2.1. Four diagnoses, one shared structure

Recent philosophical work on epistemic injustice in psychiatry has produced a set of convergent diagnoses about the exclusion of psychopathologized perspectives from psychiatric classification. The diagnoses differ in their theoretical frameworks and in the specific forms of injustice they identify, but they share a structural feature that is significant for the argument of this paper: each correctly identifies a problem and proposes a remedy that operates at the level of disciplinary self-revision without specifying the institutional conditions under which that remedy could be implemented. This is consistent with Anderson’s (Reference Anderson2012) argument that epistemic injustice is sustained by features of institutional design rather than by individual attitudes alone, and that institutional remedies require attention to the structural properties of deliberative processes. The accounts examined below identify the injustice but do not specify those properties for the case of psychiatric classification.

Bueter (Reference Bueter2019) argues that the exclusion of patients from DSM revision constitutes a case of pre-emptive testimonial injustice. Patients are excluded from taxonomic decision-making in virtue of their status as laypeople, on the assumption that they lack the expertise to contribute to scientific practices. Bueter shows that this assumption is misguided for two reasons. The role of values in psychiatric classification means that non-empirical considerations are already operative in the process, and first-person knowledge can function as a corrective against inaccurate or incomplete diagnostic criteria. The remedy she proposes is increased integration of patients and advocates into taxonomic decision-making.

Bueter’s analysis intersects with what Dotson (Reference Dotson2012, Reference Dotson2014) calls contributory injustice, the form of injustice in which marginalized speakers’ alternative interpretive resources are wilfully ignored by hearers who could access them. Where pre-emptive testimonial injustice excludes speakers in advance, contributory injustice operates on speakers who are formally included but whose contributions are filtered through hermeneutical resources that systematically discount them. Gagné-Julien (Reference Gagné-Julien2022) develops the contributory-injustice diagnosis for psychiatric medicalization specifically. The institutional conditions I develop below are designed to address both modes of exclusion.

Knox (Reference Knox2022a) shows that the problem extends beyond pre-emptive testimonial injustice. Even if patients were no longer excluded on the basis of purported lack of expertise, their most radical perspectives could still be dismissed through hermeneutical ignorance, a refusal to take seriously alternative conceptualizations that conflict with dominant psychiatric frameworks. Knox argues that social objectivity, understood through Longino’s (Reference Longino1990) account, requires integration of these alternative conceptualizations into the revision process if it is to subject its own background assumptions to transformative criticism (see also Pohlhaus Reference Pohlhaus2012, on wilful hermeneutical ignorance). The remedy is substantive inclusion, including perspectives hostile to mainstream psychiatry, in a process that gives the psychopathologized meaningful power over outcomes.

Tekin (Reference Tekin2022, Reference Tekin2025) arrives at a related conclusion from a different direction, and her proposal reconceives objectivity itself rather than demanding participation within an unchanged conception of it. Participatory Interactive Objectivity treats objectivity as a property of appropriately structured epistemic interaction, in which patients participate as experts alongside clinicians, and her Multitudinous Self Model renders first-person disruptions scientifically tractable as central manifestations of psychopathology rather than peripheral data.

Ballesteros (Reference Ballesteros2025) identifies a structural factor that the other accounts leave implicit: the conception of objectivity as standardization that operates in psychiatric practice promotes epistemically unjust treatment of patients by reducing their contributions to data points within predetermined assessment instruments. On her analysis, the problem is not primarily a matter of individual clinicians’ attitudes or skills; it is built into the theoretical commitment to standardized assessment as the privileged mode of objectivity. Ballesteros argues that rethinking objectivity beyond standardization is necessary for epistemic justice, and she explicitly notes that proposals like those of Knox and Tekin, while representing a step forward in the context of DSM revision, remain insufficient for achieving epistemic justice in clinical practice because they address the process that produces the manual without addressing the conception of objectivity that governs how the manual is used.

The four diagnoses are correct, and they are complementary rather than competing. What these accounts share is underdetermination at the level of institutional implementation.

2.2. What Knox already shows about implementation

Knox (Reference Knox2022a) goes further than the other accounts in articulating what inclusion must look like if it is to produce social objectivity, formulating three implications of their view: first, that the contributions of the psychopathologized cannot be limited to extrascientific matters; second, that hermeneutical ignorance of alternative conceptualizations presents an independent threat that must be addressed beyond simply extending participation to non-experts; and third, that integration of pathologized perspectives must be deep and substantive, reflecting real power to shape outcomes. Knox explicitly recognizes that the implementation of these implications remains an open problem (Reference Knox2022a: 253).

Aftab (Reference Aftab2022) and Bueter (Reference Bueter2022) respond in commentaries published alongside Knox’s article, to which Knox (Reference Knox2022b) replies in the same issue. Aftab asks what role scientific expertise should play in determining inappropriate exclusions and whether the goal is to diversify rather than dilute expertise. Bueter asks whether transformative criticism requires shared standards and what happens when the most radical critics do not share them. Aftab raises the question of whether uptake must be bidirectional and what that implies institutionally. He also draws attention to the possibility that the psychopathologized and the psychiatrically distressed or disabled may have different relationships to diagnostic authority, a distinction whose implications for the proposed conditions I address in Section 4. Each of these questions identifies a point at which Knox’s argument for inclusion generates a further problem about the form that inclusion should take.

Knox’s response to their commentators formulates what I take to be the central puzzle:

The answer requires institutional conditions that preserve critical contestation without collapsing deliberation, but they are derived from an analysis of the institutional structure of the revision process (Section 3), not from Knox’s framework alone.

2.3. From diagnosis to implementation

Recent work has begun to address the puzzle concretely.

Gagné-Julien and Friesen (Reference Gagné-Julien and Friesen2026) address the question of implementation in a way that is directly relevant but operates at a different level of analysis. Writing in the Hastings Center Report, they argue for the inclusion of people with lived experience in the DSM revision process and propose two concrete structures. The first is a committee composed entirely of people with lived experience (a ‘PLE committee’) to inform ongoing revisions. The second is the inclusion of at least two representatives with relevant lived experience in each existing review committee. Their analysis explicitly identifies the risks of tokenism and failures of uptake and recommends that the decision-making authority of representatives be clarified in advance, that roles and expected expertise be specified, and that mechanisms such as veto power or implementation pauses be considered. Gagné-Julien and Friesen derive specific participatory structures (ethical, epistemic, and epistemic justice-based) from a taxonomy of reasons for inclusion, whereas my argument specifies the epistemic conditions that any such structure must satisfy. The four conditions I develop in Section 4 are derived from the epistemic puzzle that Knox formulates as open, not from a taxonomy of reasons for inclusion. They address the same question that Gagné-Julien and Friesen address through structural prescription, but they answer it through the specification of conditions applicable to multiple possible structures.

Aftab’s (Reference Aftab2023) warning about elite capture provides the political urgency that motivates the distinction between genuine and tokenistic inclusion. In an editorial essay for Philosophy, Psychiatry, & Psychology, Aftab notes that ‘concepts in philosophy of psychiatry such as “epistemic injustice” show signs of elite captureFootnote ⁴ by the psychiatric establishment, resulting in token participation but no systemic changes’ (Reference Aftab2023: 6). If the philosophical vocabulary developed to diagnose epistemic injustice can itself be captured by the institutions it was meant to criticize, then the operationalization of conditions that distinguish genuine from tokenistic inclusion becomes a practical necessity. The conditions I propose in Section 4 are designed to prevent the specific capture Aftab identifies, namely the adoption of the language of inclusion without the institutional changes that would make inclusion epistemically productive.

3.1. The multi-stage architecture of DSM revision

The DSM revision is a multi-stage institutional process in which different actors exercise authority over different aspects of the outcome. Understanding where the conditions proposed in Section 4 would need to operate requires a schematic account of how the process currently works.

The DSM-5 revision process was organized around several institutional bodies with distinct mandates. Work groups, composed primarily of researchers and clinicians with expertise in specific diagnostic areas, were responsible for reviewing evidence, proposing changes to diagnostic criteria, and producing revised text. The Scientific Review Committee (SRC) evaluated whether proposed changes met evidential thresholds. The Clinical and Public Health Committee (CPHC) assessed proposals on grounds of clinical utility, public health impact, and coherence with the broader manual. The Board of Trustees of the American Psychiatric Association (APA) retained final approval authority over all changes. Each stage configured the relationship between deliberation and authority differently, and the perspectives of those affected by diagnostic decisions entered the process through different channels and with different degrees of influence.

Two features of the process that emerged after 2013 matter for what follows. The publication of DSM-5 was followed by the adoption of a continuous improvement model. A standing steering committee now oversees revisions on an ongoing basis, review committees perform the evaluative work previously assigned to work groups, proposed changes are posted for public comment, and approved modifications pass to the APA Assembly and Board of Trustees, a process that produced the DSM-5-TR in 2022 (Gagné-Julien and Friesen Reference Gagné-Julien and Friesen2026). Proposals enter through an open web portal and must satisfy criteria specified in advance, including a scholarly review of the literature supporting the change (Oquendo et al. Reference Oquendo, Abi-Dargham, Alpert, Benton, Clarke, Compton, Drexler, Fung, Kas, Malaspina, O’Keefe, Öngür, Wainberg, Yonkers, Yousif and Gogtay2026). The apparatus is permanent rather than convened for a single revision, a fact whose significance for the reproducibility of participatory arrangements I take up in Section 4.2.

The documents of the APA’s Future DSM Strategic Committee, which I analyse in detail elsewhere (Zeller Reference Zeller2026), reveal that the process now taking shape preserves this basic architecture while adding layers of consultation and stakeholder engagement. The committee’s initial strategy states that it is critical to include people with lived experience ‘as experts’, and the committee is developing frameworks for soliciting their input (Oquendo et al. Reference Oquendo, Abi-Dargham, Alpert, Benton, Clarke, Compton, Drexler, Fung, Kas, Malaspina, O’Keefe, Öngür, Wainberg, Yonkers, Yousif and Gogtay2026). The structure the strategy publishes tells a more specific story. People with lived experience appear among the non-voting ex officio members to be integrated into the strategic committee, alongside the committee’s fourteen voting members (Oquendo et al. Reference Oquendo, Abi-Dargham, Alpert, Benton, Clarke, Compton, Drexler, Fung, Kas, Malaspina, O’Keefe, Öngür, Wainberg, Yonkers, Yousif and Gogtay2026). The question is whether an institutional structure of this shape can produce the kind of epistemic engagement that the language of expertise implies.

3.2. The SRC/CPHC division

The division between the SRC and the CPHC during the DSM-5 revision is instructive because it institutionalized a distinction between ‘scientific’ and ‘extrascientific’ considerations that Knox (Reference Knox2022a) identifies as epistemically problematic. The SRC was charged with assessing the scientific evidence supporting proposed changes; the CPHC addressed everything else. Patient and advocacy perspectives were channelled primarily through the CPHC, through public comment periods, and through informal engagement with individual work group members. They were not represented in the SRC.

This institutional arrangement is the organizational expression of the hermeneutical ignorance that Knox describes. The division presupposes that the boundaries of scientific relevance are already settled, and it allocates authority over those boundaries to a body from which the psychopathologized are excluded. Solomon’s (Reference Solomon, Kendler and Parnas2017) point, developed through her examination of the decision to remove Asperger’s syndrome in favour of Autism Spectrum Disorder and endorsed by Knox, bears repeating in institutional terms. Solomon argues that framing the work of a body like the CPHC as extrascientific is misleading because we cannot tell a priori which considerations will prove scientifically relevant, and the Asperger’s case supplies the demonstration. A change presented as an internal scientific matter reshaped how autistic people are conceptualized and treated, and considerations the division had assigned to public acceptability turned out to bear directly on what the revision was deciding. A structure that settles the boundaries of scientific relevance by institutional fiat therefore undermines the epistemic goals it claims to serve. The SRC/CPHC division defined the institutional space in which scientific review occurs as a space from which psychopathologized perspectives are procedurally absent.

The first condition I propose, constitutive participation, is designed to address precisely this mechanism. The division between scientific and extrascientific review is the point at which the process determines whose epistemic contributions count as bearing on evidential questions and whose are relegated to considerations of language, stigma, and public acceptability.

3.3. Process and output: defining the scope of reform

A diagnostic manual must produce a specific kind of output, namely codified categories that function within reimbursement systems, regulatory frameworks, forensic proceedings, and clinical decision-making protocols. This constraint is real and non-negotiable in any practically relevant reform proposal. A revision of the DSM that produced categories unusable by insurance systems or incomprehensible to clinicians would fail at its institutional function regardless of the epistemic virtues of the process that generated it.

The conditions I propose in Section 4 operate on the epistemic process that produces categories, not on the operational form of the categories themselves. The distinction defines the space within which reform remains compatible with the institutional function of the DSM. The output constraint tells us what the manual must deliver. It does not tell us how the epistemic process that determines the content of that delivery must be organized. A DSM produced under these conditions would still generate ICD-compatible diagnostic codes and operationally defined criteria. The manual would remain usable within reimbursement, regulatory, and forensic systems.

The revision process would nevertheless incorporate a wider range of epistemic inputs, and the substantive content of the manual could shift over time as a result. The distinction between epistemic process and operational form remains analytically useful even granting this feedback. The operational format of the output (coded categories, defined criteria, administrative usability) stays stable, while what is categorized, where boundaries are drawn, and what counts as a symptom may gradually shift across revision cycles. This gradual modification is the mechanism through which an epistemic process responsive to diverse perspectives produces classification that better reflects the phenomena it aims to capture.

I do not want to suggest that this process is frictionless. Constitutive participation in agenda-setting will make the early stages of revision slower and more contested. Procedurally differentiated translation will generate disagreements about which stage-specific norms apply in particular cases. Bidirectional uptake with documented accountability will impose administrative costs that the current process does not bear. These are real costs, and they are predictable consequences of expanding the epistemic base of the revision process. They are also the ordinary costs of social objectivity in any domain where it has been implemented. Longino’s model does not promise that socially objective processes will be faster or smoother than alternatives; it holds that they will be more epistemically reliable. Douglas (Reference Douglas2009) makes the related point that the operation of values in scientific deliberation is unavoidable rather than eliminable, and that the relevant question is whether their operation is subject to appropriate procedural constraints. The conditions developed in Section 4 specify what such constraints would require in the context of psychiatric classification. The trade-off between procedural efficiency and epistemic quality is not unique to psychiatry, and accepting it is part of what it means to take social objectivity seriously as a desideratum rather than as a rhetorical gesture.

3.4. The limits of downstream consultation: the ASAN case

Participation of the psychopathologized in the DSM revision process occurs predominantly at consultative and downstream stages. Public comment periods allow advocacy organizations and individuals to respond to draft proposals, but these responses enter the process after the fundamental epistemic decisions (what questions to ask, what evidence to seek, what constructs to investigate) have already been made. The psychopathologized participate as respondents to an agenda set by others.

The case of the Autistic Self Advocacy Network (ASAN), a non-profit organization run by and for autistic people, illustrates this dynamic with particular clarity. ASAN’s engagement with the DSM-5 Neurodevelopmental Disorders Workgroup was a significant if partial success: the organization influenced the final form of autism-related diagnoses through sustained, sophisticated advocacy (Kapp and Ne’eman Reference Kapp, Ne’eman and Kapp2020). The advocacy was grounded in direct engagement with the research literature. ASAN’s policy briefs to the workgroup reviewed the evidence bearing on the draft criteria and set out research-backed options for revising them (Kapp and Ne’eman Reference Kapp and Ne’eman2012), a fact that shows that the capacity for evidence-level contribution that constitutive participation presupposes has already been demonstrated. Yet as Knox (Reference Knox2022a) documents, ASAN’s influence operated through persuasion of work group members who retained unilateral decision-making authority. Kapp and Ne’eman (Reference Kapp, Ne’eman and Kapp2020) describe the goal as having psychiatry acknowledge the autistic community ‘as an equal, not as a junior partner’ (189), a formulation that reveals the gap between what was achieved and what genuine inclusion would require. An equal partner has standing to shape outcomes through deliberation; ASAN had standing to lobby decision-makers who could accept or reject their input without institutional accountability for that decision.

The ASAN case simultaneously illustrates the potential of psychopathologized participation and its structural precariousness under current arrangements. What ASAN achieved depended on organizational capacity, political timing, and the receptivity of specific individuals. None of these conditions are reproducible by institutional design in the current governance structure. The conditions that I will propose in the next section are designed to make inclusion structurally reproducible rather than contingently achievable and to extend the epistemic scope of participation from response to agenda-setting.

4.1. Constitutive participation

Knox (Reference Knox2022a) argues that the contributions of the psychopathologized cannot be limited to extrascientific matters if psychiatry is to achieve social objectivity. The argument follows directly from Longino’s account: if normative outlook and background assumptions shape how a scientific community assesses what counts as good evidence, then restricting pathologized input to matters of language and stigma while reserving evidential questions for credentialed experts reproduces the very division that social objectivity is meant to overcome. As Section 3 documented, the DSM-5 revision process institutionalized this division through the separation of the SRC from the CPHC, creating an organizational structure that presupposes settled boundaries of scientific relevance and excludes the psychopathologized from contesting them.

There is a further reason why this exclusion is particularly damaging in psychiatry. The phenomena that psychiatric classification aims to capture are not exhausted by subpersonal mechanisms; their characterization depends in part on how they are experienced, interpreted, and negotiated by the subjects to whom the categories apply, and this makes first-person perspectives epistemically relevant to how the classificatory domain is delineated.Footnote ⁵ Excluding psychopathologized perspectives from evidential deliberation risks narrowing the epistemic access through which the classificatory domain itself becomes intelligible, a risk that goes beyond the loss of useful experiential information.

Tekin (Reference Tekin2025) reaches a convergent conclusion by a different route. Her Multitudinous Self Model treats disruptions of agency, narrative coherence, embodiment, relationality, and self-concept as central manifestations of mental disorder rather than ancillary features, and on that basis she argues that the testimony of those who live through such disruptions constitutes expertise by experience that psychiatric science needs rather than merely accommodates. The claim defended here is complementary and operates one level up. Where Tekin establishes that first-person reports are scientifically tractable evidence about the self, my concern is the delineation of the classificatory domain. If the phenomena are partly constituted at the personal level, then decisions about what belongs in the domain at all, and not only decisions about how to explain what has been admitted into it, require epistemic access that runs through the perspectives of the classified.

The condition that follows is constitutive participation. The psychopathologized must be involved in epistemic activities that shape what questions are asked, what evidence is sought, and what counts as a successful answer, not only in activities that respond to questions already formulated by credentialed experts. In institutional terms, this means formal representation in bodies that determine research agendas and evidential standards and dedicated funding for research programmes led by individuals with lived experience at early stages of the revision cycle, before the parameters of inquiry have been set. Aftab (Reference Aftab2022) rightly emphasizes that achieving this requires investment in a ‘pipeline’ of researchers with lived experience (citing Jones et al. Reference Jones, Atterbury, Byrne, Carras, Brown and Phalen2021). Standpoint theory explains why the investment is epistemically necessary rather than merely equitable. On the account that Friesen and Goldstein (Reference Friesen and Goldstein2022) defend for the psy sciences, the epistemic advantage associated with marginalization is an achievement of critical engagement rather than an automatic endowment of social position, and those who have attained such a standpoint therefore need tools, funding, and sustained opportunities to develop research programmes of their own. The condition cannot be met by including psychopathologized participants in existing deliberative bodies while leaving the pipeline that produces qualified participants undeveloped. The constitutive dimension requires both structural presence in decision-making and prior investment in the conditions that make that presence epistemically productive.

The stakes of the early stage are not hypothetical. Spectrum 10K, a large genomic study of autism in the United Kingdom, was paused shortly after its launch and ultimately abandoned following sustained objections from autistic people and their organizations, whose concerns included the possible eugenic uses of the data and the absence of meaningful community involvement in the design of the research (Knox Reference Knox, Tsou, Shaw and Fehr2025; Natri et al. Reference Natri, Chapman, Heraty, Dwyer, Walker, Kapp, Dron, Martinez-Agosto, Mikkola and Doherty2023). The objections targeted decisions about whether and how to pursue the study, decisions settled before there were any findings to evaluate. Gopakumar (Reference Gopakumar2026) argues that community participation in this act of gatekeeping was justified and that gatekeeping in potentially harmful research legitimately extends to decisions about pursuit rather than only to decisions about publication. The episode translates directly to classification. Agenda-setting is the stage at which the framings, priorities, and risks that will structure subsequent inquiry are fixed, and it is the stage from which the psychopathologized have been most systematically absent.

The process for the next DSM supplies the immediate test. The initial strategy affirms that people with lived experience should be included as experts, and the committee structure it publishes places them among the non-voting ex officio members (Oquendo et al. Reference Oquendo, Abi-Dargham, Alpert, Benton, Clarke, Compton, Drexler, Fung, Kas, Malaspina, O’Keefe, Öngür, Wainberg, Yonkers, Yousif and Gogtay2026). Under the continuous revision model, one patient representative sits on the steering committee, and none sit on the review committees that evaluate proposals (Gagné-Julien and Friesen Reference Gagné-Julien and Friesen2026). Constitutive participation sets a determinate standard against this arrangement. It requires representation with voting standing in the bodies that will fix the evidential standards and research priorities of the revision and funded research programmes led by people with lived experience commissioned while those parameters remain open. Input solicited from non-voting seats after the voting membership has framed the questions is consultation, whatever vocabulary of expertise accompanies it.

4.2. Structural reproduction

As I documented in Section 3, ASAN’s engagement with the DSM-5 Neurodevelopmental Disorders Workgroup achieved significant influence but created no institutional precedent: a future workgroup could exclude analogous organizations without violating any procedural norm. The question this raises is what structural property the deliberative process would need to possess for inclusion to reproduce itself across revision cycles.

Aftab’s (Reference Aftab2022) distinction between diversifying and diluting expertise provides the analytic frame for this condition, and Longino’s (Reference Longino2002) criterion of tempered equality of intellectual authority states the underlying norm precisely. Equality of intellectual authority does not mean that every participant carries equal weight on every question; it is tempered by relevant expertise and by the quality of arguments. What the criterion rules out is disqualification on the basis of social position rather than epistemic merit, and this is the mode of exclusion that Knox (Reference Knox2022a) documents for the psychopathologized. The goal is therefore not to place non-experts on equal footing with experts in matters that require specialized competence. The goal is to ensure that the community of experts is diverse enough to subject its own background assumptions to the kind of transformative criticism that Longino’s model demands. Diversification in this sense is a structural property of the deliberative body, achieved through codified governance rather than individual discretion.

What follows from this analysis is a requirement of structural reproduction. The mechanisms that sustain inclusion must be procedurally embedded so that they survive changes in institutional leadership. In institutional terms, this means formal representation with deliberative authority at specified stages of the revision process, codified in standing governance instruments rather than dependent on the discretion of committee chairs. The criteria for adequate representation must be specified independently of the individuals who happen to occupy leadership positions at any given revision cycle. When representation depends on the orientations of individual leaders, inclusion becomes a gift that can be withdrawn. When it is codified in procedure, withdrawal requires an explicit institutional decision with corresponding justificatory costs. The difference is between an arrangement that reproduces itself by default and one that must be actively sustained against institutional inertia.

An objection makes visible how demanding the condition is. The DSM has no codified revision procedure that carries over from one major revision to the next. Each revision has constituted its own governance, and where a process has resembled its predecessor, the resemblance reflects leadership choosing to honour loose precedent rather than complying with standing institutional policy. Structural reproduction therefore cannot be achieved by inserting a participation clause into governance documents that do not persist between revisions. The objection is descriptively correct, and it identifies the reason the condition takes the form it does rather than a reason to abandon it. The precariousness it describes is not confined to participatory arrangements. It is the general condition of the revision process’s own procedures, and it is exactly what makes inclusion a gift that each cycle may decline to renew. Meeting the condition accordingly requires creating standing governance where there has been none, through instruments that persist across major revisions, such as provisions written into the mandate of the standing steering committee or into the operating rules of the continuous revision apparatus described in Section 3.1. That apparatus shows that permanence of this kind is institutionally available. The steering committee, the review committees, and the proposal portal have operated continuously for over a decade, spanning the production of the DSM-5-TR, without being reconvened from scratch. What has never been codified within them is participation. The condition names that omission, and the design of the process for the next DSM, which is being settled now, is the occasion on which it could be repaired.

4.3. Translation interfaces that preserve content

Knox formulates the puzzle that motivates this condition in the passage already cited from their response to Bueter and Aftab (Knox Reference Knox2022b: 277): how psychiatric science can engage productively with alternative interpretations of its domain without reducing them to the dominant framework or stripping them of their critical force.

The puzzle concerns the relationship between the critical perspectives that inclusion brings into the process and the epistemic framework within which the process operates. If participants must translate their contributions into the dominant framework before those contributions can receive uptake, the translation strips away the very features that made the contributions epistemically valuable. A neurodiversity advocate who must express their perspective in the language of disorder and dysfunction in order to be heard by a workgroup has already conceded the conceptual terrain that their participation was meant to contest. This is not a hypothetical risk. Diagnostic categories encode evaluative commitments that cannot be neutralized simply by stipulating that one uses them descriptively rather than evaluatively.

The exchange between Bueter and Knox makes visible the difficulty without resolving it. Bueter (Reference Bueter2022) argues that social objectivity requires shared standards for criticism, invoking Carrier’s (Reference Carrier2013) notion of ‘epistemic attitude’ as the minimal shared commitment that enables productive disagreement. Knox (Reference Knox2022b) accepts the criterion but insists that the standards themselves must be open to contestation, since one of the functions of including radically different perspectives is to interrogate the background assumptions embedded in current epistemic standards. The disagreement is about what must be shared for disagreement to be productive. Is it shared standards of assessment or shared commitment to the goals of inquiry without agreement on the standards?

I propose that this disagreement can be dissolved procedurally rather than resolved theoretically. Any revision of a diagnostic manual performs three distinguishable epistemic functions, whether or not its institutional organization separates them. Agenda-setting determines which phenomena deserve investigation and which questions are worth asking. Evidence evaluation assesses the empirical support for proposed diagnostic constructs. Criteria formulation translates the results of evaluation into operationally usable diagnostic criteria. These functions have different epistemic requirements, and a single set of participatory norms applied uniformly across all of them will either be too restrictive for the first function or too permissive for the last. The distinction is analytic. Whether the functions are also carried out as institutionally separate stages is a fact about process design, and I return below to the fact that in the current process they are not.

The condition is procedurally differentiated translation. Participation formats must vary by stage, with explicit criteria for how non-standard perspectives can be articulated and considered at each stage. What distinguishes this condition from ordinary deliberative procedure is its directionality. Standard deliberative protocols assume a shared framework and ask how diverse inputs can be processed within it. Procedurally differentiated translation asks a prior question. At which stages of the process must the framework itself be open to contestation, and at which stages may the framework legitimately constrain the form of contributions?

A neurodiversity perspective may reject the assumption that a given condition constitutes a disorder or dysfunction at all. Such a perspective contests the criteria through which the framework distinguishes disorder from non-pathological variation, and its contribution cannot be fully articulated within a framework that already presupposes deficit and dysfunction. Requiring it to adopt the operational language of deficit and dysfunction at every stage of deliberation neutralizes the form of criticism that made its inclusion epistemically valuable in the first place.

Framework-level rejection is one position within neurodiversity discourse rather than the whole of it. Advocacy movements more commonly occupy a hybrid position, mobilizing diagnostic categories to secure recognition, services, and accommodation while resisting the pathologizing normativity those categories encode, and ASAN’s campaign to revise rather than abolish the autism criteria is an instance (Kapp and Ne’eman Reference Kapp, Ne’eman and Kapp2020). The prevalence of that hybrid position is itself evidence of the pressure this condition addresses. Where institutional legibility runs through diagnostic vocabulary, perspectives that contest the vocabulary must adopt it in order to be heard at all, and they carry the resulting conceptual concessions into every setting where recognition is at stake.

The differentiation by stage responds to this problem. At the agenda-setting stage, perspectives should be expressible in their own frameworks. The relevant question at this stage is whether a perspective identifies phenomena, experiences, or concerns that the current framework overlooks. Existing evidential standards apply at later stages, where their function is to assess proposals, not to filter which proposals can enter the process.

At the evidence-evaluation stage, criteria for relevant evidence should remain explicit and contestable. Participants must be able to challenge the assumptions governing evidential assessment. Deliberation at this stage still requires standards that allow proposals to be comparatively evaluated.

Those standards emerge partly from earlier stages of the revision process. Agenda-setting activities shape which phenomena, experiences, and forms of evidence enter subsequent deliberation. Constitutive participation therefore influences later adjudication by expanding the range of considerations treated as epistemically relevant.

Procedural differentiation localizes disagreement at the stage where the epistemic resources for managing it are generated. Early stages permit broader contestation of classificatory assumptions. Later stages operate under stronger constraints connected to reliability, codability, and administrative usability.

The tension that Knox identifies survives procedural differentiation in transformed form. The constraints that apply at later stages follow from the institutional function of the manual rather than from the epistemic superiority of any particular framework. Radical reconceptualizations therefore have their greatest role at the agenda-setting stage, where they can reshape what phenomena become visible to inquiry.

The condition now faces an institutional objection that must be met directly. The three functions are not carried out as separate stages in the current process, and the first is not carried out as a distinct activity at all. Revision work begins from the existing criteria. Review committees, like the work groups before them, take the current categories as given and ask whether there are reasons to change them so that agenda-setting, evidence evaluation, and criteria formulation operate simultaneously and implicitly within a single deliberative frame. The proposal portal codifies this default. A submission must arrive with structured documentation of validity, reliability, and clinical utility for a change to the existing categories, which means it must be formulated in the manual’s own evidential vocabulary before it can be considered at all (Gagné-Julien and Friesen Reference Gagné-Julien and Friesen2026; Oquendo et al. Reference Oquendo, Abi-Dargham, Alpert, Benton, Clarke, Compton, Drexler, Fung, Kas, Malaspina, O’Keefe, Öngür, Wainberg, Yonkers, Yousif and Gogtay2026). Consider what this structure does to a radical challenge. A neurodiversity advocate who contests whether autism should be classified as a disorder at all can raise that challenge only at the point where the relevant committee is reviewing the existing autism criteria, a point at which evidence evaluation and criteria formulation are already on the table and the only question the process can register is whether the criteria should be modified. The framework-level challenge is not refuted; it is rendered procedurally unintelligible, and this is the institutional form of the reduction Knox warns against. The criteria-first default is not a neutral organizational convenience. It is the mechanism by which hermeneutical ignorance is built into the sequence of deliberation itself. Procedurally differentiated translation therefore demands more than adjusting participation within the process as it stands. It requires creating an agenda-setting stage that does not currently exist, institutionally separated from the evaluation of specific proposals, in which framework-level challenges can be articulated and considered before the existing categories set the terms of discussion. This is a restructuring of the revision process, and I present it as such. The design of the process for the next DSM is an occasion on which such a stage could be created deliberately, and the existence of a strategic committee whose mandate is to reconsider how the future manual should be produced shows that process design at this level is already on the institutional agenda (Oquendo et al. Reference Oquendo, Abi-Dargham, Alpert, Benton, Clarke, Compton, Drexler, Fung, Kas, Malaspina, O’Keefe, Öngür, Wainberg, Yonkers, Yousif and Gogtay2026).

Knox (Reference Knox, Tsou, Shaw and Fehr2025) has recently proposed a related but distinct response to the same puzzle, which they call ‘hermeneutical pluralism’: the view that activist knowledge systems such as the Neurodiversity Movement should be maintained as partly independent of psychiatric science, serving as external critical checks on a system that might otherwise adopt inclusive language without changing its practices. The conditions proposed here are compatible with hermeneutical pluralism. They address the internal epistemic structure of the revision process; hermeneutical pluralism addresses the external relationship between the revision process and the knowledge systems it engages with. Both are needed: the conditions specify how inclusion within the process can be genuine, while hermeneutical pluralism explains why inclusion within the process cannot be the whole story.

4.4. Bidirectional uptake with epistemic attitude

Longino’s requirement of uptake of criticism has a feature that is easy to overlook: it cuts both ways. As Aftab (Reference Aftab2022) emphasizes, quoting Longino (Reference Longino2002: 130), ‘not only must the community be responsive, but the claims of advocates of a line of criticism must take account of those responses’. The psychopathologized, as they enter critical dialogue with the psychiatric community, must be prepared to receive and respond to criticism of their own conceptualizations. This bidirectionality distinguishes epistemic participation from mere consultation. In consultation, one party provides input and the other decides what to do with it. In social objectivity, both parties subject their positions to critical scrutiny and demonstrate responsiveness to the criticism they receive.

Bueter (Reference Bueter2022) anchors the criterion for productive participation in Carrier’s (Reference Carrier2013) notion of epistemic attitude, which she glosses as a ‘shared striving for the goals of knowledge and consensus’ (273). Knox (Reference Knox2022b) accepts this criterion in their response. The convergence is significant. It identifies a threshold below which inclusion cannot produce the epistemic benefits it promises. Participants who refuse on principle to engage with the epistemic goals of the process they have joined cannot contribute to its social objectivity. This threshold provides an internal criterion for distinguishing participants whose radical criticism enriches the process from participants whose engagement is incompatible with the process as an epistemic undertaking. Bueter (Reference Bueter2022) formulates the exclusion principle precisely: it remains possible to ‘exclude participants whose behavior openly goes against the norms of social objectivity, or whose beliefs are incompatible with participating in the critical process in good faith’ (273). The criterion applies to participants’ engagement with the epistemic-pragmatic goals of the revision process rather than to the substantive content of their views. A participant who holds that current diagnostic categories systematically misrepresent the phenomena they classify, or that the psychiatric framework is itself conceptually flawed, satisfies the criterion as long as they remain willing to expose that disagreement to critical engagement within the process. The criterion excludes refusal to participate in deliberation as such; it does not exclude radical disagreement with prevailing assumptions.

The convergence between Bueter and Knox can be translated into an institutional requirement of bidirectional uptake with an epistemic attitude, as the participatory approach proposed by Gagné-Julien and Friesen (Reference Gagné-Julien and Friesen2026) similarly recognizes. Uptake must be documented. When the revision process considers and declines to incorporate a perspective contributed by psychopathologized participants, the reasons for that decision must be publicly articulated and made available for further scrutiny. Documentation of uptake is the mechanism through which the process demonstrates that non-incorporation reflects reasoned engagement rather than hermeneutical ignorance. The analogy is to peer review in other scientific contexts, where the norm is that rejection is accompanied by reasons that the rejected party can evaluate and respond to. A DSM revision process that incorporates psychopathologized perspectives without systematic documentation of how those perspectives were received, evaluated, and either integrated or set aside with reasons cannot claim to have achieved uptake in Longino’s sense. The demand has precedent within the field’s own institutions. Reviewers and editors with lived experience have begun to codify standards requiring that publications claiming participatory methods document who was involved, in what capacity, and with what degree of influence, precisely so that the participatory vocabulary cannot cover tokenistic practice (Colder Carras et al. Reference Colder Carras, Machin, Brown, Marttinen, Maxwell, Frampton, Jackman and Jones2023). The requirement proposed here extends an emerging documentation norm from the reporting of participatory research to the governance of classification.

The epistemic attitude criterion also functions as a safeguard against the concern that inclusion will paralyse the revision process through interminable disagreement. Not every perspective has an unconditional entitlement to indefinite participation. What every perspective contributed in good faith has is an entitlement to documented engagement, meaning the right to know that it was considered, on what grounds it was assessed, and why, if it was not adopted, the process reached a different conclusion. This is a demanding requirement, but it is not an unfamiliar one. Scientific institutions routinely produce documented justifications for their decisions when those decisions affect the interests of identifiable parties. The DSM revision process has simply not been held to this standard with respect to the psychopathologized.

The process for the next DSM supplies the test once more. Under the continuous revision model, proposed changes are opened to public comment, but there is no transparency about how comments are handled, no published account of the weight they receive, and no practice of returning reasons to those whose input is not taken up (Gagné-Julien and Friesen Reference Gagné-Julien and Friesen2026). A process that intends the engagement language of its strategy documents to mean more than consultation would commit in advance to a public record of the perspectives received, the assessment they were given, and the grounds on which they were incorporated or set aside. The published strategy contains no commitment of this kind.

5.1. The process/output distinction, qualified

The conditions modify the epistemic organization of the revision process rather than the operational requirements of the manual itself. Changes in the classification process will, over time, produce changes in the substantive content of the output. If constitutive participation in agenda-setting modifies what counts as a phenomenon worth investigating, and if translation interfaces allow non-standard perspectives to inform the assessment of evidence, then different diagnostic constructs may emerge, existing constructs may be revised, and some category boundaries may shift. This is not a defect to be minimized. It is the point.

The relevant question is the pace and mechanism of change. The conditions produce change gradually, through the normal operation of a revised deliberative process, across successive revision cycles. They do not produce change by fiat or by external imposition. The distinction between gradual change through a legitimate epistemic process and abrupt change through institutional disruption is precisely what enables professional evolution without disciplinary dissolution. Abbott (Reference Abbott1988) documents how professions routinely reconfigure their jurisdictional functions in response to changing epistemic and social conditions. What they cannot survive is abrupt loss of their core jurisdictional claim. The conditions proposed here do not threaten psychiatry’s jurisdictional claim to diagnostic authority; they modify the deliberative process through which that authority is exercised.

Knox (Reference Knox2023) provides additional support for this argument from within the philosophy of psychiatry. Their account of the institutional definition of psychiatric conditions holds that the legitimacy of psychiatric classification already depends on the quality of the social process that produces it. If psychiatric conditions are partly constituted by the institutional processes through which they are identified, classified, and responded to, then improving the epistemic quality of those processes realizes the normative commitments the institution already claims to hold. A discipline that defines its conditions through a socially objective process has a stronger, not weaker, claim to epistemic authority than one that defines them through a process from which relevant perspectives have been excluded.

Anderson (Reference Anderson2006) develops a parallel argument for democratic institutions more generally. The legitimacy of democratic decision-making depends not only on procedural inclusion but also on the epistemic quality of the deliberation that inclusion makes possible. Diverse participation contributes to better collective epistemic outcomes when it is structured by procedures that allow participants’ contributions to inform the substantive deliberation rather than merely to be registered. The conditions I propose are an application of this general principle to the institutional case of psychiatric classification.

5.2. The diluted expertise objection

The diluted expertise objection holds that the proposed conditions, by expanding participation beyond credentialed experts, threaten the jurisdictional function that constitutes psychiatry as a profession. On this view, psychiatric authority to diagnose rests on specialized expertise, and diluting the deliberative process with non-expert perspectives undermines the epistemic basis of that authority.

My response operates on two levels. At the first level, the conditions do not dilute expertise. They diversify the community of experts, following Aftab’s (Reference Aftab2022) formulation. Constitutive participation (4.1) combined with the pipeline investment Aftab identifies means that the psychopathologized enter the process as participants with relevant epistemic contributions, not as untrained interlopers. Structural reproduction (4.2) ensures that diversity is a property of the deliberative body rather than a contingent feature of particular revision cycles. The conditions expand the epistemic base of diagnostic authority; they do not erode it.

The position is supported by work in the sociology and epistemology of expertise. Collins and Evans (Reference Collins and Evans2002) distinguish contributory expertise from interactional expertise and from referee expertise, arguing that effective participation in technical deliberation can be grounded in different kinds of competence relevant to different functions in the process. Their framework supports the present argument in two ways. First, it makes clear that expertise is not exhausted by formal credentials. Second, it provides analytic tools for matching participatory roles to relevant competences at different stages of revision, which connects directly to Condition 4.3 on procedurally differentiated translation.

At the second level, the conditions include an internal limiting principle. As Condition 4.4 establishes, the criterion of epistemic attitude (Bueter Reference Bueter2022; Carrier Reference Carrier2013) provides a threshold below which participation cannot produce epistemic benefits: participants whose engagement is incompatible with the shared goals of the process can be excluded without violating the norms of the deliberative process. The diluted expertise objection assumes that inclusion means abandoning quality criteria. The conditions specify inclusion under quality criteria that are themselves subject to collective scrutiny.

The objection has a stronger form, an incommensurability variant, that is worth addressing directly. It holds that even if the conditions do not dilute expertise, they introduce perspectives whose epistemic frameworks are incommensurable with psychiatric science, making productive deliberation impossible. The response is that this is precisely the problem that Condition 4.3 (translation interfaces) is designed to address. Procedurally differentiated participation formats distribute the challenge of incommensurability across stages in which it can be locally managed. The incommensurability variant gains its force from the assumption that all participation must meet a single evidential standard. Once that assumption is abandoned in favour of stage-differentiated norms, the objection loses its force.

5.3. The radical objection

The radical objection comes from the opposite direction. Gagné-Julien (Reference Gagné-Julien2022) argues that inclusion of marginalized perspectives within existing institutional structures risks co-optation: the perspectives are absorbed, domesticated, and used to legitimate a process that remains fundamentally unchanged. On this view, what Mad Studies and related movements need is autonomous spaces for the production of knowledge about mental distress, madness, and neurodivergence, rather than better inclusion in psychiatric governance. These spaces would operate outside the epistemic authority of psychiatry.

The objection identifies a real risk, and nothing in my argument denies it. My argument concerns the epistemic governance of inclusion within existing classificatory institutions, not the full political reorganization of psychiatric authority. The conditions do not address injustices that occur outside the revision process, they do not create autonomous spaces for alternative knowledge production, and they do not claim to resolve the deeper political questions about the relationship between psychiatric authority and the communities it classifies.

What the conditions do is specify the minimum institutional requirements without which any form of inclusion, however well-intentioned, will reproduce tokenism. Gagné-Julien and Friesen (Reference Gagné-Julien and Friesen2026) themselves identify the risks of tokenism and failures of uptake as central dangers of participatory approaches. The conditions that I propose here are designed to prevent exactly those outcomes. If autonomous spaces for alternative knowledge production are also necessary, the conditions do not preclude them. Knox (Reference Knox, Tsou, Shaw and Fehr2025) has independently arrived at a compatible conclusion: their hermeneutical pluralism holds that movements like the Neurodiversity Movement serve as indispensable external checks on psychiatric science, and that this function requires that they maintain partial independence from the institutions they critique. The Spectrum 10K case, introduced in Section 4.1, illustrates the point concretely in Knox’s analysis: even when researchers explicitly endorsed a Neurodiversity framework, they failed to anticipate risks that independent activists identified immediately. The conditions proposed here operate within the revision process; hermeneutical pluralism operates alongside it.

A stronger version of the objection targets the distinction between epistemic process and operational output itself. On this view, procedural reform leaves intact a classificatory structure organized around concepts of deficit and dysfunction (LeFrançois et al. Reference LeFrançois, Menzies and Reaume2013).

Two replies are available. The first is that the operational constraint does not entail the conceptual commitment. What the output must deliver is codified categories usable within reimbursement, regulatory, and forensic systems, and nothing in that requirement dictates that the categories be organized around disorder or dysfunction. Knox (Reference Knox2023) shows that a classification could instead be organized around the concept of a psychiatric condition, institutionally defined in terms of whether psychiatric involvement would serve well-being, which carries no commitment to pathology, dysfunction, or disorder. A revision process satisfying the four conditions could therefore move the conceptual basis of the manual in that direction across successive cycles while its operational form remained intact. The distinction between process and output does not freeze the framework in place; it is what makes framework-level change achievable through a legitimate epistemic process rather than by fiat.

The second reply concerns the institutional horizon. Even for those who reject disorder-based classification outright, my argument presupposes only that the DSM, or an institution functionally equivalent to it, will continue to operate within legal, regulatory, and reimbursement systems.

Under those conditions, the organization of the revision process remains philosophically consequential even for those who advocate more fundamental forms of transformation.

Questions about institutional replacement therefore do not eliminate questions about epistemic governance. The radical objection correctly identifies the limits of procedural inclusion within existing classificatory systems. The governance of those systems nevertheless remains a distinct and unavoidable problem as long as they continue to structure psychiatric authority.

Bueter’s (Reference Bueter2022) criterion of epistemic attitude is relevant here as well. The shared goal she proposes for the DSM case (creating a diagnostic manual based on current scientific knowledge that facilitates the best care possible) can be accepted without requiring participants to renounce alternative frameworks. A participant can hold that their own framework offers a better account of the phenomena while nonetheless accepting the epistemic-pragmatic goal of the process as a shared point of departure. Accepting the goal does not mean endorsing every assumption embedded in the current implementation of that goal. It means agreeing that the disagreement about those assumptions can be conducted within a process oriented toward that goal. This makes the disagreement about epistemic frameworks productive rather than paralysing, because it provides a criterion for what productive engagement looks like without predetermining the outcome.