Corps de l’article

Challenging how persons at the intersection of race and dis/ability are regularly misrecognized, Kuppers (2011) recounts an experiment in which the artist collective The Olimpias visited UCLA’s Telematic Laboratory, which is equipped with a studio named the Cave that contains a 24 camera imaging system designed to assemble human movements into a single 3D avatar. Dancing naked within this space, performers discovered that this imaging system was habitually calibrated to best render standing dancers with pinkish skin tones. Reflecting on their experiment in (mis)recognition, Kuppers notes how the Olimpias “push[ed] a form to where it breaks down,” while also highlighting “what faces, bodies, narratives, or techniques stress it and pull it out of shape” (p. 8). The effects of this experiment reappear again and again in the routine misrecognition of non-White faces by facial recognition algorithms. While such errors could be dismissed as unintended programing mistakes, they should instead be understood as forms of sociopolitical violence, the effects of which are best evidenced in a 2015 apology issued by Google after it was discovered that its facial recognition algorithm misidentified Black human faces as gorillas.

While these examples may seem like a strange way to introduce an essay about autism’s racializing assemblages, they mirror how “the existing public discourse about autism is glaringly incomplete” (Onaiwu, 2020, p. 244). Reflecting on the publication of All the Weight of Our Dreams: On Living Racialized Autism, an anthology authored and edited by autistics of colour, Onaiwu (ibid.) elaborates further: “The experiences, stories, and images of people of colour on the autism spectrum are conspicuously absent both in the public view and even within the sphere of disability” (p. 244). Just as the Cave and facial recognition algorithms are designed without intentionality toward recognizing non-White faces, bodies, narratives, or techniques, clinical and cultural definitions of autism are developing in ways that deny the very existence of autistics of colour. Dorothy Groomer recalls this ontological erasure in the documentary film Refrigerator Mothers (2002):

According to my doctor, my son could not be autistic. I was not white, it was assumed that I was not educated and therefore he was labeled emotionally disturbed. Here your child has a disability that you recognize and they said, nah, you can’t be that. You can’t even be a refrigerator mother [laughs], the irony of it all.

quoted in McGuire, 2016, p. 41; emphasis added

Groomer’s recollection that her son could not be autistic because she was non-White and uneducated instantiate what Weheliye (2014) describes as racialization, that is, “those sociopolitical relations that discipline humanity into full humans, not-quite-humans, and nonhumans” (p. 5). This hierarchical ordering of humanity is not only apparent in the deliberately indifferent uses of facial recognition algorithms but also in data about which children receive autism diagnoses and which children are labelled as emotionally disturbed (ED) or intellectually disabled (ID). In addition to reflecting the ontological erasure of her son’s autism and herself as a knowing subject, Groomer’s recollection also testifies to autism’s racializing assemblages. According to Weheliye, “racializing assemblages articulate intensities between human physiology and flesh, producing racial categories, which are subsequently coded as natural substances, whether pure or impure, rather than as the territorializing articulations of these assemblages” (pp. 50–51). The territorializing effects of autism are evident in Groomer’s linking autism to Whiteness and associating the ED label with Blackness. Thus, despite the theory that so-called non-nurturing or “cold” mothers caused autism being roundly discredited, not even the condemned psychological label “refrigerator mother” was available to Groomer.

Groomer’s recollection demonstrates that autism knowledge cannot be understood apart from the racialization of so-called high-incidence or subjective disabilities (i.e., ED or ID). As Baynton (2001) insists, to treat race and dis/ability as if they were unconnected is to articulate humanity as an evolutionary hierarchy within which non-White races are rendered as non- or not-quite-human because of their presumed ontological proximity to disability. Disability labels are thus not isolated descriptors of human difference; they are ways of articulating humanity as a “relational ontological totality” (Weheliye, 2014, p. 32). Within this totality, autism knowledge is assembled at the juncture of a cultural and clinical feedback loop. Within this “looping effect of human kinds,” more critical attention is needed to autism’s racializing assemblages than to the unraced “making up” of autistic people (Hacking, 2007). If Hacking (2009b) is correct that “we are participating in a living experiment in concept formation of a sort that does not come more than once in a dozen lifetimes” (p. 506), then it is important that autism knowledge is expansive enough to avoid the sociopolitical violence that the above examples of technological racialization document. Groomer’s recollection provides an entry into a questioning of autism’s racializing assemblages. In what follows, autism is further interrogated as a racialized disability classification by providing evidence of disproportionality and delayed autism diagnoses for children of colour, especially for Black and Brown children, as well as through elaborating “autistic whiteness” (Heilker, 2012). These sections examining autism’s racializing assemblages, in turn raise questions about how autism is defined and narrated.

Evidence of Disproportionality and Diagnostic Delays

Groomer’s recollection provides motivation to give critical attention to how disability labels classify people differently, a process referred to as disproportionality and which results in children of colour being both overrepresented in certain disability classifications (i.e., ED and ID) and underrepresented in others, like autism. Establishing the need for an anthology about autism and race, Onaiwu (2017), addresses this concern: “‘Autism is autism, right? Does race really make a difference?’” (p. x). As the authors collected within All the Weight of Our Dreams argue: no, autism does not adhere to people of colour in the same way that it coheres around Whiteness, which means that yes, race really does matter. Attending to these questions further necessitates a contrapuntal reading that “links ideas and practices that are regarded as being opposites or in contradiction, thereby revealing points of contact across these bodies of work that are not readily apparent” (Artiles, 2013, p. 332). The diagonal reading of autism knowledge begun in this section follows similar efforts to read autism against the grain toward fugitive practices of resistance (Roscigno, 2019). Onaiwu’s questioning of autism’s relation to race invites such readings in its suggestion that because autistics of colour are located within multiple and overlapping systems of oppression, they experience both a double bind that compounds structural disadvantages and a double perspective that facilitates connecting discrepant points of contact (e.g., autism, racialization, Whiteness, and ED or ID classifications). This section begins a contrapuntal reading of autism’s racializing assemblages by providing evidence of disproportionality and delayed autism diagnoses for children of colour while also suggesting reasons to be skeptical of deficit-based solutions to these problems.

There are an increasing number of studies which provide evidence that children of colour are either misdiagnosed or receive delayed autism diagnoses. Within this scholarship, studies conducted by Mandell, et al. (2002; 2007; 2009) are regularly cited. While these studies rely on clinical rather than school-level data, they correspond with research that students of colour are more likely to be both placed within segregated special education environments and overrepresented within ED or ID disability classifications (see Harry & Klinger, 2006; Losen & Orfield, 2002). A first step for disproportionality research is determining whether observed diagnostic differences in race and ethnicity result from true group differences in prevalence (making autism a disability classification that is not applicable to children of colour) or from disparities in diagnostic practices (making autism a disability classification that is not applied to children of colour). While autism is a complex neurodevelopmental disability without any known biomarkers, according to Mandell, et al. (2002; 2007; 2009), there are no statistically significant differences by race or ethnicity in the prevalence of autism, which means that most research is guided by questions of why autism diagnoses are applied less frequently to children of colour.

For example, Mandell, et al. (2009) found that children of colour with IQs below 70 are less likely to be diagnosed as autistic than White children with similar IQs. This is because children of colour within this IQ range are less likely to receive further assessments that might support autism diagnoses. A reason for this diagnostic disparity is a belief amongst clinicians that children of colour are statistically more likely to be intellectually disabled, an ontological presumption that diagnostically echoes Groomer’s above recollection. The existence of other labels allows clinicians to individually assess children of colour while stopping short of evaluations that could produce autism diagnoses. Because diagnosing autism is complicated by the fact that it co-occurs with other symptoms, Mandell et al. (2007) note that the American Psychiatric Association (APA) specifies that autism should be ruled out before reaching other conclusions (i.e., ID) and prior to assigning children conduct or adjustment disorders (i.e., ED). However, this same study found that Black children were three times more likely than White children to receive diagnoses other than autism when presenting with co-occurring symptoms, indicating that this best practice regularly goes unfollowed. Disaggregating such findings, Bobb (2019) notes that Black girls are even less likely to receive autism diagnoses.

In addition to providing evidence of misdiagnoses, studies into the nexus of autism and race also indicate that even when children of colour receive autism diagnoses, it is often later than White children, which can result in autistic children of colour not receiving free and appropriate educations as mandated by the Individuals with Disabilities Education Act (1990). As Mandell, et al. (2002) note, while there is evidence of delayed diagnoses for both White and non-White children, Black and Brown children are often diagnosed later than White children. Furthermore, White children are more likely to both enter the diagnostic process earlier and remain within this process for less time than Black and Brown children. For instance, 72% of White children received an autism diagnosis upon their first clinical visit compared to 57% of Black children, whose diagnoses can require three times as many clinical visits. Importantly, earlier entry into the diagnostic process does not account for disparities in the amount of time Black and Brown children spend within this process, nor redress why these children are diagnosed one to two years later than White children. It is more likely that children of colour receive a “wait and see” approach from clinicians, which can result in schools misrecognizing, and then labelling, students of colour as ED due to perceived behaviour problems.

While Mandell, et al. (2002; 2007; 2009) advocate that clinicians working in underserved communities and with marginalized families should be educated to better recognize autism in children of colour and trained against relying on statistical presumptions that intellectual disabilities and conduct or adjustment disorders are more likely for children of colour, they stop short of interrogating how autism has been historically constructed since Kanner’s (1943) foundational study as a White condition. Critiquing such diagnostic circularity, Bobb (2019) argues: “As fewer intellectually abled Black youngsters are diagnosed, professionals come to believe that mainstream Black communities are less affected by autism … It is a self-fulfilling expectation leading to under-diagnosis” (p. 39). Despite special education being complicated by questions of how student identities manoeuvre within, by, and for educational ecologies, deficit-based assumptions continue to inform how current research explains disproportionality and diagnostic delays (Artiles, et al., 2010). For instance, Dababnah, et al. (2018) conclude that disparities in autism diagnoses result from: 1) primary healthcare providers inattentiveness to Black caregivers’ concerns, 2) perceived racism and poor caregiver-provider interactions, 3) stigma in the Black community about autism, and 4) legal and custodial issues amongst Black caregivers. Artiles, et al. (2010) are skeptical of such recommendations. In addition to exhibiting circular reasoning, there is also a negative feedback loop that individually locates diagnostic responsibility. Explaining disproportionality or diagnostic delays by citing, for example, data that Black and Brown families are less likely to have regular access to pediatric visitations leaves unanswered questions of why children of colour have more difficulty receiving autism diagnoses during visitations or why this diagnosis more easily coheres around Whiteness. Recommendations for improving healthcare access, reducing clinician bias, or increasing parental awareness also leave autism’s racializing assemblages uninterrogated, raising the definitional question: What is autism?

What Is Autism? Definitional Questions

While there are numerous psychological theories that claim to answer this question, few philosophers have waded into this classificatory debate. One problematic example is Barnbaum (2008), whose deficit-based bioethical philosophy equates having a theory of mind (ToM) to being human. Barnbaum then wrongly deduces that because autistics lack a ToM, they cannot be human. As Yergeau (2013) points out, this definition requires the circular reasoning that ToM only exists because autistics do not possess it. Taking a different tact, Cushing (2013, 2018) defines autism according to a metaphysical distinction between realism and constructivism. Similar to the medical model of disability, realists maintain that autism is knowledge-independent even if its essence has yet to be understood entirely by psychology, neurology, or genetics. As such, autism is a naturally occurring condition similar to Down syndrome, even if it has only recently been named. In contrast, constructivism claims that autism is conceptually characterized by and constructed through historical knowledge. The claim that autism is knowledge-dependent corresponds with the social model of disability.

Following this distinction, most contemporary definitions of autism are realist and can be further divided according to categorical questions about autism’s genus and differentia, that is, what kind of thing is autism and what makes autism its own special instance of that kind. Determining the former requires asking about the psychological, neurological, or genetic causes of autism, whereas establishing the latter necessitates isolating which behaviours categorically constitute autism. Cushing (2013) suggests that autism can be differentiated according to whether behaviours are bundled or clustered. Kanner (1943) is an example of using the former, and the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) (American Psychiatric Association, 2013), specifically the categories of “social communication and social interaction” and “restrictive, repetitive behavior, interests, activities,” is an example of using the latter. A difficulty with defining autism as a bundle of behaviours is vagueness: How many and what combination of behaviours must be present to merit an autism diagnosis? Additionally, given that autism can co-occur with other symptoms, how do clinicians and educators isolate behaviours? Are there behaviours not identified by Kanner that are suggestive of autism? More problematic, however, is that defining autism as a set of behaviours perpetuates erroneous fears of an autism epidemic, which supports efforts by the organization Autism Speaks to find a cure (McGuire, 2016).

Erroneous fears of an autism epidemic are not alleviated by clustering behaviours. The DSM-5 is an example of how autism is clustered by broad types of behaviours (categories A and B) that unify autism into a single spectrum diagnosis. While clustering allows for more diagnostic flexibility, it lacks specificity. For example, how can two children with different behaviour types share the same diagnosis? Or, more pressing, how can two children with similar behaviour types have different diagnoses? Clustering thus presents a difficulty of ascertaining how behaviours categorized as “social communication and social interaction” (category A) and “restrictive, repetitive behavior, interests, activities” (category B) definitionally cohere into a single diagnosis. Additionally, given that children must meet all criteria from category A and at least two criteria from category B, the problem of vagueness persists, as there are an exponential number of possible diagnostic combinations, as does the belief that interventions, like applied behavioural analysis (ABA), can cure autism by training autistic children to be more intentional in their movements and more communicative in their social interactions.

While both bundling and clustering define autism as its behaviours without asking why behaviour or these specific behaviours can be categorized as such, more problematic are efforts to define autism by analogizing it to race. For example, in parsing constructivism, Cushing (2013) posits that autism is socially constructed similar to race, that is: “[a] phenomenon that arbitrarily groups traits together that have no underlying common cause, genetic or otherwise” (p. 24). This comparison maintains the conspicuous absence of autistics of colour from the public view and the sphere of disability. Such analogies also perpetuate a racialization mythos that misrecognizes autistics of colour as emotionally disturbed or intellectually disabled because autism is always analogous to, but never inclusive of being Black and Brown. Efforts to analogically define autism serve to abstract the lived experiences of autistics of colour while leaving uninterrogated assumptions of White innocence that normalize autism as a White condition by racializing Black and Brown children as ED or ID. Embedded within such definitions is a “racecraft of disability labeling” (Kearl, 2019) that imagines autism first as a White condition, which is then compared to racialized identities. This strange manoeuvering further coheres autism around Whiteness by analogically linking autism knowledge and race rather than interrogating autism’s racializing assemblages. Traversing this hierarchical terrain requires intersectional reasoning, like dis/ability critical race studies (DisCrit) (Annamma, Connor, & Ferri, 2013), which contrapuntally attends to how racism and ableism validate and reinforce each other and how Whiteness both habitually narrates disability classifications and arbitrates humanness.

Evidence of Autistic Whiteness

While perhaps a strange beginning, the introduction to this essay implicitly critiques conventional geographical understandings that locate autism in Silicon Valley and which culturally situate autism within STEM fields and philosophy departments. For example, Steve Silberman (2015) introduces his book Neurotribes: The Legacy of Autism and the Future of Neurodiversity by recounting his realization that reporting on technology for Wired magazine also meant indirectly (un)covering autism’s “geek chic.” Anderson and Cushing (2013) likewise introduce The Philosophy of Autism by sharing how their son’s diagnosis alerted them to the prevalence of autism within their academic field. Embedded within this diagnostic cultural geography are media representations, like the techno-savant character Sheldon Cooper from the sitcom The Big Bang Theory (2007–2019), that narratively assemble together autism, intelligence, and Whiteness. The linking of autistic children to educated White professionals has persisted since Kanner (1943), a historical failure that Arnold (2017) describes as “Autism’s Race Problem.” Sheldon offers a contemporary example of how autism is characterized in media representations as well as a rhetorical exemplar of how a culturally constituted aesthetics of Whiteness habituates autism’s diagnostic space.

While tropes like the techno-savant are terrible psychology, they are nevertheless “an aspect of our times that we are only beginning to think about” (Hacking, 2010, p. 654). Going further, Onaiwu (2017) points out how autism’s racializing assemblages inform how autistics of colour understand themselves:

Autism = (white male-presenting) toddler wearing a Thomas the Train t-shirt; autism = (white male-presenting) quirky teenage gamer; autism = (white male-presenting) geeky computer programmer; autism = (white male-presenting) adult rocking and staring off into space … autism = Temple Grandin, puzzle pieces, ABA therapy, and Autism Speaks.

p. xv

These equations make possible critical interrogations of both how autism is culturally assembled and the clinical effects of these assemblages. Returning to All the Weight of Our Dreams, Brown (2017) materially responds to Cushing’s (2013, 2018) metaphysical definition of autism knowledge by noting how intersections of race and dis/ability can “indelibly mark the lives of people of color and disabled people” to such an extent that constant reminders are needed that autistics of colour are real.

The need for such reminders raises questions of how Whiteness narrates who is autistic, a consideration that Matthews (2019) explores in arguing that Sheldon Cooper conserves a White male hegemony by “establishing a viable survival model for the beleaguered white male” (p. 67). As an embodiment of a human-digital hybrid aesthetic, Sheldon “cement[s] whiteness as misunderstood and persecuted, yet also as necessary and inevitable” (ibid., p. 65). The visual celebration of Sheldon’s paleness, thinness, and computer-like symmetry defines what autism looks like in rhetorical opposition to dark, thick, and asymmetrical bodies. While the show does not identify Sheldon with Aspie supremacism, The Big Bang Theory nonetheless equates Whiteness with intellectual superiority by portraying Sheldon as super-abled, as exceeding the limitations of neurotypicality, thus “pav[ing] the way for the apotheosis of the white male who, significantly, has the most to lose from cultural trends toward neuro and ethnic diversity and the most to gain from portrayals of autistic techno-savants as intellectually gifted and relentlessly white” (ibid., p. 62). While Aspie supremacist rhetoric does not overtly mention race, “the general orientation and thrust of this discourse aligns uncomfortably well with commonplace tropes of white superiority” (Heilker, 2012, para. 8). The difficulty with Aspie supremist rhetoric, then, is that it situates autism within an evolutionary hierarchy.

While racializing assemblages of Sheldon Cooper as intellectually superior are troubling, they are most likely performed for and with neurotypical identification in mind. Still, Hacking (2009b) notes the danger of such characterizations: “It encourages the image of the autist as gifted with a secret knowledge or wondrous powers. It can lead to sliding from a genuine fact to a foolish fiction” (p. 514). Matthews’ (2019) analysis of a clinically informed, culturally constituted autism aesthetic echoes this concern as well as other critiques of how autism more easily coheres around Whiteness. Heilker (2012), for instance, argues that the rhetorical power of characters like Sheldon is that they invite armchair autism diagnoses of historical figures, the list of which is exclusively White. Assemblages of autism, Whiteness, and intelligence thus succeed in perpetuating the myth that one must be “a white, upper-middle class man from a western country in order to be ‘legitimately autistic’” (Gardiner, 2017, p. 11). This myth continues to narrate autism through the form of the “shiny autistic,” that is, that characters like Sheldon, but also autistics like Temple Grandin, definitionally embody autism. Autism Speaks furthers this myth by foregrounding their White, neurotypical leadership and regularly featuring White autistics in their public service announcements.

Echoing Onaiwu’s (2017) equations, above, such representations invoke a circular reasoning whereby “autism presents itself as a white condition because it is sought out as a white condition, which, in turn, reinforces it as a white condition and incentivizes it as a white condition within mainstream media” (Matthews, 2019, p. 64). As already noted, statistical presumptions racially assemble who is diagnosed as autistic, who is diagnosed as intellectually disabled, and who is diagnosed with a conduct or adjustment disorder. More explicitly, Matthews (ibid.) cites a clinical study in which researchers purposefully screened for White autistic participants. In the opposite direction, Draaisma (2009) observes how the White autistic character Simon Lynch, in the film Mercury Rising (1998), was developed in consultation with the University of Chicago’s head of Pediatric Psychiatry to ensure this portrayal’s clinical accuracy. It is almost as if Kanner’s (1943) case studies are being repurposed to develop legitimately autistic characters, indicating the power of autistic Whiteness to habitually narrate who is autistic.

Following Ahmed (2007), education can be understood as a series of habit spaces assembled by and through Whiteness. Envisioned from the founding vantage of child psychologists like G. Stanley Hall, special education was designed to eugenically control for the threats that non-White children posed to civilization (Baker, 1998, 2002). On this point, Oaks’ (2005) history of tracking, best illustrated by the political cartoon “School Begins” (1899), remains instructive for understanding the present educational reach of Baynton’s (2001) historical analysis, as well as how, extending Weheliye (2014), disability classifications continue to hierarchically order students according to protean evolutionary standards of humanness. US education not only tracks children of colour into segregated educational spaces, it also introduces new diagnostic spaces for White children. For example, Sleeter (1987) shows how the specific learning disability emerged from a desire among White, middle-class parents to see that their academically failing children were not labelled according to the existent psychological categories used for children of colour. Similarly, Metzl (2010) demonstrates how schizophrenia was initially used to describe an “‘emotional disharmony,’ that negatively impacted White people’s abilities to ‘think and feel’” (p. xii). It was only after the civil rights movement that this diagnosis shifted from being a general descriptor of disharmonic Whiteness to a specific category of Black anti-sociality. Leonardo and Broderick (2011) similarly note the capacity of Whiteness to expand in oppositional relation to Blackness, creating conditions of “immunitary whiteness” that negate the educational lives of students of colour (Kearl, 2019).

These examples support Ahmed’s (2007) phenomenological thesis that when spaces conform to Whiteness, whoever can inhabit them is more likely to be recognized. Autism is thus a diagnostic space whose present coherence is constituted by its historical inhabitations. This critique is shared by All the Weight of Our Dreams. Because autism has been historically narrated as a White condition, its diagnostic space has culturally cohered around Whiteness and its clinical habituation has escaped counter-narration. Recognizing the narrative habits of diagnostic spaces requires regular reminding. All the Weight of Our Dreams is an example of this labour. Autistic Whiteness is a needed relational analytic that shows how diagnoses cohere around their inhabitants toward the construction and maintenance of racializing assemblages of education’s ontological totality. That disability labels not only differ according to race but also use race to cohere disability classifications means that it is impossible to answer the question of who is autistic without asking the relational question of how autism is narrated.

Who Is Autistic? Narrative Questions

As already alluded to, the best answer to the question “Who is autistic?” comes from autistics of colour themselves, whose various life-writings provide counter-narratives to definitions of autism restricted to analogical reasoning and constrained by representations of autistic Whiteness. Broderick and Ne’eman (2008) provide a further counter-narrative that autism is a metaphor for neurodiversity, which contests efforts by Autism Speaks to narrate autism as a disease. McGuire (2016) also questions autism’s cultural veridiction by analyzing it as a relational space that is interactional, historically contingent, socially mediated, and sociopolitical. Treating autism as an “object of interpretive analysis” (p. 21), McGuire moves questions of autism knowledge from analogical definitions toward genealogical interrogations of how psychological, neurological, and genetic theories variously constitute the autistic subject as disordered. Rather than asking what is autism, McGuire’s critical genealogy invites questions about how autism’s racializing assemblages narrate who is autistic. Heeding this invitation, this section argues that previous evidence of disproportionality and delayed diagnoses for children of colour as well as the above elaboration of autistic Whiteness demonstrate the need for narratives that tell a different truth about autism and that tell this truth differently.

Attending the above question, Hacking (2007) notes that “we make ourselves in our own scientific image of the kinds of people it is possible to be” (p. 305). This observation posits a dynamic nominalism consisting of complex interactions among classifications, people, institutions, knowledge, and experts. For Hacking, autism exemplifies this interactive process. Beginning with an observation that autism diagnoses have steadily increased, Hacking posits a two-part distinction, first, between (a) there not being any high functioning autists in 1950 and (b) experiencing oneself as a high functioning autist not being an ontological possibility in 1950; and second, between (c) there being many high functioning autists in 2000 and (d) experiencing oneself as a high functioning autist being an ontological possibility in 2000. The second half of this two-part distinction (b and d) demonstrates the looping effect, which Hacking arrives at by hypothesizing that some of the 11 children diagnosed as autistic by Kanner (1943) overcame their autism to become exemplars of a new kind of person – the high functioning autist. These “recovered” autists then allowed others who may not have previously understood themselves to be autistic, but perhaps wondered about their way of being in the world, to declare: “‘That’s me!’” (Hacking, 2007, p. 304).

While this use of the looping effect differs from previous examples of circular reasoning, the effects are similar. For instance, Hacking (ibid.) pays little attention to the first two halves of this distinction (a and c) except to declare as false claims that there were not autistic people prior to 1943. Considering this claim seems less important than interrogating who is assembled as autistic by such armchair diagnoses. For example, reflecting on a Google image search for “autism” in which 47 of the first 50 images exclusively featured White people, Heilker (2012) questions the likelihood that autistics of colour will individually declare “That’s me!” Hacking (2010) has also discussed how the manifold of autism fiction mirrors the rise of the Internet. Central to this analysis is a desire to replace autism metaphors like “alien” or “shell” with the interactive description, “classification-in-motion” (ibid., p. 640). Yergeau (2017) also notes how the Internet makes this motion possible in observing how autists find both themselves and other autistics through various blogs and chatrooms. This is Hacking’s (2010) point as well: both autism fiction and the Internet are making it ontologically possible to experience oneself as autistic in ways not possible in 1950, which is changing how we are learning to “talk” about autism, thus making it possible for autistic people “to be, to exist, to live” (Hacking, 2009b, p. 501). Questions remain, however, regarding whose lived experiences will narrate this motion and whether new autism knowledge will tell a different truth about who is autistic and tell the truth of autism’s relational ontological totality differently.

Who Is Human? Re-Assembling Autism Knowledge

Yergeau’s (2013) critique of Barnbaum’s (2008) bioethical definition of autism helps to conceptualize the question “Who is human?”:

The whole point of ToM is that autistic people do not have it … The whole point of ToM is that humans do have it … The whole point of ToM is that autistics are not human … This last item – that the autistic is not human – is the unspoken premise of the ToM enthymeme. There exists an autistic mind, but not an autistic human”

paras. 57–60; original emphasis

This syllogism exemplifies how autism knowledge is governed by an evolutionary hierarchy designed to render autistics as non- or not-quite-human. Despite being helpful toward thinking about autism as a classification-in-motion, Hacking (2009a, 2009b, 2009c) not only leaves the cultural and clinical Whiteness of autism’s habit spaces uninterrogated but also misses Yergeau’s point that definitions of autism participate in a violent sociopolitical hierarchical ordering of humanity. Manning (2019) elaborates further: “The violence is in the exclusion of black life, of neurodiverse life, from neurotypicality and the category it upholds at all costs, the human” (p. 18). In addition to providing evidence of the sociopolitical violence that (re)produces autism’s racializing assemblages, this concluding section gestures toward a re-assembling of autism knowledge.

Extending evidence of disproportionality and delayed diagnoses for children of colour as well as the above elaboration of autistic Whiteness, Yergeau (2017) notes: “Temple Grandin’s routine proclamations that autism teachers should emulate the social practice of the 1950s is … a demonstrably racialized orientation toward the world (p. 5). Likewise, Brown (2017) argues that while autistic community activism has focused on closing the Judge Rotenberg Center (JRC),[1] activist literature rarely mentions that over 75% of its inhabitants are Black and Brown. These data provide a needed reminder that disability labelling schemes regularly result in students of colour being dis-located by carceral logics (Adams & Erevelles, 2014) and highlight how the JRC functions as an Agambenian state of exception whose violent anti-Blackness is unfortunately all too mundane (Adams & Erevelles, 2017). The correlation that exists, for instance, between the ED disability label and the school-to-prison nexus correlates with the imprisonment of autistics of colour Neli Latson (in 2010) and Matthew Rushin (in 2019), as well as the murder of Elijah McClain (in 2019). Indeed, All the Weight of Our Dreams foregrounds concerns about how autistics of colour will be treated by police. Hannon (2016) reinforces this concern in describing how her autistic son’s Blackness materially matters because it effects whether teachers will allow him to use his coping skills and determines the amount of recovery time (if any) schooling affords him. Blackness is ultimately the factor marking the difference between Hannon’s son simply being misunderstood or perceived as a threat by his White teachers and classmates. Similar examples to Hannon’s lived experience of being a Black mother raising a gifted Black son with autism can also be found in the work of #AutisticWhileBlack, as well as the advocacy and educational efforts of the Autistic Women & Nonbinary Network and the Color of Autism Foundation.

These examples support Weheliye’s (2014) argument that racializing assemblages require “the baring of nonwhite subjects from the category of the human as it is performed in the modern west” (p. 3). Artiles (2011) extends this thesis in describing how educational policies like No Child Left Behind (2002) (re)produce a “race-ability binomial ontology” (p. 435), which both strategically parses race and dis/ability and pursues their categorical alignment through practices of ontological erasure (for example, all students should have equal educational outcomes) that purposefully employ dis/ability to surveil students of colour with the intent of seeing them labelled, incarcerated, or spirit-murdered. While Yergeau’s (2013) above syllogism shows how explanations of autism (re)produce a hierarchical ordering of humanity, Weheliye holds in critical attention how Blackness delimits which children can inhabit what diagnostic spaces. Rather than being ironic, then, Groomer’s above recollection all too mundanely reveals how autism’s racializing assemblages delineate which diagnostic spaces children of colour can inhabit, as well as habituating the shape of these same spaces. Weheliye provides a much-needed reminder that autism knowledge must be understood in ontological relation to other diagnostic classifications, given the potential of disability labels like ED and ID to capture children of colour on sight even before dehumanizing psychological theories like ToM can be applied. This reminder helps toward understanding how such theories are predicated on non-White subjects already being barred from the ontological category, the human. Çelik (2017), for example, recounts how growing up queer and Turkish within educational ecologies that essentialize “difference” from the vantage of the white, cismale individual not only meant being excluded from autism knowledge but also resulted in a violently mutilated self-image such that they “felt inherently different to the ‘human’ altogether” (p. 355). Reinforcing how not even the condemned psychological label “refrigerator mother” was available to Groomer, Çelik’s recollection is a reminder of “a political violence that acts hierarchically to position certain bodies as always harmful, always dangerous, always flesh” (Roscigno, 2019, p. 12).

There is little accounting for this relational totality in Hacking’s (2007) historical ontology, suggesting that Hacking’s (2010) classification-in-motion is sociopolitically constrained by the habituated cultural and clinical Whiteness of autism’s racializing assemblages. To this point, Artiles (2011, 2013) shows how the very being of students of colour can morph depending upon how educational policies and practices understand identities that exist along intersections of race and dis/ability. While Hacking (2009a, 2009b, 2009c) can account for motions in classificatory knowledge about autism, such accounts remain individual (e.g., “That’s me!”) and in name only because they leave uninterrogated how this knowledge has been racially assembled via a race–ability binomial ontology that can materially transform how students of colour exist within educational ecologies, including precluding their recognizability as “human” (Çelik, 2017). Returning to Hacking’s (2009b) opening observation about participating in a living experiment in concept formation, it is possible that counter-narratives like neurodiversity and the life-writings of autistics of colour collected in All the Weight of Our Dreams simply reflect the workings of dynamic nominalism. Following Weheliye (2014), however, this conclusion leaves unasked questions about why autism diagnoses do not adhere to children of colour who have already been made up upon sight by racializing assemblages that pinion Blackness to emotional or intellectual capabilities through circular logics predicated on individualizing forms of diagnostic personal responsibility, or how the authors collected in All the Weight of Our Dreams express a shared desire to collectively displace (i.e., “That’s us!”) rather than to individually inhabit extant ways of being autistic.

Telling a different truth about autism means starting from a place that does not use autistics of colour as analogical limit cases for defining autism knowledge, but instead begins with Blackness as constitutive of neurodiversity such that to be pro-neurodiversity is to be anti-racist. Perhaps, then, more critical attention is needed toward understanding the hierarchical ordering of humanity, within which autism is implicated, than to the nominal making up of human kinds. Racializing assemblages, rather than looping effects, more helpfully conceptualize present motions of autism knowledge, given the potential of the former to recognize (even if contrapuntally and diagonally) the approximate proximity of neurodiversity and Black life (Manning, 2019, 2020). To conclude, then, by beginning again differently, Gordon (2020) not only helps to redefine what is autism and counter-narrate who is autistic but also helps to re-assemble autism knowledge: “I get the pleasure of simultaneously celebrating the spoils of black culture and autpunk (or you can also call it autistic culture); I like to juxtapose the two cultures every chance I get.” The pleasure Gordon is describing is #BlackAutisticJoy:

Whether I show off my Autism Acceptance logo based on the outline of my motherland Africa, display my cooking abilities, narrate scenes from movies on TikTok, shake my locs when I hear crunk music, or record myself playing NHL 94 on my PC, I carry both identities with pride. I am not afraid to show the best of both worlds; no wonder why some of my peers view me as strange or weird. I’ll take that!

para. 14

Gordon’s (ibid.) way of being autistic and Black provides an assets-based re-assembling of Groomer’s above recollection while also illustrating how there is more to understanding the nexus of autism and race than providing evidence of disproportionality and delayed diagnoses for children of colour and elaborating autistic Whiteness – not the least of which is that such racializing assemblages (re)centre Whiteness and tend to terminate by recounting sociopolitical violence against autistics of colour, a comparative move that posits “acting white” (Ogbu, 2004) while autistic as necessary to the survival of Black and Brown neurodiverse life. The hope in Gordon’s re-assembling of autism knowledge is how it relationally re-maps humanity’s ontological totality beyond evolutionary hierarchies that territorialize humanness into diagnostic habit spaces and then code inhabitants as naturally belonging to the definitions these same spaces narrate. There is a mutual inclusiveness to Gordon’s life-writing that refuses to be strategically parsed or categorically aligned because it is always already a being of relation. Instead, “black sociality,” imagined here through #BlackAutisticJoy, “invents sites of collective expression rather than inhabiting them” (Manning, 2020, p. 6; original emphasis). Questioning autism’s racializing assemblages helps to de-naturalize and de-territorialize the sociopolitical violence that not only governs autism knowledge but also hierarchically orders who is human. Gordon’s celebratory fusing of being Black and autistic is a needed rejoinder to Groomer’s ontological erasure as a knowing subject that also stresses autism’s individualizing classificatory motion, pulling it out of shape and carrying it collectively forward toward re-assemblages that value autistic existence differently.