The idea of autism—as a metaphor, a neurological disorder, a mental state, an increasingly prevalent diagnostic category, or a species of neurological difference—has an almost ready-made appeal for social scientists, concerned as they are with questions of human identity, difference, perception and subjectivity within a social and cultural context. This interest can work both ways: Dawn Prince-Hughes (2004), an anthropologist with autism, has explained how she learned about human emotional reciprocity through her work with primates. Temple Grandin, another woman with autism who is best known for her memoirs (1986, 1995) and her work designing livestock handling facilities, famously remarked that she has often felt like an ‘anthropologist on Mars’ among typically-developing people.Footnote 1

Autism is defined as a neurodevelopmental disorder of communication, behavior and social interaction, with onset before the age of three (APA, 2000). Although it was first described in the 1940s, diagnoses of syndromes on what is now known as the ‘autism spectrum’ have risen dramatically since the 1990s. This is at least partially the result of improved awareness among parents, pediatricians and educators, and of a broadening of the diagnostic category to include disorders like Asperger Syndrome and Pervasive Developmental Disorder-Not Otherwise Specified (Gernsbacher et al., 2005). In this sense, autism could be seen as an exemplary case of ‘making up people’ (Hacking, 1999), in which a previously obscure disorder was defined in broad terms by medical authorities just as parents learned to see their children's unusual behaviors as potential signs of autism (Hacking, 2006). If changes in social and medical practices can explain the increase, the possibility that environmental changes also play a role remains open (Rutter, 2005). With the increased visibility have come public controversies regarding the causes of autism, appropriate treatment strategies and the desirability of aggressive treatment as opposed to social policies focused on accommodation and acceptance.

Social scientists have begun to turn their attention to the autism spectrum disorders in earnest as a means to discuss identity politics, subjectivity and interpersonal relationships, and a great deal of suggestive work has emerged as a result of this interest. In this commentary, I survey the existing work on the social studies of the autism spectrum disorders by characterizing some central themes. They include the idea of autistic culture and the role of autism in different cultural contexts, the use of treatment strategies as starting points for theories about the causes and mechanisms of autism, the ways that autism has been used to explore different forms of subjectivity, and the importance of sociological and anthropological studies of autism researchers. I also look at studies of social movements and advocacy groups concerned with autism. In the process, I suggest some ways that this research might be broadened, deepened, brought to bear on new issues or repositioned in such a way as to enable researchers to address new questions. I think that it is particularly important to ask how both the work of clinical and basic researchers studying autism and the work of social scientists has been influenced by the preoccupations of their particular disciplines and the research traditions that have preceded and in some cases informed them. Finally, I return to some reasons that studies of autism and neurodevelopmental disorders in general may be productive subjects for the social studies of the life sciences and biomedicine.

Autistic cultures

Perhaps the most significant focus of anthropological studies of autism has been inspired by the advocacy of individuals with diagnoses on the autism spectrum. Jim Sinclair (1993), Martijn Dekker (2000) and Amanda Baggs (2007), among many other self-advocates, have argued for the validity of autistic experience from the perspective of the neurodiversity movement. They believe that autism represents one acceptable cognitive style or mode of interaction among many. Inspired by the disability rights movement and specifically Deaf advocacy (see Dekker, 2000), self-advocates propose that autism is a valid identity that ought to be protected against the encroachment of aggressive behavioral therapies (Baker, 2006: 27; Dawson, 2004) or prenatal genetic tests that might be developed in the future. Self-advocates tend to invoke a social model of disability, in which disabilities largely result from a lack of social accommodations to human difference, as opposed to a medical model that conceives of disabilities primarily in terms of the physical and cognitive disorders of individuals (Chamak, 2008). In 2007, a video posted to YouTube, the video sharing site, caused a minor sensation that resulted in a TV appearance for its director (and subject), Amanda Baggs. The video, titled ‘In my language’, showed a young woman holding her hands under running water, rocking, rubbing her face against an open book and engaging in other typically ‘autistic’ behaviors while humming a repeating sequence of notes. In the second half of the video, Baggs used a voice synthesizer to explain to viewers that the video is in her ‘native language’, but that her language is ‘not about designing words or even visual symbols for people to interpret. It is about being in a constant conversation with every aspect of my environment’ (Baggs, 2007). Viewers responded positively to the idea of autism as a foreign culture that only seemed deficient to those who refused to learn its norms.

Self-advocates have also argued that basic medical and psychological claims about autism represent unfounded assumptions on the part of researchers. These include the theory that autism represents a primary deficit of empathizing against a background of cognitive strengths in systemizing (e.g. Baron-Cohen, 2003), the conventional wisdom that the majority of people with ‘low-functioning’ autism have mental retardation, and the very idea of separating the autism spectrum into categories of ‘low-functioning’ and ‘high-functioning’. In some cases, self-advocates have been joined by parents who have come to resent the standard portrayal of autism as a growing epidemic and ‘tragedy’ with few redeeming aspects for diagnosed individuals or their families (Gernsbacher et al., 2005). In online communities of people on the autism spectrum, topics of conversation include the validity of professional expertise against shared experiences of living with a diagnosis, and the validity and value of people with autism defined in contrast, often, to ‘neurotypicals’ (Brownlow and O'Dell, 2006). Similarly, the most common metaphors that have been used to describe autism in both popular and medical discourses, including military metaphors of ‘fortresses’ and ‘sieges’, as well as allusions to emptiness, aloneness and isolation, reinforce dominant images that are inaccurate from the perspective of people with autism (Nickrenz, 2007; Waltz, 2003, 2005).

By emphasizing both the validity and distinctiveness of autistic experience, self-advocates have encouraged both anthropologists and journalists to explore the idea of autism as a culture, a project that has its precedents in classic studies of Deaf culture (e.g. Cohen, 1995). Scholars and activists alike are interested in the prospect of challenging standards of normality and deviance, and also in considering the ways that particular disorders can rapidly become culturally salient through the influence of advocacy and increased awareness, as in Lawrence Osborne's (2002) journalistic account of Asperger Syndrome subcultures, or Nancy Bagatell's ethnographic case study of one young man's efforts to define his identity as a person with Asperger Syndrome through the information that he gained from support group meetings and Internet sites (Bagatell, 2007).

Although the limitations of the medical discourse surrounding autism may seem most in need of critique in view of the concerns of the people that it purports to describe, it is also important to ask how signifiers of autistic identity are themselves shaped and transmitted. We should not assume that everything that people with autism say is a simple reflection of the disorder. Mark Haddon's 2003 bestseller The curious incident of the dog in the night-time, although it never mentioned a specific diagnosis, was widely hailed as a faithful depiction of Asperger Syndrome because of its deliberate presentation of the protagonist's symptoms. Contemporary readers easily interpreted these symptoms as indications of the character's underlying diagnosis. Social scientists should think about how the biosocial construct of ‘autistic culture’ combines externally produced medical definitions of autism, ideas about autism that are produced by people with the diagnosis, descriptions of autism in publications (websites, blogs, handbooks and self-help manuals, scores of which are available through publishers like Jessica Kingsley) and representations in popular culture. If autism is indeed a culture, we should ask how it is constructed, transmitted and maintained, what its local manifestations are and what relationship it bears to biological facts about autism as presented in the medical literature or understood by members of the community.

Approaches that take the culture of people with autism as their object of study are relatively recent, but important work has been done since at least the 1990s using social science methodologies to focus on the experience of the parents of children with autism and on the adequacy of services for people on the autism spectrum. These studies tend to reflect the opinion of the majority of clinical researchers (and of parents of children with autism) that autism is a severe disorder rather than a benign form of difference, and that treatments and social services are often woefully inadequate. These studies also reflect the truism that disabilities are experienced by both individuals and their families. All members of a family are affected and most adapt in response to the disability and to an often-hostile society that makes few provisions for full-time caregivers. Families often experience social isolation and stigma, are criticized by observers who mistakenly attribute a child's unusual behaviors to poor parenting (Gray, 2002), and see themselves as ‘voiceless’ in many of their encounters with medical professionals (Feinstein, 2006). The families of children with autism seem almost inevitably to take on traditional gender roles, as mothers assume the bulk of caregiving work (Gray, 2003). At the same time, parents have also proven themselves to be resourceful navigators of service systems and inhospitable communities. For instance, Michal Shaked (2005) has shown how parents in ultraorthodox communities in Israel work to justify the inclusion of their children in social life by making claims for their competence in culturally significant religious domains. Parents employ a wide range of strategies for coping with the experience of having a child with autism (some of which are common to the parents of children with other disabilities, e.g. Landsman, 2005) including a range of narrative strategies for conceptualizing their child's autism and their relation to it (Gray, 2001).

Competencies and how to define them have also occupied researchers. An insightful 1996 study by Ruble and Dalrymple critiqued the ideal of independence in outcome measures for autism treatment. Individuals with autism whose life histories suggested the most resilience and best overall quality of life were often those who, rather than being independent, benefited from accommodations in their surroundings and a complex network of social and family supports. Likewise, Billington (2006) argued for a ‘partnership model’ for evaluating the success of supports and therapeutic measures. This would take into account the perceptions of people with autism themselves, including their own assessment of their strengths and weaknesses. Morton Gernsbacher (2006) has argued that parents also need to be taught to view their children not as defective but as different, allowing them to engage in a reciprocal relationship of understanding and accommodation, while Townson and collaborators (2007), using a model of participatory research, worked with a group of young adults with autism diagnoses to evaluate the quality of services from the perspective of those on whom the services were focused. Ochs et al. (2001) found that when children with high-functioning autism were included in mainstream educational settings, their negative reactions to mistreatment at the hands of classmates indicated a capacity for emotional responses that has not always been attributed to them or figured into programs designed to promote inclusion. All of these studies, although many originated from more clinically oriented perspectives, used ethnographic, case study or other qualitative research methods to argue for the importance of the perspective of people with autism in designing treatments and services. They also point to the value of an open-minded attitude toward the needs of people with autism that allows for the possibility that conventional medical definitions may not always be the most useful ones where the concern is to create a welcoming and safe environment.

A complementary perspective focuses on the experience of autism across cultures. Tamara Daley (2002a) in particular has argued for the importance of cross-cultural research on the autism spectrum disorders. The tendency of some professionals has been to assume that autism, because of its basis in biological difference, must look the same everywhere. In reality, comparative, cross-cultural studies have the potential to demonstrate differences in the manifestation of the supposedly invariable symptoms of autism (Daley, 2002a). This has implications not only for our understanding of the boundaries between the ‘biological’ and the ‘cultural’, but also for the design of services and accommodations. Her studies, based in India, of the trajectory from parental recognition of symptoms to receipt of an official diagnosis of autism, as well as of the ways that different types of specialists understand and deploy diagnostic criteria, provide effective justification for approaches that look at variety in clinical practices both within and across cultures (Daley, 2002b, 2004). Similarly, although there has been more than one study that has retrospectively diagnosed cases of autism in the eighteenth and nineteenth centuries, at least two studies focus on very different aspects of those earlier cases. One study argues for the consistency of the behavioral symptoms of autism across time (Houston and Frith, 2000) while another focuses on the contrasts between a nineteenth-century doctor's understanding of autism and contemporary practices, emphasizing the ways that the doctor noted disrupted physiological processes in the three possible cases of autism that he saw (Waltz and Shattock, 2004).

Probably the most significant study to date of autism across cultures is Richard Roy Grinker's Unstrange minds (2008). The book is a beautifully written combination of ethnography and memoir of the author's experiences raising his daughter, who is on the autism spectrum, and the experience of families living with autism in Korea, South Africa and India. Grinker's explicit aim is to examine the emergence of the idea of an autism ‘epidemic’, a claim that has been persistently made by parent organizations. Many self-advocates (as well as Grinker himself) believe that this view underestimates the significance of broadened diagnostic categories and better public awareness, and devalues the unique attributes of autistic minds. Autism may not be a ‘culture-bound syndrome’ that has emerged as a product of only one location or manifests in markedly different ways in different places, but the ways in which people with autism are understood and integrated into communities can differ radically. Indeed, Grinker finds that the families of children with autism in these very different cultural contexts have embraced the biomedical definition of autism rather than the culturally situated accounts of their relatives, who may attribute a child's autism to possession, for instance.Footnote 2 These families experience many of the same struggles for acceptance and services that are faced by parents living in the United States.

If an important task for scholars investigating communities of people with autism is to remember to ask questions about the specificity of autism as a culture, another significant focus should be the way that communities of people with autism and their families are part of broader networks, both national and transnational. The diagnostic instruments used to classify people with autism, the services that are currently provided for them and the measurements of the efficacy of those services emerge from institutional and professional contexts that, it often seems, have only imperfectly taken into account the needs and priorities of people with autism and their families. Social scientists thus potentially have a great deal to contribute to clinical and policy discussions if they can produce work that is accessible to psychologists and other professionals.

Theorizing treatment

Majia Nadesan (2005), in her comprehensive history of the conceptual frameworks that have been used to explain autism since it was first described, argues convincingly for the close relationship between trends in the fields of psychology and education and changes in the ways that autism has been interpreted and treated over time. In a session at the 2007 meeting of the American Anthropological Association, four of the panelists centered their discussion on a treatment method, broadly defined. One author structured her paper around an account of a session of sensory integration therapy (Park, 2007). A second presenter described the complex interplay between a child with autism and an experienced therapist using the Rapid Prompting Method developed by Soma Mukhopadhyay (Engelke and Mangano, 2007), another explored the dynamics of animal-assisted therapy (Solomon, 2007), and yet another presenter, while she did not document a particular treatment method, made it clear that her choice to home-school her son was a response to the very real possibility that if he remained in the school system, he had a good chance of acquiring a diagnosis and eligibility for more formalized methods of treatment (Prince-Hughes, 2007).Footnote 3 Instead, she provided her son with learning opportunities that were organized around his interests, which, if it was not a form of treatment, was at least a practice developed in response to a particular cognitive style that the presenter shared with her son. In each case, the effectiveness of the treatment practice in question led the author of the paper to develop, elaborate upon or modify a theory about autism, subjectivity or cognition more generally.

Noticing the focus on treatment led me to think about the way that this development, from observations about treatment efficacy to theories about the underlying disorder, has worked in the history of autism in particular and the history of medicine and psychiatry more generally. Among historians of medicine it is now commonly understood that the perceived efficacy of therapies changes over time. Therapeutic encounters are often intimate exchanges that can affect the beliefs of both patients and doctors. Bruno Bettelheim is known for promoting the idea, now mostly dismissed, that psychotherapy had the ability to help children with autism, as well as the theory that autism was caused by cold, affectively distant mothers. However, Bettelheim wasn't alone in his convictions. Counselors at the Orthogenic School also believed that they were helping the children with autism diagnoses that they worked with (Sanders, 1996). Bettelheim's book, The empty fortress (1967) received laudatory reviews in leading US magazines and newspapers, but decades later Bettelheim has come to stand for failed theories of the etiology of autism.

During the years that Bruno Bettelheim had funding from the Ford Foundation to investigate the viability of milieu therapy as a treatment for autism, the Orthogenic School hosted an anthropologist-in-residence, Jules Henry. Henry observed counselors at work with the students in the residential treatment program, and he also observed the families who had placed their children in the school. In Pathways to madness (1971), Henry contended that he could understand the family environment that had fostered the development of mental illness in children, even without observing the parents’ direct interactions with them. He thought that the pathogenic tendencies were evident in a parent's tone with their typically-developing children, or with their spouse, or in the details of their daily lives.

In the 1970s and 1980s, Martha Welch's Holding Therapy, a method that consisted of intensive and often confrontational sessions of enforced eye contact while parents cradled their affected child in their arms, was popular (Welch, 1988). Catherine Maurice, a mother and memoirist, provides a moving account of the treatment's promise and her eventual frustrations with it (Maurice, 1993). Welch enjoyed a great deal of public and professional support—she was featured in a BBC special—and Niko and Elisabeth Tinbergen enthusiastically promoted her methods in their book, suggesting that the technique closely followed the implications of their theory of autism, which drew on the insights of ethology (Tinbergen and Tinbergen, 1983). Ivar Lovaas, who was instrumental in carrying out studies of and arguing for the utility of Applied Behavior Analysis, proposed a behavioral theory of autism based at least in part on the apparent success of the interventions that he had developed (Lovaas and Smith, 1989). Simon Baron-Cohen and colleagues have based a proprietary intervention on the theory that a salient feature of autism is an inability to empathize. Their technique uses a computer program to teach people with autism to recognize facial expressions and tones of voice (Golan and Baron-Cohen, 2006). Clinical researchers, meanwhile, have expressed some bewilderment that contemporary parents of children with autism are still willing to swear by Secretin, a pancreatic hormone that in several clinical trials did not demonstrate efficacy for treating symptoms of autism (Esch and Carr, 2004).

Facilitated Communication (FC), a method that involves a helper supporting the arm of a person with autism as they type on a keyboard, has been a controversial method in the care of people with impaired communication. It was developed and initially advocated as a technique for use with people with cerebral palsy and later was attempted with individuals with autism, some of whom developed the ability to type independently (Biklen, 1990). By the mid-1990s, the American Psychological Association had issued a statement discouraging the use of FC amid growing evidence that facilitators were unconsciously prompting answers in FC sessions (APA, 1994; Mostert, 2001). There were also cases in which individuals who were being facilitated produced accusations of abuse against caregivers that were not substantiated (Howlin and Jones, 1996; Konstantareas, 1998). Despite this, the technique continues to be popular among some parents and educators. Ralph Savarese (2007) based much of his recent memoir on the typed communications of his adopted son, and Douglas Biklen's (2005) edited collection of facilitated writing uses the texts to argue for a radically different interpretation of the experience of autism. Sue Rubin, in the documentary Autism is a world (Wurzburg, 2005), is shown using a modified form of facilitated communication with verbal prompting—she is mostly nonverbal—and in one scene she interacts with an expert in the field who is visibly impressed at the ways that Rubin's ability to communicate confounds her own expectations about the mental capacities of many people with autism.

Bilu and Goodman (1997) have described the use of facilitated communication in a Jewish ultraorthodox community in Israel, explaining how the community has to a certain extent come to regard the typed messages of children with autism as oracular pronouncements. Children using FC are consulted on matters concerning metaphysical questions, religious practice of faith and the reasons for their disability (often explained as punishment for sins committed in a previous life) (Bilu and Goodman, 1997). Although it may be easy to read this faith in FC as so firmly situated in the experience of the ultraorthodox community as to be irrelevant to other contexts, contemporary Americans are also often impressed by the ways that FC-mediated communications appear to reveal insights on the part of people with autism that are stunning, if not psychic.

Controversies over therapies or radical changes in therapeutic norms have not only occurred with autism—during the 1950s and 1960s, doctors tried to treat schizophrenia with psychoanalysis and later family therapy, while now it is managed primarily with antipsychotic medications. My point in providing this array of therapeutic trends and associated theoretical frameworks isn't that therapies are bound to change as we gain new insights and information about the causes of a disorder, nor that every promising new therapy is destined to seem naïve in time. My point is that, based on the effectiveness of therapy, doctors have often been tempted to construct new theories about the mechanisms underlying a disorder. They do so at times without asking the question of who responds and who doesn't, or whether it is the experience of administering the therapy itself that accounts in part for the perception that it works. Given the history of fairly radical change in approved therapies for autism, and the fact that the population that is defined as having autism is so heterogeneous, social scientists should probably hesitate to generate theories from their observations of therapeutic encounters, or at least should be thoughtful about the stakes involved when they choose to do so.

I also think that it is worth asking what happens when we consider the experience of something working, of efficacy, as worthy of anthropological or social investigation and explanation in its own right. As social scientists, we might think about technologies and diagnoses as functioning synchronously to produce results that are difficult to understand outside of the historical and social context in which they occur. We should ask questions about how particular therapies emerge as preferred options for particular contexts, how techniques are designed, and what types of acquired expertise are required to successfully develop and administer therapies (e.g. Engelke and Mangano, 2007; Silverman and Brosco, 2007), even as we look at the new perspectives on autism that are made possible by particular interventions.

Autism and subjectivity

Most experts take for granted that autistic disorders occur along a spectrum of functionality or severity. Heterogeneity in the population of people diagnosed implies a number of things. Self-advocates have raised important points about the appropriateness of defining autism along a single spectrum when individuals can fit descriptions of high- or low-functioning at different points in their lives or in different settings, and when given individuals may have specific tasks or areas of cognitive or social ability in which their skills are at very different levels. A single person may have real problems with verbal communication but have required relative ease in social interactions. There may also be an underlying physiological heterogeneity in autism—at the genetic level or even the level of brain structure or metabolism (Herbert, 2005). It is entirely possible that when we talk about autism we may be talking about a complex syndrome where what we call cognitive differences are actually cognitive responses to differences, or differences in combination with environmental stressors, that affect entire physiologies but result in similar cognitive outcomes or behavioral responses. It is beginning to seem unlikely that there is only one way for a person to develop symptoms that are consistent with an autism diagnosis (Happe et al., 2006).

Self-advocates have themselves promoted the notion that people with autism occupy a social and sensory sphere that is so foreign to ‘neurotypical’ experience as to be nearly extraterrestrial. Jim Sinclair (1993) at one point called people with autism ‘aliens’, and a 2003 anthology of writing by women with autism is titled Women from another planet. Many self-advocates prefer to eschew the person-first language advocated by the disability rights movement (i.e. ‘a person with autism’) for labels like ‘autistic’ or ‘Aspie’ in order to emphasize the idea that their autism is inextricable from their identity.

Many researchers now talk about autism in terms of a kind of meaningful distinctness of experience, cognition, interaction and perhaps also subjectivity. For this reason, it may be worthwhile to ask ourselves what we think—theoretically and functionally—of difference. Is what Dawn Prince-Hughes (2007) has referred to as ‘connectedness’ to the outside world something that we want to take as different in kind from a more typical awareness of self and others, or is it an experience from which we want to infer rules about people in general? The question concerns the degree to which social scientists see it as important to take into account both those processes and modes of interaction seen as universal and those seen as particular or unique to people with autism, and how we distinguish between the two in our work. Studies of autism that proceed from the idea that the unique elements of autistic experience arise from the unique biology or cognitive attributes of people with autism run the risk of producing an essentializing discourse. At the extreme, it threatens to treat people with autism as members of a distinct species of human to which our obligations and responsibilities might be different. We need to be careful to pay attention to the ways that ideas about the biology of autism are reproduced and reinforced through medical literature, clinical and self-help handbooks, and even support groups.

Research from the social sciences can counter essentializing discourses by showing how the behavior of people with autism is in fact less different from their typically-developing peers than has been assumed, or how people with autism are also participants in social milieus that play a role in constructing (and defining) the strengths and weaknesses associated with autism (Ochs et al., 2004). This approach sees the symptoms that characterize autism as becoming visible, in part, as a result of socio-culturally based behavioral norms, especially those that demand intuitive comprehension of social expectations—with implications for our understanding of the ‘practical logic’ important in navigating social life (Ochs and Solomon, 2004; Ochs et al., 2004: 148). It also highlights the fact that people with autism, their families and other experts can play an active role in educating others on how to create a more tolerant environment (Ochs et al., 2001). A special issue of Discourse Studies edited by Ochs and Solomon in 2004 brought together the work of researchers using methods from linguistic anthropology to study the ways that people with Asperger Syndrome and high-functioning autism participate in conversational interactions. Contributors emphasized that conversational competence and impairments in autism are significantly more complex than previously recognized, and that the conversational practices of people with autism may violate commonly accepted ideas about their level of empathy and sensitivity to social nuances (Kremer-Sadlik, 2004; Sirota, 2004; Solomon, 2004).

Recent work on psychological testing suggests, likewise, that estimates of the level of mental retardation in the population diagnosed with autism may be a reflection of the testing instruments, which demand performance in areas that are most affected by autism and downplay those domains that reflect autistic strengths (Dawson et al., 2007). People with autism may perform exceptionally well on tests for which the lists of necessary skills ‘read like compendia of the cognitive processes that autistics are assumed to lack’ (Dawson et al. 2007: 658). In effect, work in this area presents a social analysis of the assumptions that are built into intelligence tests by professionals, who tend to see strong performances in particular domains by people with cognitive disabilities as incidental effects of systemic deficits. Similarly, people with autism may characterize their diagnosis in terms of symptoms (for instance perceptual differences and emotional sensitivity) that are unrepresented in standard psychiatric descriptions of the disorder (Chamak et al., 2008). If, in addition, people diagnosed with autism tend to see few relevant distinctions between categories of ‘low-functioning’ and ‘high-functioning’, social scientists need to find ways of conducting research with people with disabling behavioral symptoms and few or no verbal communication abilities, in addition to people with strong verbal language skills who may be easier to work with. In practice, this might involve close observation of interactions with caregivers and support staff.

In considering autism and subjectivity, social scientists are and have good reason to be drawn to the theory-of-mind hypothesis in autism, first proposed by Baron-Cohen et al. in 1985 and expanded upon in a number of publications, including Simon Baron-Cohen's Mindblindness: An essay on autism and theory of mind (1995), in addition to a notable collection, Understanding other minds: Perspectives from autism (Baron-Cohen et al., 1993). Although other cognitive psychologists have disputed the hypothesis that people with autism have a primary deficit in their capacity to attribute mental states to others (and, indeed, that the brain has a modular organization that would lend itself to such localized and specific deficits), and psychologists have challenged the finding that people with autism in general lack a functioning theory of mind, the hypothesis continues to be referenced in writing in the social sciences and humanities (Belmonte, 2008).Footnote 4

A theory that proposes that people with autism have an innate inability to attribute internal mental states to others suggests that autism provides an unprecedented opportunity to study alternate forms of subjectivity. These various subjectivities might be studied as they emerge in the narrative writing of people with autism, for instance Jurecic (2007: 425–427), or by looking at how interspecies relationships can mediate intersubjectivity between those with autism and other human interlocutors (Solomon, 2007). Social scientists are understandably tempted to draw on those theories that have particular resonances with the traditional preoccupations of their disciplines, and most use them quite responsibly, with caveats pointing to concerns from both the scientific community and the self-advocacy community (e.g. Jurecic, 2007: 426). However, an equally valid focus (drawing on science studies) may be the anthropology of scientific research on the autism spectrum disorders. This approach should not be limited to exploring perspectives from cognitive psychology, but should look critically and in detail at how autism is constructed within the fields of genetics, neuroscience, linguistics and immunology, as well as by practitioners working in the fields of developmental pediatrics, occupational and behavioral therapies, complementary and alternative medicine and neurology.

Social movements

The autism spectrum disorders have been an important locus of organization and activism. Beginning in the 1960s in the US and the UK, parents organized to share information and lobby for improved services for their children, including behavioral therapy and administrative reclassification from categories like ‘Severely Emotionally Disturbed’ to biologically based categories like developmental disabilities. More recently, autism researchers, the media and parent advocates have converged in identifying autism as an object of urgent concern not just for affected families but for the public at large, although the way that these different groups identify priorities may be startlingly divergent (Singh et al., 2007). Although the activism preceded (or at least coevolved with) any concept of autism as a specifically genetic disorder, it still seems natural to write about the phenomenon in terms of concepts like biosociality or biological (or genetic) citizenship (Rabinow, 1996; Rapp and Ginsburg, 2001; Rose and Novas, 2003).

The most interesting aspect of biosocial relations around autism is the sheer variety of ways in which biological information is used as a means to articulate group membership. Organizations like the US-based Autism Speaks are wealthy enough to have a significant impact on the direction of research through their choice of funding targets. Parents of children with autism have worked together to shape research on the genetics of autism (Silverman, 2008), but they have also come together around interpretations of autism as a systemic medical disorder that may respond to biomedical interventions. Pru Hobson-West (2007), Sharon Kaufman (2007) and Laura Senier (2008) have written about parent activism centered on the possibility that autism is caused by vaccines as a kind of risk management or ‘worst-case thinking’ (Senier, 2008), which has much to tell us about the importance of trust in parent assessments of medical advice and public understandings of scientific authority. Many of the scholars mentioned above who have looked at the emergence of ‘autistic culture’ have framed it as a type of identity-based social movement. Brigitte Chamak (2008) has wisely looked at the interplay between parent organizations, which have focused their activism on services and research funding, and self-advocacy groups.

The question for autism and, indeed, for many of the disorders for which biosociality or biological citizenship have appeared to be relevant terms is, first, to determine what specificity these types of organizations or movements have, and whether biological concepts and information play a formative role for these organizations. We need to ask whether groups formed around autism (or breast cancer, or depression, or cystic fibrosis) are different from other types of political or social organizations in the degree to which members refer to biological facts in the process of their legislative or cultural advocacy. Do their members engage in what Joe Dumit (1997) has called ‘objective self-fashioning’ when they refer to biological information such as images from brain scans and diagnostic checklists? How do parents make use of printouts from commercial laboratories that measure the quantities of heavy metals in their children's baby hair, or identify immune sensitivities to dietary substances, when they argue for the relevance of biomedical treatments?

We might also ask about the variety in the types of biological information that are put to use, and with what political intent. For instance, self-advocacy groups and parent groups make different use of genetic information. Some parents are inclined to trust epidemiological reports that find, for instance, that the rate of autism diagnoses continued to increase following the removal of thimerosal from vaccines in Canada (Fombonne et al., 2006), while others remain focused on experimental studies that demonstrate the potential toxicity of the ethylmercury-based vaccine preservative thimerosal (e.g. Goth et al., 2006; Hornig et al., 2004). Finally, we need to ask whether social movement models best serve our purposes for evaluating the strategy, composition and organizational trajectory of these groups. If group aims are narrowly focused on political or medical changes, and if they have a relatively poor track record of forming alliances with members of groups associated with other, possibly related disorders, perhaps these associations are better described as interest groups rather than as social movements (despite their preferences), but this is also an open question.

Conclusion

The recent upsurge in public interest in the autism spectrum disorders, and the outpouring of memoirs and literature written by people with autism, mean that scholars have good reason to pay attention. As with many neurological or psychiatric disorders, there is a tension between biologically centered explanatory modes and those that focus on the social construction of the disorder. Specialists in disability studies distinguish between the social model and the medical model of disabilities. The most successful studies are those that refuse to situate their claims firmly within one model or the other, but instead pay attention to the strategic uses of different models by various interested groups. More importantly, the distinction—whether we choose to think about autism as a devastating neurological condition or as a disorder that emerges only within the context of a medical and social system that makes autistic symptoms appear deviant—has a limited impact at best on some of the most crucial political questions surrounding autism. These questions relate to the way that societies deal with dependence and accommodation. As many of the researchers described above have argued, cognitive disabilities in general, and autism in particular, present challenges to traditional accounts of personhood, especially when defined in terms of the potential for attaining complete independence or for forming meaningful social relationships. The fact that some individuals will require significant support throughout their lives, with no reasonable expectation that they will eventually carry out productive work, has implications for theories of human rights and citizenship that have yet to be fully understood (Nussbaum, 2006).

Scholars working on autism or on other complex diseases and disabilities share another problem. This could be described as the tension between two academic perspectives. We can choose to describe the problems particular to autism, where the task is to better represent people with the disorder by critiquing the types and efficacy of services and treatments that are available, and advocating for cultural accommodation and acceptance. The other choice is to describe autism as one instance of cultural formation, advocacy, and biosociality around a particularly complex disease entity, thus drawing connections between autism and other contested and underserved illness categories. As social scientists we may be inclined toward the latter alternative because it presents the most opportunities for our work to relate to the work of others in our field. However, it is important to remember that people with autism in particular have been vulnerable to enrollment in the theoretical programs of ambitious researchers, often without being consulted, let alone given the option of collaboration or partnership in research. For this reason, paradoxical though it may seem in the case of a condition first characterized as featuring an ‘innate inability to form the usual, biologically provided affective contact with people’ (Kanner, 1943: 250), researchers need to recognize that ‘[i]n order to act ethically in the field of autism … it is necessary to hold on to fundamental human issues—of feeling, thinking and meaning’ (Billington, 2006: 10). In other words, researchers might find that friendship is a wise first step in any project that presumes to represent people with autism.