and misrepresentation among them) serve to delegitimise TGNC lives. Relevant is also the understanding that by being granted control to self definition, TGNC respondents are offered a possibility to do gender in their own manner (Magliozzi et al., 2016), or to reclaim their agency as term composers in the collective repertoire (McKinnon, 2017).
My argument has also focused on the situated character of epistemic marginalisation. Easy distinctions between marginally situated people and dominantly situated people ought not to be seen as a strict binary, but rather to take into account the intersectional and contextual character of social identities. Inside trans communities, different ways of expression and identification, individual and collective resources, opportunities to pass as cisgender, access to supportive networks and healthcare are, among many other factors, decisive in rendering some experiences more dominant and others more marginal. Therefore, the binary distinction can be particularly problematic, for it may lead to secondary marginalisation and obscure knowledge that comes from alternative pathways.
As recent epistemologies of ignorance have emphasized, it is not always the case that hermeneutical gaps render experiences unintelligible for everybody equally and in every communicative dynamic. As epistemologists (such as standpoint theorists) writing on inter-racial, inter-gender, and inter-sexual (mis)communication and (mis)understanding have emphasized, when we encounter hermeneutical problems in situations of oppression, it is of the utmost importance to keep in mind that a complex society often contains diverse publics with heterogeneous interpretative resources and practices (Medina, 2012, p. 210).
The variety of forms resistance can take depends on people's situatedness towards normativity. In the cisgenderist system we find ourselves immersed into, cisgender people (along some transsexual people) have the privilege to conceive of sex and gender as a naturalised entity. This normative stance often makes them ignorant against alternative gender positioning (Aultman, 2019;Johnson, 2016). Unwillingness to acknowledge other gender identities and lack of references create a vicious circle of suspicion, which reinforces cisgender people's safety inside their stable boundaries, while excluding nonconformists.
When this stability is questioned, the gender binary's entitlement to normativity is threatened. Thus, cisgender people's comfort with being intelligible, credible, and epistemically trustworthy is also threatened. Nonetheless, if cisgender people explore the costs of their own comfort, through a process Medina calls self-estrangement (Medina, 2013, p. 311), they might learn to assume their privilege differently. By acknowledging the hermeneutical marginalisation other people face because of their nonconforming identities and experiences, they help complicate their own positioning towards gender.
This study wished to sustain that this self-estranging process is reparative, for it not only provides justice to TGNC people for the marginalisation and microaggressions they have to tolerate, but also reconnects, or relocates, cisgender people inside the cisgenderist system. To negotiate gender identity not as persecutory ("some people try to convince us the non-normative is the normative"); but as reparative ("we can understand our own gender better, thanks to TGNC people") is crucial in this sense. This negotiation may as well begin through major systemic changes, as I suggested with the introduction of Gender Fair Language to administrative and official accounts of gender. It may still seem unintelligible for many to have to answer questions both on their sex assigned at birth and on their current gender identity and expression, but it may also cater for a new hermeneutical resource through which to thoroughly understand other people's realities.
GFL has been seen across the text as an epistemic resource of significant value for TGNC people. Throughout 2020, social media evidenced a major shift towards GFL worldwide, with massive incorporation of preferred pronouns in user names, zoom meeting IDs and Twitter hatches. The effort to integrate a new linguistic system of more "politically correct" forms necessitates a mere exposure effect, in order to overcome counter-arguments that tend to find it useless, hard or uncomfortable to read or to pronounce, or even unnecessary believing that equality might now be a given. As Sczesny et al. suggest: In all, there is a lack of transfer of scientific knowledge which prevents the understanding of linguistic asymmetries as part of a broader gender imbalance and hinders social change. Education and policy-making therefore need to increase the efforts of circulating new scientific insights about GFL to break the vicious circle of illinformed controversies and discussions about GFL. (2016, p. 7) Epistemic affirmative action promulgates justice by shifting attention from the speaker who is seen as dissident or impertinent to the context, to the hearers who are unable to understand. It is not the person who self-defines their gender through personal terms that is unintelligible, but rather the cisheteronormative society that is unwilling to embrace this process.
Relational and situated contributions notwithstanding, the intrinsic paradox of the concepts of GFL, epistemic reparation, microaggressions and hermeneutical marginalisation is that they are themselves epistemic resources (Pohlhaus, 2011). This is one of the major counter-arguments against Fricker's theorisation of epistemic injustice: people who are not familiar with the notions she accounts for, are condemned into a vicious circle of further epistemic exclusion. Terminological, cognitive and distributive limitations might paralyse restorative attempts; a TGNC person lacking metanarrative on their own gender diversity might be incapable of reporting their gender in an assessment. The same limitation undoubtedly impregnates the present reflexion as well, as several other terms have been invoked but not examined empirically. Hence, future research should incorporate aspects of TGNC respondents' familiarisation with terminology on microaggressions, epistemic injustice and GFL. Likewise, issues of term translation, adaptation and evolution should be contemplated.