Damian Milton, an autistic sociologist working from Kent without an academic post, submitted a five-page paper to Disability & Society in 2012 that proposed something the discipline had not yet permitted itself to consider: that the well-attested social difficulties of autistic adults might be measurement artefacts of a field whose instruments had been calibrated by non-autistic researchers asking non-autistic questions of subjects who could not be heard inside the questions they were being asked. The paper called the artefact the double empathy problem. It argued, with the kind of structural precision a sociologist trained in interactionism can deploy when the personal stakes are high enough, that empathy across neurotypes is a bilateral failure rather than a unilateral deficit, and that the entire literature of autistic theory-of-mind impairment had been telling only one half of a two-handed story. Milton's paper has now been cited more than seventeen hundred times,1 a number that no five-page article in a small disability-studies journal had any reasonable expectation of achieving when an autistic outsider submitted it from a position of professional precarity.
The citation count is not the point. The point is that the paper, against the polite expectations of its field, became the conceptual hinge on which a generational paradigm shift turned. Eight years after Milton's article, Catherine Crompton's Edinburgh group ran the diffusion-chain experiments that gave the theory empirical teeth: information transferred between autistic-autistic dyads at rates indistinguishable from neurotypical-neurotypical dyads, with the cross-neurotype condition producing the only measurable degradation.2 By 2021 Liz Pellicano and Jacquiline den Houting had argued in Journal of Child Psychology and Psychiatry that the field needed to commit, formally and methodologically, to a paradigm shift away from the medical model and toward what they called the neurodiversity paradigm.3 By 2022 Robert Chapman's Empire of Normality had located the entire deficit-discourse inside what Chapman, drawing on Karatani and Marxist disability theory, called neuronormative empire: a regime of cognitive evaluation produced by industrial capitalism's need for standardised labour and now metastasised into every domain of clinical and educational life.4 The field, in other words, has spent the decade after Milton arguing about whether the deficit model is a research error, a paradigm error, or a political error, and the three positions are not the same answer.
This review reads the field through one operational question: what does the literature license for the design of an adult affinity group that gathers neurodivergent people, takes the liberation pedagogy of Tricia Hersey and Sonya Renee Taylor and Devon Price as primary readings rather than ornamental epigraphs, and treats allies as first-class participants? The question is operational because the curriculum exists, the design problem is real, and the literature is being asked to do work it has not always been asked to do.
The work splits into three demands. The first is methodological: what counts as evidence in a field where the most frequently cited primary source on autistic burnout is a community-participatory qualitative study5 rather than a randomized trial, and where the principal trade-press anchors of liberation pedagogy do not appear in PubMed at all? The second is political: what counts as suffering, and what counts as cure, in a literature that argues with itself about whether autistic distress is a developmental disorder requiring intervention6 or a minority-stress phenomenon produced by a hostile ecology7, or a structural feature of the political economy that the curriculum is meant to interrupt rather than reform.4 The third is design-clinical: what does a closed-cohort ten-session pan-neurodivergent group with first-class-allies architecture have to inherit, and from where, when no published protocol has previously attempted that integration?
The literature divides into six positions rather than six topics, and the distinction matters because reviews that treat them as topics produce the genre Sonya Renee Taylor warned about: bibliographies that gesture at disagreement without metabolizing it. The pathology paradigm, represented in the contemporary clinical literature by Solanto's metacognitive therapy for adult ADHD,8 Safren's CBT-for-adult-ADHD protocols,9 and the Cambridge Autism Research Centre's social-cognition program,10 treats autism and ADHD as developmental disorders whose features are dysfunctions to be remediated. The accommodation-affirmation paradigm, articulated through Pellicano's editorial program at Autism, the AASPIRE practice-based guidelines,11 and the Dallman framework for affirming autism therapy,12 treats autistic features as natural variation requiring environmental fit rather than individual remediation. The critical-disability-studies and abolitionist position, carried by Robert Chapman, Nick Walker, Melanie Yergeau, and Monique Botha,13 treats the regulatory ideal of normality itself as the producing condition of autistic suffering and refuses the reformist horizon of accommodation. The autistic-phenomenological and standpoint-epistemic position, originated in the Murray-Lesser-Lawson monotropism account14 and developed by Milton, Raymaker, Crompton, and Kapp, treats autistic first-person report as primary epistemic material rather than as data to be interpreted by a non-autistic clinician. The liberation-pedagogy position, anchored by Hersey, Taylor, Price, and the bell hooks lineage, sits outside the clinical frame entirely and treats neurodivergent suffering as politically produced exhaustion that demands organised rest rather than therapeutic adjustment. The mad-studies and critical-psychiatry tradition, represented by Beresford, McWade, Mills, Moncrieff, and Bracken,15 is the broader anti-pathology lineage of which the neurodiversity paradigm is a younger sibling.
These positions disagree with each other in ways the curriculum cannot wish away. The pathology paradigm would call the curriculum's first-class-allies architecture a category error, since allies have no diagnosis and therefore no clinical indication. The accommodation paradigm would accept the architecture but ask whether the curriculum is sufficiently outcome-instrumented to demonstrate fit. The abolitionist position would interrogate whether a ten-session closed-cohort container reproduces the disciplinary logic it claims to interrupt. The phenomenological position would ask whose first-person report grounds the design and at what cost to the AAC-user theorists whose accounts seldom enter Anglophone clinical literature. The liberation-pedagogy position would ask whether the integration of Hersey's specifically Black-liberation-theological grounding into a secular pan-ND container constitutes an honoring or a flattening. The mad-studies position would ask why the McWade-Milton-Beresford 2015 dialogue16 has not been mined more carefully by either field. The review takes each of these challenges seriously and locates the curriculum's response in tensions.html.
The dates that matter are these. Judy Singer's master's thesis at the University of Technology Sydney in 1998 introduced neurodiversity as a sociological term, building on the Independent Living movement and the contemporary autism-rights organising that Sinclair, Baggs, and others had been doing through ANI. Murray, Lesser, and Lawson published the monotropism account in Autism in 2005, proposing attention as the cognitive primitive that organises autistic experience.14 Baron-Cohen's extreme-male-brain hypothesis appeared in 2002 and remained the dominant theoretical statement in clinical autism research for the better part of a decade. Milton's double-empathy paper landed in 2012. The Kenny et al. language-preferences survey in 2016 documented Anglophone autistic preference for identity-first language, a methodological finding that the field has since been arguing about whether to treat as universal.17 Hull's masking studies appeared from 2017 onward, with the CAT-Q masking questionnaire validated in 2018.18 Yergeau's Authoring Autism in 201719 and Walker's Neuroqueer Heresies in 202120 opened the rhetorical and queer-theoretical fronts. Autism in Adulthood, founded under Christina Nicolaidis's editorship in 2019, became the affirming flagship journal. Raymaker's autistic-burnout paper appeared in the new journal's first volume in 2020.5 Crompton's diffusion-chain trilogy followed the same year. Pellicano and den Houting's paradigm paper appeared in 2021. Bottema-Beutel's avoid-ableist-language guidance landed the same year and has since become the most-cited piece of methodological discipline in the field.21 Chapman's Empire of Normality arrived in 2022, the same year Hersey's Rest Is Resistance reached a mass audience and Devon Price's Unmasking Autism moved liberation-pedagogy framing into autistic-adult trade reading. The 2024 Botha-Chapman-Onaiwu-Kapp-Stannard Ashley-Walker correction on the origins of the neurodiversity movement returned credit to the autistic-of-color, intellectually-disabled, and AAC-user organisers whose work the field had been routinely attributing to Singer alone.22 Flower's 2025 Delphi study on autism-affirming therapy is the most recent attempt to operationalise the affirming paradigm into therapist-facing practice guidance.23. The genealogy document carries the full timeline with influence lines and paradigm shifts.
Four paradigm shifts have happened in the literature since Milton, and each shift has cost the field something. The deficit-to-neurodiversity shift, accelerated by the 2018-2022 wave, cost the field a clear consensus about what counts as autistic suffering that the clinic should treat. The theory-of-mind-to-double-empathy shift, completed by Crompton 2020, cost the cognitive-science programme its dominant explanatory framework and has not yet produced a unified replacement.2 The individualist-clinical-to-ecological-niche shift, completed by Chapman 2021 and 2023, cost the field its claim to political neutrality, since once you describe neuronormativity as a regulatory ideal produced by industrial capitalism you have made a political argument that no quantitative outcome study can resolve.4 The researcher-on-autism-to-autism-researcher shift, accelerated by AASPIRE's community-based participatory research methodology and Botha's 2022 autism research is in crisis piece,24 has cost the field the methodological convenience of treating autistic adults as objects of research rather than collaborators, and has begun to surface the citation-cartel and hyper-credentialing dynamics that excluded autistic-of-color, intellectually-disabled, and non-Anglophone scholars from the prior canon. Each shift has been a gain. None has been free.
The curriculum belongs, in the schools-as-positions taxonomy, to a position the literature does not yet have a stable name for: a synthesis of the autistic-phenomenological position with the liberation-pedagogy position, held inside a closed-cohort clinical container that imports the abolitionist critique as facilitator self-audit rather than as theoretical decoration. The curriculum is not abolitionist; it accepts that ten-session containers, intake batteries, and clinical risk management are real constraints under which adults will continue to gather. It is not pure accommodation; the readings of Hersey, Taylor, and Price commit the curriculum to a political claim about why neurodivergent adults are exhausted, namely that the exhaustion is produced rather than constitutional. It is not pathology paradigm; it does not treat the affinity group's purpose as the remediation of social-skills deficits. It does, however, retain four loyalties to the clinical: psychometric intake (CAT-Q, ASIS, MQ), structured session architecture (a five-part template repeated across ten weeks), explicit risk management (active-suicidality contraindications, between-session check-ins), and outcome instrumentation in a register the affirming paradigm has begun to develop but has not yet stabilised.23
The position has costs the curriculum has to own. The first is the cohort-honesty cost: the design serves a thin slice of the neurodivergent population, namely verbally fluent, late-diagnosed Anglophone adults with sufficient executive resource to attend a ten-week closed cohort. It does not serve non-speaking autistic adults, intellectually disabled neurodivergent adults, or adults whose primary communicative modality is AAC. The review names this cost rather than dressing it up. The second cost is the BIPOC-late-diagnosis under-treatment cost, surfaced in All The Weight of Our Dreams25 and the work of Onaiwu, Brown, and Giwa Onaiwu, which the curriculum's reading list addresses but which a single ten-session group cannot resolve. The third cost is the Hersey-secularization risk: Hersey's Rest Is Resistance is grounded in Black liberation theology and womanist tradition, and a curriculum that imports her into a secular pan-ND container without naming the theological grounding flattens what it claims to honor. The review's recommended ruling, carried in tensions.html, is to honor Hersey inside the lineage that produced her: Lorde's A Burst of Light, Brueggemann's Sabbath as Resistance, Howard Thurman's Jesus and the Disinherited, Amos Yong's theological work on disability, John Swinton's pastoral-theology corpus, and Robin Wall Kimmerer as a parallel Indigenous source on rest as ecological reciprocity. The fourth cost is the autism-centric evidence-base cost: roughly eighty per cent of the literature on neurodivergent group work is autism literature, with adult ADHD a thinner secondary base and the rest of the ND umbrella (Tourette's, OCD-as-ND, dyspraxia, dyslexia) thinly represented in adult clinical research. The pan-ND design choice is licensed by Dwyer 2024's trans-diagnostic monotropism evidence26 and by Wise's practitioner architecture for ND-affirming DBT,27 but the empirical floor under it is real and the review names it.
Irvin Yalom's eleven therapeutic factors for group psychotherapy were worked out for verbal, predominantly neurotypical adult outpatient cohorts in the 1960s and 1970s, and the question for an affinity-group designer is which factors translate cleanly, which translate with adaptation, and which become iatrogenic when imported without translation. Universality, instillation of hope, altruism, imparting of information, and the existential factors translate cleanly: a neurodivergent adult sitting with other neurodivergent adults receives the universality factor more potently than they have ever received it elsewhere, since the stranger-status they carry into mixed-neurotype rooms relaxes when the room is matched. Cohesiveness translates, but only with neurotype matching, since the Crompton diffusion-chain evidence shows that information transfer between autistic-autistic dyads exceeds cross-neurotype information transfer.2 Catharsis translates with pacing and window-of-tolerance management, since autistic interoception literatures suggest that affective overwhelm in unstructured catharsis can produce shutdown rather than integration. Two factors become iatrogenic without serious adaptation: interpersonal learning and the corrective recapitulation of the family group, both of which were designed around neurotypical interpretive heuristics that the literature now recognises as cross-neurotype-failure-prone. Development of socialising techniques, which Yalom retained from earlier social-skills traditions, is directly iatrogenic in the affinity context, since the entire point of an affinity group is that the masking labour the wider world demands is the thing being interrupted. The curriculum's design takes Yalom's factors as a checklist of what to translate, and the frontier document records the translation move as a candidate contribution.
AANE separates ND and ally groups by default. Hirvikoski's PEGASUS protocol includes significant others as supporters whose role is to scaffold the autistic protagonist rather than to occupy primary speaking time.28 No published curriculum the bibliographic mapper located treats allies as first-class participants with weighted speaking time, witnesses-not-protagonists framing, and explicit risk-mitigation architecture for the asymmetries that arise when a partner, a sibling, or an adult child sits in a room where they are not the diagnosed party. The curriculum's first-class-allies move is, in literature terms, a candidate frontier contribution rather than an established practice. Its risk-mitigation architecture borrows from the couples-therapy literature on bilateral accountability and from the disability-justice literature on access intimacy29 rather than from the autism-group literature, which has not yet asked the question. Whether the move survives empirical contact with cohorts is a question this review cannot answer; it can only locate the move at a defensible point in the literature and name the risks the architecture is asking to absorb.
Tema Okun's White Supremacy Culture document, distributed since 1999 and revised continuously since, names a constellation of organisational habits that the curriculum borrows as facilitator self-audit rather than as content delivered to participants: perfectionism, sense of urgency, defensiveness, worship of the written word, quantity over quality.30 bell hooks's engaged pedagogy31 supplies the relational architecture. Mia Mingus's access intimacy29 supplies the ethical horizon. adrienne maree brown's Emergent Strategy supplies the small-is-good design heuristic.32 Hersey's Rest Is Resistance, Taylor's The Body Is Not an Apology,33 and Price's Laziness Does Not Exist34 and Unmasking Autism35 become primary readings rather than epigraphs, with the five-part session structure designed to braid text engagement, somatic practice, dialogue, action commitment, and closing ritual. The literature has not yet attempted this integration in published form, which means the curriculum is asking the field to recognise a candidate frontier contribution. The cost is that the curriculum cannot yet point to outcome data; the gain is that the curriculum can point to a clearly articulated theoretical commitment that the integration is meant to test.
The review will not claim that ND-affirming therapy outperforms cognitive-behavioural therapy in head-to-head trials, because Linden 2023's network meta-analysis does not contain a single ND-affirming arm.36 The review will not claim that polyvagal theory grounds the curriculum's somatic moves, because Porges's polyvagal claims have been contested in the recent neuroscience literature, with Grossman 2023 and P. Taylor 2022 raising serious objections to the cardiac-vagal mechanism Porges proposed.37 The curriculum has therefore dropped polyvagal framing and grounds its somatic and pacing logic in predictive-processing and allostatic-load literatures instead.38 The review will not claim that allies-inclusive groups are safer or more effective than ND-only groups, because no head-to-head evidence exists. The review will not claim that the curriculum's intake battery is psychometrically validated for the pan-ND population, because the CAT-Q and ASIS are autism-specific instruments18 and the trans-diagnostic monotropism literature is younger than the evidence required to make that claim.
What the review will claim, and what the AE-tier table in methodology.html documents in full, is that autistic-autistic dyadic rapport is well-supported (AE2), that identity-first language is preferred among Anglophone autistic adults (AE2), that masking is associated with mental-health burden in cross-sectional and longitudinal data (AE3, with the causal-language hedge), that autistic burnout is now sufficiently operationalised for clinical recognition (AE3), that monotropism has gained empirical traction (AE3), that group cohesion in autistic-only spaces is supported by qualitative and small-quantitative work (AE3), that rest as clinical intervention is theoretically licensed (AE4) but not yet RCT-tested in autistic burnout (AE5), and that ND-group work for active suicidality is contraindicated in the absence of crisis-protocol scaffolding (AE3). The curriculum's claims are pinned to that grading.
The review's visible absences, named in full in methodology.html, are these: non-Anglophone scholarship is thinly represented because the citation-graph review reproduces the field's English-language dominance; AAC-user theorists (Mel Baggs, Ido Kedar, Naoki Higashida, the broader non-speaking-theorist tradition) are present in the review but not weighted as heavily as their importance to the field warrants; intellectually disabled autistic adults are under-served by the entire literature, not only this review; behavioural-analytic affirming voices (the compassionate-ABA work of Hanley and Rajaraman) appear thinly because the review's home territory is the affirming and abolitionist traditions rather than the reformed-behaviourism dialogue that work belongs to;39 faith-based and theological critics of secular Hersey-uptake (Brueggemann, Howard Thurman, Amos Yong, John Swinton) are named in the curriculum's reading list but not engaged in depth in the body of the review; Tourette's, OCD-as-ND, dyspraxia, and dyslexia adult clinical literatures are thinly represented in the umbrella; the citation-graph methodology itself reproduces high-impact-factor journal weighting and citation-cartel effects that the field's own meta-research is beginning to surface.22 The review does not paper over these. It names them, locates them, and treats them as conditions of the curriculum's design rather than as curiosities to be resolved later.
The field has become, in the fourteen years since Milton's paper, a discipline with an internal politics. The internal politics is not a defect; it is the trace of a body of work that has begun to take seriously the question of who is permitted to make claims about whom. The curriculum that this review's literature is meant to support sits inside that politics rather than outside it, takes a position rather than gesturing at neutrality, and accepts the costs of its position rather than dressing them as features. Whether the curriculum survives empirical contact is a downstream question. The review's task is finished when the curriculum can be located, defended, and held accountable.
Milton sat outside the academy in 2012 reading theory-of-mind tasks back to himself and noticing that the tasks had been written by people who had not read themselves first. The reviewers who returned his paper let it through. The field that has formed since has not yet decided whether it knows what to do with the room he opened. This curriculum is one attempt to find out.
Footnoted references in this essay correspond to entries on the references page. Inline numbering is footnote-style rather than parenthetical APA for readability; full APA-7 entries are available on the references page.