Recent representations of autism frequently include an assumption that autism is the result of a “theory of mind” deficit (i.e., an inability to understand others' mental states). This notion is examined using a social constructionist perspective. The belief that autism is a sort of “mind-blindness” has much in common with earlier representations of autism that depict it as a puzzle and, paradoxically, as a single entity defined by core characteristics. Theory of mind theorists also, like their predecessors, define autism as a form of insufficiency and as requiring fixing rather than accommodation. Alternative narratives about autistic minds that incorporate the perspectives of people labeled autistic are an important counterbalance to the limitations of such professional viewpoints.
Put the pieces together one way, and you end up with a normal child. Put them together another way, and you end up with a child with autism. (Nash & Bonesteel, 2002, p. 56)
For years, the logo of the Autism Society of America, formerly the National Society for Autistic Children, was the bowed head of a child superimposed on puzzle pieces. One piece does not fit. The puzzle piece logo was first introduced by the National Autistic Society (of England) in 1963 (Allison, 1988). Today, the use of puzzle imagery, depicting autism as mysterious or enigmatic, is nearly inescapable in the world of autism. With few exceptions, autism-related societies and organizations incorporate jigsaw puzzle shapes into logos and other artwork on their literature and websites. More often than not the image is not simply that of a puzzle, but of a puzzle with a piece missing. Sometimes the idea of a missing puzzle piece is expressed in terms that are quite stark, as in the following passage describing MRI images: “When you look at these images, you can see what's not there she [Dr. Nancy Minshew] says, conjuring up an experience eerily akin to looking at side-by-side photographs of Manhattan with and without the Twin Towers” (cited in Nash & Bonesteel, 2002, p. 52). Comparing the presence of autism to the absence of the World Trade Center interprets autism as devastating.
In its most extreme form, the puzzle metaphor is extended to the idea that the “real” child is absent. When describing an autistic child, adults commonly refer to a “real” (i.e., nonautistic) child that they “know” is “in there” and might be released if only one were to work hard enough at it. Clara Claiborne Park's portrait of her autistic daughter as a young child (1972) is perhaps the best known of several such accounts that use the image of “The Changeling” (her title for Chapter 1 of The Siege), which suggests that the autistic child is in some way a substitute for a “real” child who has been stolen away. (With regard to my use above of the words “autistic child,” I often make a choice not to use so-called “person first” language based, in part, on the arguments of Sinclair , who objects to such usage. Sinclair affirmed that autism is a positive and essential part of who he is. He prefers “autistic person,” which implies that autism is inseparable from the person, to “person with autism,” which he believes makes the autism sound either unimportant or negative. On the other hand, I also view categories such as autism, which assign labels in an attempt to describe human variety, as socially constructed. When I especially wish to emphasize the social construction of autism, I use “people with autism labels.”)
The label of autism is typically assigned to children who demonstrate pervasive differences in communication, social interaction, and sensorimotor responses to their environment. Autistic people often have intense and compelling interests that can dominate their behavior. In recent years, it has become common for professionals, media accounts, and to some extent parents, to explain these autistic characteristics as resulting from an inability to form something called a “theory of mind.” A theory of mind is a person's awareness and understanding that he or she, and other people, have thoughts, beliefs, desires, intentions, feelings—the full range of mental states. Current representations of autism are increasingly dominated by the notion that autistic people lack this awareness and understanding and, therefore, cannot “mindread,” or easily and automatically interpret the mental states of others, a talent that is presumed to come naturally to nonautistic people. I do not believe that autism can be adequately explained by the absence of a theory of mind and will argue that this notion is simply the latest manifestation of a missing puzzle piece metaphor that should have been discarded long ago.
My argument is based on three premises. First, I approach categories that people use to describe human differences (e.g., “autism”) from the standpoint of symbolic interactionism (Blumer, 1969), which holds that meaning in general, and in particular the meaning of such categories, is created through a process of social interaction and communication. Although such distinctions may have an empirical basis, what matters about them is what people decide matters; such categories are what are often term social constructs.
A second premise is that the ways we choose to describe and represent people with disabilities are never neutral (Duchan, 1998). Whether consciously made or not, our representational choices both result from and perpetuate dominant perspectives that dictate who is valued and how power and privilege are distributed.
My final premise is that social representations change over time, and it is this impermanence that allows us to evaluate them. When a particular perspective comes to dominate, it becomes normalized, which makes it especially difficult to regard critically. Normalization is a form of conceptual camouflage; most of our social constructs strike us as self-evident. Like fashions in clothing, social understandings do not appear odd until we look back at them over time. Change over time in the ways that human differences are understood creates both the opportunity and the obligation to reinterpret socially constructed knowledge.
My autistic son was born in 1978, shortly after a great shift in conceptions about autism had occurred. By the late 1970s, an assumption that autism is the result of specific neurological differences had largely replaced the earlier belief that autism was a form of psychosis, a child's drastic withdrawal from cold and unfeeling parents (Bettelheim, 1967). My wife and I were fortunate not to endure as much blame as did those parents from the 1960s, when it was so often taken for granted that “refrigerator parents” caused autism. However, like those earlier parents, our child has been continually explained to us in ways that do not match our experiences living day to day with him. The “theory of mind” explanation is the latest example of this.
A Brief History of the Theory of Mind Explanation of Autism
Autism as a Category
Kanner (1943/1973), a psychiatrist at Johns Hopkins University, published case studies of 11 children whom he had observed over a period of 5 years. He claimed that characteristics displayed by these children had enough in common that they represented a distinct syndrome, never identified before. Kanner offered the name “early infantile autism” for his syndrome, which eventually came to be known simply as “autism.” The word comes from the Greek autós, meaning self and refers to Kanner's perception that the children he observed were withdrawing in a dramatic way from social contact.
In 1944, one year after Kanner's paper, Asperger, an Austrian pediatrician, published a postdoctoral dissertation describing a group of 4 children very similar to Kanner's 11. Both Kanner and Asperger independently believed that they were identifying a new syndrome, and both chose to use the word autism to describe it. Although Kanner was also born in Austria, he and Asperger were unlikely to have seen each other's work during World War II. Many of our conceptions regarding autism have changed since 1943, but the descriptions of autistic children that Kanner and Asperger produced sound quite familiar today to those working with children so labeled.
Autism is usually defined as a behavioral syndrome; children are assigned the label based solely on their behavior. The contribution of Kanner and Asperger was to recognize that a seemingly disparate set of behavioral characteristics often appeared to cluster in certain children. These traits include unusual ways of interacting and communicating with others, sensorimotor differences, intense interests, and a compelling appreciation for rules and routines. Although individuals are identified as autistic on the basis of their behavior, it has long been assumed that autistic behavior has its origin in mental function. People do not simply “act autistic.” They “are autistic.” This is because the behaviors associated with autism are so pervasive that they cannot be separated from the person. Today, autism is widely believed to be the result of neurological differences, although the exact nature of these differences is yet to be defined. Despite much recent research into the biological basis of autism, there is still no blood test or genetic test that identifies it (Akshoomoff, 2000; Minshew, Johnson, & Luna, 2000).
Does the Autistic Child Have a “Theory of Mind?”
The first use of the term theory of mind in the sense that has become common today was in a study by Premack and Woodruff 1978), two primatologists at the University of Pennsylvania, entitled “Does the Chimpanzee Have a Theory of Mind?” In the following passage, they defined the term:
In this paper we speculate about the possibility that the chimpanzee may have a “theory of mind,” one not markedly different from our own. In saying that an individual has a theory of mind, we mean that the individual imputes mental states to himself and to others.… A system of inferences of this kind is properly viewed as a theory, first, because such states are not directly observable, and second, because the system can be used to make predictions, specifically about the behavior of other organisms. (p. 515)
While exploring the extent to which chimpanzees' thoughts may resemble those of human beings, Premack and Woodruff (1978) assumed that we (homo sapiens) do possess the intuitive understanding of others' mental states that they called a “theory of mind.” This trait is seen as adaptive because an organism that understands the intentions of another can make better predictions about the behavior of potential predators or prey.
Premack and Woodruff (1978) described an experiment in which they showed chimpanzees videotapes of humans in problem situations that the animals could presumably understand (e.g., trying to retrieve bananas that are placed above their reach). The animals were then shown a series of photographs, one of which depicted a possible solution to the problem (e.g., a moveable box that allowed the human to reach the bananas). The authors argued that the fact that the chimpanzees they tested tended to choose the best answer meant that they were able to take the perspective of the person in the video. One statement in their conclusion regarding study with other populations has proved prophetic: “Although here we have talked only about the chimpanzee. …[a]re at least some retarded children deficient in specifically this form of theory building? What is the developmental course of such theory building in the normal child?” (pp. 525–526).
The study of theory of mind in different populations has indeed absorbed a number of researchers since Premack and Woodruff (1978). As Gopnick (1997) observed, “In the last few years there has been an explosion of interest in children's ideas about the mind” (p. 3). Accordingly, a large body of research about the theory of mind has emerged. These investigated described a developmental process in which children create representations based on understanding others' mental states and do so with increasing sophistication as they mature. A number of researchers have found that children typically appear to cross a theory of mind threshold between the ages of 3 and 4; that is, they are newly able to demonstrate competence on theory of mind tasks created by researchers (see, for example, Perner, 1991, pp. 240–241).
Wimmer and Perner (1983) reasoned that simply predicting another's actions, as Premack and Woodruff's (1978) chimpanzees were asked to do, does not demonstrate understanding of mental states unless there is a discrepancy between reality and what the other person believes is true. This is because subjects could simply be making choices based on what they believe rather than on what the other person believes, which does not require understanding another individual's belief state. In order to better test for the presence of a theory of mind, they developed a “false belief task” in which subjects must show understanding that another person's behavior results from a prediction that is logical but false.
Baron-Cohen, Leslie, and Frith (1985) were the first to try this procedure with children labeled autistic. In their article, “Does the Autistic Child Have a ‘Theory of Mind?’” a title that echoes Premack and Woodruff (1978), they used a simplified version of the false belief task developed by Wimmer and Perner (1983) to create what has become nicknamed the “Sally–Anne experiment.” Using dolls named Sally and Anne, an experimenter showed a child a scenario that portrayed one of the two characters as likely to have a false belief. In the scenario, Sally put a marble away in a box and then left. While she was gone Anne transferred the marble from the box to a basket. Shortly after, Sally returned. The experimenter then asked the child who had been watching this, “Where will Sally look for the marble?”
Baron-Cohen and his colleagues (1985) compared three populations: children without disabilities, children with Down syndrome, and children labeled autistic. The children were selected so that, on average, the autistic children had higher IQs than did members of the other two groups. The striking result of the experiment was that most children in the first two groups stated that Sally would look in the box, but the autistic subjects were much more likely to choose the basket. Baron-Cohen and his colleagues concluded that the difficulty autistic children had answering the question correctly stems from an inability to grasp that Sally had a false belief because she did not see the marble transferred. The autistic children presumably could not take Sally's perspective because they were lacking a theory of mind.
Both in their 1985 article and in their subsequent prolific work, Baron-Cohen, Leslie, and Frith argued that the presumed absence of a theory of mind, which they believed was demonstrated by the Sally–Anne experiment, lies at the heart of autism. Differences related to interaction, communication, and intense areas of interest are all construed to be the result of a specific, neurologically based deficit that causes autistic people not to understand that others have perspectives. More recently, Baron-Cohen (1990) coined the term mindblindness, and he continued to claim that an inability to “mindread” is indeed the central feature, the sine qua non of autism. (Note that in using this term, Baron-Cohen is not talking about a literal ability to read the exact thoughts of others, but a talent for predicting behavior based on understanding others' mental states.)
Features of a “Missing Theory of Mind” Representation of Autism
What happens when we characterize people as lacking a theory of mind? Do autistic people ever benefit from such representations?
Although some conceptions about autism seem to have shifted dramatically over time, certain underlying assumptions have changed very little, as each new theory enjoys its heyday. Theory of mind theory shares at least the following eight characteristics with earlier representations of autism. (a) Its proponents view “autism” as a unitary phenomenon that results from a single, core disability. (b) Its adherents claim to have the professional authority to explain autism. (c) They assert a scientific basis for such claims. (d) Autism is portrayed as exotic and extreme. (e) Autism is represented not as difference but as deficiency (i.e., autistic people are considered damaged or inadequate versions of “normal” human beings). (e) The intelligence of autistic people is often regarded as questionable or entirely discounted. (f) Professionals and other “helpers” are depicted as endlessly patient while autistic people are rarely viewed as patient. (g) All of the above is normalized, such that it becomes a dominant perspective that is not questioned.
Autism is a Unitary Phenomenon, the Result of a Single, “Core” Disability
The notion that people are autistic because they lack an ability to understand others' minds is typically presented as a sort of grand unifying theory. It claims to be a parsimonious explanation for the social, communicative, and behavioral differences typical of people with autism labels:
In 1985, Uta Frith, Alan Leslie, and I proposed that three of the cardinal symptoms in autism—the abnormalities in social development, in communication development, and in pretend play—might be the results [sic] of a failure in the development of mindreading (Baron-Cohen, 1995, p. 63).
If we met someone who lacked an understanding of mind … we might expect abnormalities in socialization … An individual who lacked an understanding of mind would also be very poor at communication. An individual who lacked an understanding of mind would probably … display impoverished imaginative abilities … In sum … if we wish to know what an individual would be like who lacked an understanding of mind, then perhaps we need look no further than those unfortunate individuals who have autism. (Mitchell, 1996, pp. 165–166)
Baron-Cohen (2001), in particular, has championed the idea that “difficulty in understanding other minds is a core cognitive feature of autism spectrum conditions” (p. 169).
People with autism labels are not as quick to construct autism as a unitary phenomenon. They are more apt to describe their experience as individual and diverse:
I have a deficit in my thinking and I don't know if it is common in autism or is just unique to me. When I remember events, I can't tell if they really happened or I imagined them. (Rubin, 1998, p. 3)
These are the reasons why, to me, there is no single thing called ‘autism’. Some camels have a lot of one type of straw on their backs. Some camels have a whole collection of different types of straws on their backs. (Williams, 1996, p. 6)
Cognitive Psychologists are Uniquely Positioned to Understand and Explain Autism
Professional bids for authority are often involved in changing constructions of autism. Developers of successive models each challenge the authority of earlier representations in order to establish their legitimacy. This is most evident from an historical perspective.
As a psychiatrist Kanner constructed autism as a psychiatric disorder and, initially, the profession of psychiatry was seen as authoritative regarding the nature of autism. Beginning in the 1960s, however, behavioral scientists challenged the authority to define autism claimed by the psychiatric profession, based on the failure of psychotherapy to produce change in autistic children (Lovaas, Schaeffer, & Simmons, 1965). Cognitive psychologists, who have produced the notion that autism is the result of a “theory of mind” deficit, can be understood as competing with behavioral scientists for a position of professional authority. Their claims regarding autism are, in part, constructed as a challenge to claims of behaviorists, whom they characterize as “reductionist.” Premack and Woodruff (1978) framed their interest in theory of mind in part as a challenge to behaviorism and asserted that “to admit that animals are mentalists compels the admission that behaviorist accounts of animals are at best profoundly incomplete” (p. 526). Demonstrating the connection to primate research that introduced the concept of theory of mind, Baron-Cohen (1995) located his work on mindblindness in the study of “evolutionary psychology” (p. 9). As a group, cognitive psychologists claim to have created a more global theory of autism than earlier representations. According to Frith (1991):
The cognitive explanation of autism provides the most complete understanding of the cause of this disorder so far. … Alternatives to cognitive theories in the explanation of autism are psychodynamic theories and behaviourism. Neither of these has succeeded, partly because they focus on only some aspects of autism and partly because they ignore biological factors. (p. 16, emphasis in the original)
Claims and competition regarding professional authority occur within a single profession as well as between professions. Professions also evolve; psychiatry now bears little resemblance to the profession in Kanner's time. However, what remains common throughout this complexity is that professional perspectives typically gain legitimacy at the expense of earlier constructs. In the case of autism, although this can create an illusion of “progress,” the underlying features of professional representations do not change.
Autism Can be Defined and Understood by an Objective, Scientific Process
Practitioners of so-called “soft” sciences, like sociology, education, and psychology, have often pursued the legitimacy and prestige that a designation of “science” can offer (Lagemann, 2000). In these fields experimental design is constructed and justified with exquisite care. Results are presented quantitatively in a narrative style and format appropriated from the natural sciences. Over time, social scientists have increasingly employed the trappings of so-called “hard science,” in what seems to be a bid for greater legitimacy. Thus, Baron-Cohen (1995) presented his work as related to research in evolution and genetics and proclaimed, “psychology is finally becoming a genuine natural science” (p. xv).
Scientific experimentation is a powerful tool for examining aspects of our world, but decisions about what aspects of the world to consider or why are never entirely neutral. When examining something as complex as human behavioral variety, scientists are never value-free (Danforth, 1999), and what is presented as truth always involves representational choices (Duchan, 1998). For example, consider the striking difference between how theory of mind is defined and how it is evaluated. It is defined in a very broad way, as an ability to understand others' intentions, emotional states, etc. On the most basic level, it is supposed to be an understanding that others are like us, that they reason, feel, etc. However, it is evaluated in a very particular way, primarily through false belief tasks. We are told that people who struggle with false belief tasks do so because their brains lack a specific cognitive structure called a theory of mind module—ToMM (Baron-Cohen, 1995), and, therefore, they are unable to perceive others as thinking, feeling, reasoning beings.
Research results with autistic subjects do frequently demonstrate poorer than normal performance on false belief tasks. However, in at least one study (Bara, Bucciarelli, & Colle, 2001), autistic subjects performed as well or better than a control group under certain conditions. Furthermore, other populations also typically perform poorly on such tasks in research settings, and all autistic people do not fail these tasks (Yirmiya, Erel, Shaked, & Solomonica-Levi, 1998). Although the data produced by such studies are consistent with a hypothesis that autistic people may not easily interpret the mental states of nonautistic people, these data are produced in very particular ways, and many alternative explanations have been proposed as to why children might struggle with the false belief tasks that researchers create. These include explanations related to executive function (Orzonoff, 1995), language processing and timing differences (Williams, 1996), attentional differences (Bara et al., 2001), the idea of an interfering bias toward reality (Mitchell, 1996), or problems with intentional expression of ideas (Rubin et al., 2001).
No science is ever entirely objective, its occasional claims to the contrary notwithstanding. Scientists argue from (and for) particular viewpoints. Scientific rhetoric is always located in a specific historical, political, and epistemological framework (Danforth, 1999). The use of empirical evidence does not guarantee objectivity; one must ask who decides what counts as evidence and how evidence will be selected and presented (Collins & Pinch, 1998). A clearer picture of the meaning of research is often achieved by considering the “location” of the researcher; a consideration of privilege in an analysis of the agenda of the research can often clarify what otherwise might appear as contradictory data. The question must always be asked, “Who benefits from this research?” Many have claimed that the West, and America in particular, because of its position of privilege and power, is especially susceptible to being swept up in what Danforth (1999) called the “myth of scientific progress” (p. 95). In the case of theory of mind research, it is important to note that researchers write from a position of power based on dominant social norms, and their subjects are accorded little power, voice, or respect.
Autism as Exotic or Extreme
The puzzle metaphor is an example of how common it has been to construct autism as enigmatic. Frith (2003) explained and recreated this mystique as follows: “People in the past must have encountered autism and must have attempted to come to terms with it. The chilling and fascinating combination of childhood innocence and disturbance cries out for symbolic elaboration” (pp. 17– 18).
Autism is not only exotic in most of its representations. It is also often considered to be an example of something extreme, which exists at the far end of a continuum of humanity. Theory of mind theorists commonly depict autism as both exotic and extreme. Baron-Cohen (1995) asserted that the “terrifying” description below of what it might feel like to “suffer … from mindblindness … is not an idle thought experiment or a piece of science fiction. For some people [i.e., autistic people], it is very real” (p. 5).
This is what it's like to sit round the dinner table. At the top of my field of vision is a blurry edge of nose, in front are waving hands… Around me bags of skin are draped over chairs, and stuffed into pieces of cloth, they shift and protrude in unexpected ways. (Gopnik, 1993, cited in Baron-Cohen, 1995, pp. 4–5)
Although such speculative flights of fancy are engaging to consider, it is notable that they bear no resemblance to accounts created by autistic people, who may frequently report feeling like “aliens” (e.g., O'Neill, 1999; Williams, 1996), but not because they are unaware of the mind inside “bags of skin” that surround them. Compare the imaginative passage above to the following first-person account describing how one autistic author thinks about others' minds:
Speed is a good analogy here, I think. People often impress me as a speed, a frequency … NT [“neurotypical” or nonautistic] people … are busy absorbing all the ripples on the surface of a person and making a judgment on that basis. Meanwhile, I see the ripples as blinding or irritating flashes in my eyes (so I close my eyes, if I can get away with it), but/and some other part of me, some sense, is becoming aware of the frequency at which the core of the person is vibrating … NTs … don't realize anything is happening in/with me because I am not reacting (positively) to all those ripples and they assume I am not sensitive to other people. Well, I am. But I am slowly sensitive. (Jane Meyerding cited in Blackburn, Gottschewski, George, & Niki, 2000)
In the work of many cognitive scientists, autism appears to function as a counter-example, the purpose of which is to throw an author's ideas about cognition into clear relief. For example, Gopnik, Meltzoff, and Kuhl (1999) included a brief section called “Mind-Blindness,” in their book about theory of mind in typically developing children. The authors claimed that:
there is something about the way [children with autism] understand other people that makes them very different from the rest of us … Most of us are born with the ability to link our own mind and those of others. People with autism seem to have to solve the Other Minds problem from scratch. (p. 53)
In other words, autism is provided as an example of mindblindness in order to argue for a general biological human talent for mindreading. Similarly, Mitchell (1996) introduced the last two chapters of his book on theory of mind by announcing his intention to “seek to shed light on [what the child does know about mind with increasing age and how this knowledge is acquired] by looking at what happens when development goes wrong: The case of autism” (p. 164).
In these examples, the rhetorical need for a category of autism is more important to the authors' arguments than how any particular person with the label of autism might actually respond to others. Gopnik et al. (1999) asserted that children with autism “don't seem to have the fundamental presupposition that they are like other people and other people are like them” but insisted in almost the same breath that autistic children are “very different from the rest of us” (p. 54), apparently oblivious to the irony this creates. Gopnick, Mitchell and others are guilty of what Gelb (1997) called typological thinking: “the belief that individual differences diverge around an underlying type or essence” (p. 1).
Autism as a Form of Human Insufficiency
Theory of mind theorists rely heavily on a definition of autism based on what has come to be called “the triad of impairments,” a deficit-based construction first offered by Wing (1981). As part of an epidemiological study in Camberwall, England of children with disabilities, Wing became interested in distinguishing between a subset of the children that she labeled the “sociable mentally retarded” and a second group that she believed were “socially impaired.” Describing the latter group she asserted that “The abnormalities of social interaction, verbal and nonverbal communication, and imaginative activities so consistently occurred together that they could be referred to as ‘the triad of social and language impairment’” (p. 37).
Wing referred to her triad as “the ‘core’ syndrome,” but acknowledged that as an “attempt to explain the nature of autism and autisticlike conditions [it] leaves many loose ends” (p. 42). Loose ends or not, Wing's “triad of impairments” has become a diagnostic standard that is almost invariably equated with autism in modern psychological representations.
Deficit-based definitions of autism, such as Wing's triad, explain autism not simply as different from a norm, but as deficient with regard to the norm. Other more positive constructions of the differences in these three areas are possible. Both communication and social interaction, by definition, require more than one person, and difficulties in either area should properly be located between individuals and not within one individual. It is also worth noting that what gets counted as imagination or creativity is notoriously difficult to agree upon. For example, Park (2001) said of her daughter: “Jessy couldn't invent. She could only combine—sometimes in startling ways—what she found elsewhere” (p. 115). A nonautistic person with a talent for combining old things in new ways would likely be considered to be highly “creative,” but because Park (2001) accepted Wing's triad of impairments (which includes a failure of imagination) without question, she did not see this ability in Jessy as representing true creativity.
A deficit-oriented approach is dehumanizing. Wing (1981) claimed that “the child with the triad … is strange and puzzling to other people because … he lacks the whole complex pattern of action and response that makes human beings recognizable and important to each other” (p. 40). Mitchell (1996) asserted that “impairments in communication [in people with autism] stand as an impediment to the quintessentially human means of access to the minds of other people” (p. 185, emphasis added). Baron-Cohen (2001) used almost identical language, claiming that “[a] theory of mind remains one of the quintessential abilities that makes us human” (p. 169). Elsewhere, he stated that “No other species appears capable of anything close to the sophisticated mind reading humans engage in … This all suggests to me that mind reading, like many other mental abilities, has evolved further in humans. (Baron-Cohen, 1997, p. 65). In other words, the pieces that these authors asserted are missing in autism are those that define us as human.
Because theory of mind theorists characterize autistic people as subhuman, pitiable, and deficient, they frequently employ language that portrays autism as tragic. An individual suffers from autism; autism strikes; autism is cruel. “What is this puzzling disorder that is at once so subtle and so vicious in its effects?” (Frith, 2003, p. 5). Although autistic authors also sometimes describe the effects of autism as harrowing, many autobiographical accounts are upbeat, self-affirming, and demonstrate the idea that the authors' autism can be a strong, positive force in their lives:
In an article written by Oliver Sacks in The New Yorker, I was quoted as saying, “If I could snap my fingers and be nonautistic, I would not. Autism is part of what I am.”… I would not want to lose my ability to think visually. I have found my place along the great continuum.” (Grandin, 1995, pp. 60–61)
[Autism] is not something of the deep, dark, and buried past but instead is very much in the present. I came to understand that the autism spectrum is, and always will be, a part of me. All I can do is learn better how to work with, rather than against, the traits of this differently ordered way of being. (Shore, 2003, p. 163, emphasis in the original)
Autistic authors also express with increasing frequency the idea that if they “suffer,” it is more a result of discrimination than their autism:
Personally I find the basic stumbling block on which all the other issues stem is the assumption most Neurotypical people have that Neurotypical worldview is neutral and normative. Autism is then seen in contrast to [a] given, natural, neutral, and normal society. … The autistic person is then in the position of being in a society which does not understand autism. (McNabb, 2001, cited in Kluth, 2003, p. 19)
I have no wish to be cured of being myself. If you would help me, don't try to change me to fit your world. Don't try to confine me to some tiny part of the world that you can change to fit me. Grant me the dignity of meeting me on my own terms. (Sinclair, 1992, p. 302)
Autism and Intelligence
Cognitive scientists' representation of autism as a form of mental deficiency has also been evident in how they approach the construct of intelligence. Many children with autism labels perform poorly on intelligence tests, and assumptions about the validity of IQ as a construct were not questioned in the early work of theory of mind theorists, who did not hesitate to assert a robust association of autism with intellectual dysfunction. For example, Hermelin and O'Connor (1970), who were among the first cognitive scientists to focus on a study of autism, concluded that “The association between autism and mental subnormality in children, if not complete, is certainly extensive.… Either a common factor is associated both with autism and subnormality, or one is associated with the other” (p. 121).
Data on IQ also figured highly in the original Sally–Anne experiment and similar research, in which investigators compared populations of children labeled “autistic,” “mentally retarded,” and “normal” as determined by a mental age arrived at by intelligence testing. Speaking to the acceptance of and reliance on IQ data, Frith (1989) wrote:
The conclusions we can draw from [a 1970 study by Lockyer and Rutter] are very clear: careful assessment of autistic children's intellectual abilities by clinical psychologists is trustworthy. … We must conclude that the frequently obtained low-IQ estimates and the widely diverging results on different subtests reflect stable facts. They are consistent with educational achievements and the everyday functioning of the children. The problems do not lie with the tests or the testers but reflect the children's real handicap. (p. 83)
Thus, Frith claimed authority for her profession, essentialized the notion of intelligence, located disability in the child rather than in the interaction of child and his or her environment, and asserted that autism is a biologically based intellectual impairment.
The idea that intelligence is a linear, hierarchical phenomenon that intelligence tests can measure has been challenged by many. Gould (1981) argued that it is absurd to assign a single, unitary, rankable number to something as complex as human intelligence. Hayman (1998) made a strong case for the idea that such tests serve to maintain and reproduce the position of dominant social groups and that both tests and legal structures inevitably serve the interests of those with greater social, political, or economic power. In the recent American Association on Mental Retardation definition of mental retardation, Luckasson et al. (2002) considered a broad range of information about both intellectual and adaptive functioning. To the extent that theory of mind theorists ignore such nuances in their work, they are susceptible to overgeneralization based on typological thinking (Gelb, 1997).
Not surprisingly, given the enormous social capital typically accorded to intelligence, autistic authors have often asserted that they are intelligent (Biklen & Duchan, 1994). As examples of obvious ability in people with autism labels have increased, there has been a greater willingness on the part of cognitive psychologists to broaden the way that they regard intelligence. Consequently, although theory of mind theorists continue to regard mental retardation and its obverse, intelligence, as measurable, stable phenomena, in recent years they are less likely to argue that a close association exists between autism and mental retardation. Indeed, there has been great interest recently in defining what could be described as a peculiarly autistic intelligence. Thus, in the second edition of her popular book on autism (2003), Frith explained intelligence in a more complex and nuanced fashion than she did in her earlier edition (1989). This trend is clearly visible in the statistics; in 1989 Frith made the claim, common in its time, that 75% of autistic people were also mentally retarded, whereas by 2003 that statistic had decreased to 35% (p. 135).
In recent years, Baron-Cohen has made a shift in the way that he discusses autism and theory of mind. Baron-Cohen, Wheelwright, Lawson, Griffin, and Hill (2002) alluded to their profession's increasing interest in “more intellectually capable” autistic individuals as a rationale for this shift. Baron-Cohen (2003) accorded so-called mindblindness a position as only one, albeit a very important element in a larger schema that places humans on a continuum, one end of which is called “empathizing,” and the other end, “systemizing.” This continuum, which is described as biological and not merely cultural, relates to gender as well as disability categories; not surprisingly, males are seen as systemizers and females as empathizers. Autism has become an example of the “extreme male” brain, at least in regard to this construct. The advantage of the new Baron-Cohen over the old one is that in some regards he seems willing to consider autism a form of difference rather than a form of deficiency. Nevertheless, for Baron-Cohen, autism is still a unitary phenomenon representing the extreme end of a continuum of experience. He still clearly claims a professional right to explain it based not on the experience of autistic people, but on his construction of psychological research.
Whether or not particular theory of mind theorists credit or discount the intelligence of individual autistic people, the notion that autism results from or includes a theory of mind deficit goes far beyond impugning the intellect to assert the lack of such characteristics as empathy and emotional depth—traits that are said to “make us human.” Indeed, advocates of a theory of mind model appear to assume a causal relationship between cognition and emotion. The lack of a theory of mind, a cognitive deficit, is the missing puzzle piece presumed to be at the root of affective differences that, in turn, define autism as a sub-human state. Consequently, it is not surprising that even when intelligence and talent in autistic individuals are recognized, they are still commonly pathologized. Any facility demonstrated by an autistic individual is not likely to be referred to simply as talent; rather it is a “savant” skill (e.g., Frith, 2003). Once more commonly referred to as “idiot-savant” skills, these abilities are so called because they are considered to be entirely unanticipated in a person regarded as so utterly incapable. Similar damning with faint praise shows up in terms such as islets of ability and splinter skills, which remain common in descriptions of autistic people.
Patience of the Provider
“Helpers” of autistic people, professionals and parents, are consistently portrayed as being saintly, and the enormous patience of nonautistic people is valorized. Autistic people are rarely portrayed as patient. For example, in Park's (2001) description of her daughter's development, she often referred to the superhuman patience of Jessy's many helpers over the years:
No one person, or family, could provide all that Jessy needed to grow. There was always someone else working with her in those days … Endlessly inventive, endlessly generous, without them hers would be a different and sadder story (p. 77).
In short, achievements are not credited to Jessy but, instead, attributed to her various helpers.
We often use the tone of someone's speech or writing to decide whether or not we believe the person is patient. The tone of an autistic person's communication may seem very abrupt or otherwise confusing to nonautistic people. Autistic behavior is likely to be read as “persistent,” or perhaps “perseverative” or “obsessive,” but not as “patient.” By contrast, nonautistic authors, by positioning themselves as empathic, often are read as enormously patient. In the following “brief picture of autism,” Baron-Cohen's (1995) persuaviness is largely the result of a writing style that consistently seems to communicate gentle reasonableness:
Autism is considered the most severe of all the childhood psychiatric conditions. Fortunately, it occurs only rarely, affecting between approximately 4 to 15 children per 10,000. … The condition may be associated with many biological abnormalities, such as epilepsy, mental handicap, and a variety of brain pathologies. It also appears that in many cases there is a genetic basis to the condition, since the risk of autism or related problems in identical twins or biologically related siblings is substantially higher than would be expected if autism just struck “by chance.” At present, autism is unfortunately a lifelong disorder. Thankfully, it sometimes appears to alleviate a little with age. (p. 60)
In passages like this one, Baron-Cohen gets much mileage from his adverbs. Even as he disparaged his autistic subjects by reciting a daunting list of missing pieces, he used fortunately, unfortunately, and thankfully to position himself as empathic. Words such as symptoms, condition, or pathologies indicate that his is very much a medicalized view, and those same adverbs, as well as words like risk and struck are used to convey a dire, deficit-oriented perspective. Throughout, however, Baron-Cohen conveyed that he cares, and indeed he cares in a way that the unfortunate autistic person—who lacks a theory of mind—cannot. Underlying Baron-Cohen's seemingly empathic tone is a perspective that autism is a pitiable state. Characterizing those with autism labels as “unfortunate,” privileges him, as well as his presumably nonautistic readers, as fortunate, creating unbridgeable distance and precluding real empathy.
Assumptions Not Questioned
I have no serious doubt that this theory (what I call “common-sense belief/desire psychology”) is pretty close to being true. (Fodor, 1983, cited in Baron-Cohen, 1995)
The assertion that people with autism are mindblind is considered entirely uncontroversial by the professional community of cognitive psychologists, who vigorously debate the minutiae of theory of mind theory in every other regard. For example, Mitchell (1996), who evaluated data from a variety of experiments in which false belief tasks were used to assess the development of a theory of mind in typically developing children, asserted repeatedly and forcefully that an absence of evidence of children's correct performance should not be considered evidence of absence of theory of mind and suggested that we would learn more by looking for circumstances in which competence can be displayed. When he turned to a consideration of autistic children, however, the notion that autistic children are incapable of insight into others' minds is simply assumed and presented unexamined as a definition of autism.
Similarly, Bloom and German (2000) argued compellingly that failing false belief tasks does not demonstrate that a child lacks a theory of mind, but they did not question the assumption that it is the lack of a theory of mind that makes children with autism differ from most other children, despite their vigorous critique of the false belief tasks that were the original basis for such an assertion:
Normal 3-year-olds and older children with autism both fail the false belief task, but, in all interesting regards, normal 3-year-olds are nothing like older children with autism. Normal 3-year-olds are far superior with regard to communicative and linguistic skills, the ability to pretend and understand the pretence of others, and the ability to engage in, understand and manipulate the actions of others … 3-year-olds might fail the false belief task because of general task demands, because they don't have a grasp of false belief, or both. But they surely have a “theory of mind,” in the general sense of having a sophisticated ability to reason about the mental states; this is precisely why they differ from autistic individuals in the social, communicative, and imaginative domains (p. B29).
Although theory of mind theorists are unlikely to question what have become dominant assumptions regarding autism, autistic authors are poorly served by such assumptions and are more likely to explore what underlies them:
Because Autistic people see most normal people as seeming to assume everyone is like themselves, and would react as they would in the same situation, normal people may often seem to lack “Theory of Mind”… to many … Autistic people. On the other hand, normal researchers are tempted to assume lack of or deficiency in “Theory of Mind” when Autistic people don't automatically jump to these conclusions (Blackburn cited in Blackburn et al., 2000).
These attitudes did injustice not only to those people who did not fit these stereotypes, but also to those people who did appear to display these things. Though unexpressed, some of these people did have empathy, emotions, a sense of pain, a sense of humour, imagination (and interest and curiosity). Their otherwise “bizarre” behaviours not only made sense to them but, in some cases, were actually adaptations they'd discovered to help them to make connections, calm themselves down, and remain in control better than they may otherwise have been able to. (Williams, 1996, p. 12, emphasis in original)
Mistrusting Autistic Communication
When autistic people talk about themselves, they may at times also use explanations that resemble those of theory of mind theorists to describe autism. However, other fundamentally different views also inevitably emerge from a first-person perspective based on the idea that rather than trying to cure difference, we should accommodate to it (e.g., Sinclair, 1992).
Beginning in the late 1980s, autistic people such as Donna Williams and Temple Grandin started to publish first-person accounts that described their lives as autistic people living in an often-inhospitable nonautistic world. Many readers of these early accounts questioned whether or not the authors were really autistic. The assumption underlying such responses is that autism is so incapacitating that Williams or Grandin could not be autistic and still write with such insight and sensitivity. When their books first appeared, autistic authors were characterized either as frauds or exceptions. A typical example of such controversy is an article from The Australian newspaper with the headline “Best-Selling Author Denies She Faked Autism” (Haslem, 1996). Happé (1991) hinted that we should mistrust our first impressions of Temple Grandin's first book because of the involvement of an editor:
An important point to note about [Grandin, 1986] is that it was edited by Margaret Scariano, a children's writer, who rewrote sections of the book, gave it its flashback format and generally structured it to make it easier to read. This obviously presents us with problems, casting doubt on exactly those passages which are most interesting and challenging to our ideas about autism. (p. 208)
Grandin wrote her next book about autism entirely by herself (Grandin, 1995), making such critiques irrelevant.
When voices representing a marginalized group are first heard, it is perhaps inevitable that their ideas will be dismissed, but eventually these voices have the power to transform discourse. Today, Williams' and Grandin's “credentials” are rarely challenged. Rather than being doubted, they are more likely to be commended for offering us an “inside out” view of autism. Several other voices have emerged as well and are more often published or displayed in the media (e.g., Blackman, 1999; Muckopadhyay, 2000; Shore, 2001; Willey, 1999). Despite this positive trend, it remains much more common to see autism interpreted by an autism “expert” than by a person with autism. Such experts, as nonautistic interpreters of the syndrome, however respectful of autistic individuals their approach may seem, necessarily define autism in terms of its deviance from their understanding of normality.
Autistic people, like those without the label, vary in their ability to express their ideas to others. Grandin, Williams, and similar authors are unusually articulate by any standard, but no one who has seen them communicate in person would claim that they do so in a typical, nonautistic manner. Other people with the label of autism frequently experience even more dramatic challenges with regard to communication. Indeed, it is virtually impossible that someone would be labeled autistic if no differences in communication were present, and many people with autism labels do not speak, but communicate only through writing or through their behavior. The multiplicity of communication styles in autism creates special challenges in incorporating first-person viewpoints into otherwise inadequate representations of autism. Those of us who are “neurotypical” may find it much easier to speak about, and ultimately to speak for, someone who is not speaking; we think there is no fear of contradiction!
Historically, autism has always been defined by a variety of scientific professions from the “outside in.” The label of autism was born out of clinical descriptions of behavior. Certainly, the way to transcend deficit models of autism is to formulate a new definition that is not determined so much by behavior as by experience. How does the person with autism encounter the physical world, human relationships, etc.? This is problematic, however, because the communication we can expect from people with autism labels is so often atypical; frequently described by both autistic people and their advocates as unreliable, and may be, at times, simply too hard to understand. Thus, although the idea of relying on autistic voices in order to understand autism seems self-evident, questions regarding the reliability of the autistic voice are inevitable.
A symbolic interactionist perspective (Blumer, 1969) allows us to treat problematic tasks of definition respectfully and with regard to human value because symbolic interactionism holds that meaning is generated socially, namely, at a site where Self and Other meet and interact. Thus, symbolic interactionists understand “defining” a social identity like autism as an act of communication, which, by its very nature, is collaborative. In the process of defining, both autistic and nonautistic people learn and are transformed through mutual relationships. The question to ask is not simply, “How do we know what we know?” but also “How is knowledge evoked through our relationship with others?”
Disregard for first-person experience does not only occur for those individuals who, like people with autism, possess disabilities that affect communication. Indeed, it is perhaps inevitable whenever there is an asymmetrical distribution of social power. Recognition of the ways in which the tyranny of the norm disempowers and exploits marginalized groups and individuals is the foundation of any perspective that works for increased social justice (Biklen, 2000). This is why progressive and critical perspectives strive to give voice to previously silenced constituencies. There is no reason this cannot happen with regard to autism and similar disabilities. Despite the impossibility of finally answering questions regarding how we know what we know, using what Donnellan (1984) called the “criterion of the least dangerous assumption” (p. 142) allows us, along with Linneman (2001), to “authorize” the “mindedness” of people with labels like autism. “Mind can only persist to the extent that it is experientially preserved, and who will uphold, articulate, and authoritatively assert, ‘yes, this person has a mind?’” (pp. 65–66)
Dilemmas with regard to how communicative ambiguity affects self-definition are not unique to autism. They are perhaps exaggerated by the different kind of voice we might expect from autistic people, one in which the agendas for and manners of communication are sometimes vastly dissimilar from the norm. However, all claimants to social identity ultimately have to grapple with similar questions and live with similar ambiguities. Ultimately, the question of who has the privilege to define autism—or any category of human difference—is not a scientific, but a philosophical and political one.
The models of autism that theory of mind theorists create are inadequate in large part because they fail to consider what autistic people think and feel. The theory is, in effect, “mindblind” with regard to autistic perspectives. Because assumptions about disability interfere with an ability to construct autistic people as human (Linneman, 2001), those who are most able to inform us about autistic experience are ignored.
Alternate constructions are possible when autistic voices are made part of the discourse. Bogdan and Taylor (1989) illustrated how social constructs regarding severe disabilities change depending on the relationship of an observer to the person who has been labeled. Those with intimate relationships with individuals presumed to be severely retarded (their loved ones, neighbors, and friends) often construct them as competent, likeable and, ultimately, “human.” Goode (1992) distinguished between critical professional constructions of Bobby, a man with Down syndrome, and a more intimate perspective that Goode's research team was able to achieve through intense analysis of Bobby's interactions on videotape. Kliewer and Biklen (2001) argued that when educational inclusion allows all children to be full citizens in a community of learners, regardless of the particular abilities that they demonstrate overtly, literate behaviors previously assumed to be absent begin to emerge. Similarly, Rubin and her colleagues (2001) argued that when competence is presumed, individuals with autism are more likely to engage with others in complex ways and develop more communicative sophistication.
The jigsaw puzzle metaphor, which suggests that our interactions ought to be about finding missing puzzle pieces in order to complete the picture, creates an illusion of tidiness that is belied by experience. Incorporating autistic voices into representations of autism offers an important counterbalance to professional limitations by challenging us to reinterpret autism as a form of human variation instead of human insufficiency. A process of including first-person perspectives of autistic people is the only way that definitions based on difference rather than deficits will emerge to change negative social constructions of autism. This requires a willingness to tolerate some communicative ambiguity, but no alternative will accord dignity to people like my son. A commitment to democratic and inclusive communities requires a process of communicative partnership that elicits narratives based on all voices, even those that are difficult to decipher.
Author: David Smukler, MS, Doctoral Student and Adjunct Professor, Department of Education, Syracuse University, Syracuse, NY 13244. email@example.com