Researchers have proposed numerous theories to explain autism, ranging from those that are psychologically focused to those influenced by biology and neurology. Many theories of autism share the assumption that there is a deficit in people with autism that should be researched, classified, and, ultimately, modified if the hypothesis suggests that this may be possible. Therefore, the common link among these theories is an assumption that there is something wrong with the person with autism. One highly influential theory purporting to explain the characteristics of autism is that they are caused by an inability to understand others' mental states; in other words, autistic people are considered to lack a “theory of mind.” This theory is dominant in explanations of autism, with significant influence within the professional field and in constructions of the person with autism. In this article, we critically examine the theory of mind hypothesis using empirical analysis of online material in which neurodiverse adults, including some who identified with the label of autism, reflected on theory of mind. Neurodiversity is a term that was initially conceived by people with autism in their reframing of deficits and differences and first appeared in an academic essay by Judy Singer in 1999. There has been a call for neurodiversity to be recognized and considered among with the more familiar categories of gender, class, and race (Singer, 1999).
What Is Theory of Mind?
Theory of mind was a term originally coined by Premack (1978) to describe the intentions of primates (Leudar, Costall, & Francis 2004). It has since been developed and applied across a range of disciplines within psychology, notably developmental psychology, and studies investigating “abnormal” behavior such as autism. Tager-Flusberg (1999) defined theory of mind as referring to “the ability to attribute mental states, such as desire, knowledge, and belief, to oneself and other people as a means of explaining behaviour” (p. 326). Therefore, people with autism are thought to be impaired in the ability to appreciate their own and other people's mental states (Baron-Cohen, 1998). It has been proposed that theory of mind develops as the child matures, with the ability first emerging toward the end of the first year when infants begin to view people as having intentions in their behavior patterns (Tager-Flusberg, 1999). By 3 years old, the child can understand desires and simple emotions in themselves and others and is able to discuss a person's actions in terms of the mental states that cause them. This ability improves by 4 years, whereby children can understand more complex mental states, notably the concept of beliefs, and that people may hold beliefs that conflict with reality (Tager-Flusberg, 1999). The ability to develop theory of mind is assumed, therefore, to be part of the typical developmental pattern for individuals.
Testing Theory of Mind Propositions
Several tasks have been developed to investigate the phenomenon of theory of mind and how it may explain autism. A landmark study investigating theory of mind was carried out by Baron-Cohen, Leslie, and Frith (1985), known as the False Belief Test. The test examined the hypothesis that children with autism could not understand that other people may have beliefs that differ from their own. The now-famous “Sally-Anne experiment” compared the performances of children with autism, children with Down syndrome, and typically developing children on their understanding of false beliefs. In the experiment, the researchers introduced the children to two dolls, Sally and Anne. Sally had a basket and Anne had a box. The researchers then acted out a short sequence for the children involving the two dolls. In the sequence, Sally had a marble and placed this in her basket; she then went for a walk. Unknown to Sally, Anne took the marble from the basket and placed it into the box.
The children were then asked to predict where Sally would look for the marble when she returned to the room. To give the correct answer (i.e., that Sally believed that the marble was still in the basket where she left it), the children have to had understood the concept of false belief and an appreciation that another person could have a belief that contrasted with their own. Therefore, the child must have disregarded his/her own knowledge of the true position of the marble in the box and responded that, because Sally did not witness Anne move the marble, she would believe that the marble was still in the basket where she left it.
In their study, Baron-Cohen et al. (1985) found that 80% of the children with autism failed to give the correct answer to the question concerning what Sally's beliefs were, although they could answer control questions concerning where the marble was originally and where it was moved to. The results for the children with autism contrasted with those for the other two groups of children, who gave more correct answers to the false-belief question despite the fact the children with autism scored higher than the other two groups of children on both nonverbal and verbal scale measurements. Baron-Cohen et al. (1985) concluded that the results of the experiment showed that the children with autism did not appreciate the difference between their own and the doll's knowledge. The authors further argued that such a failure by the children with autism in their experiment constituted a specific deficit that could not be attributed to the general effects of cognitive ability, because the children with Down syndrome performed well on the task and were at a lower measured cognitive level. Therefore, they proposed that they had identified in their experiment a specific cognitive impairment that is largely independent of general intellectual level and has the potential to explain both the lack of pretend play and the social impairment often witnessed in children with autism.
Several discussions have focused on the problematic nature of such tests for theory of mind. Tager-Flusberg (1999) proposed that one of the main problems associated with such tests is that some people with autism can complete the tasks. Tager-Flusberg also noted that there were developmental inconsistencies with the theory in that theory of mind should develop in children by 4 years of age, but by then, there are already many signs of autism. However, despite such reservations, Tager-Flusberg endorsed a connection between a theory of mind hypothesis and social impairment in people with autism. Tager-Flusberg argued that the social world is complex for the person with autism, because they have difficulty in understanding and predicting other people's actions. However, this may be less true for people who they know well and in situations that they are familiar with, as a more structured routine can be established that makes interaction with these significant others less uncertain for the individual; hence, a close and affectionate social interaction can be evident.
In theorizing and testing the theory of mind hypothesis, a number of assumptions are made concerning the abilities of individuals and how such deficits may be explained and understood through the hypothesis. In the rest of this article, we focus on the construction of theory of mind as a deficit model of autism.
Assumptions in Theory of Mind: Theory of Mind as a Deficit Model of Autism
Although the theory of mind hypothesis is not specific to autism and not linked to intellectual attainment specifically, literature proposing explanations for behavioral characteristics in autism draw heavily on the hypothesis. Theory of mind has been proposed as a core element explaining deficits frequently outlined in people with autism (e.g., Baron-Cohen et al., 1985; Guajardo & Watson, 2002; Tager-Flusberg, 1999). For example, impairments in theory of mind have been used to explain a lack of pretend play among children with autism and may be important in explaining communication difficulties (see, e.g., Guajardo & Watson, 2002). Therefore, the focus of the theory of mind hypothesis is on an inability, which has important consequences in a variety of situations and may influence aspects of social interaction, including the ability to understand feelings and mental states.
The theory of mind hypothesis is also used to explain difficulties in language and communication, which is cited as a core deficit in the diagnosis of autism. Tager-Flusberg (1999) proposed that people with autism have specific problems in understanding that language is a means of interacting with others and for sharing perspectives and ideas. Tager-Flusberg argued that people with autism at all ages have difficulty in taking into account the listener's perspective, which affects engagement in conversation. Tager-Flusberg concluded that such language and communicative deficits in people with autism can be directly attributed to an impaired understanding of others' minds.
It is important to note that much of mainstream research has been produced from an outsider's perspective, by “experts” who offer a particular view of autism. The voices of those with autism are rendered either marginal or silent within these research projects and commentaries. We feel that people with autism have an expertise that is often overlooked within the debate, and we focus on these within this article.
Critique of Theory of Mind
Representing Autism as a Unitary “Disability”
Reflecting on the concept of theory of mind, Smukler (2005) observed that autism is viewed as a unitary phenomenon, arising from a single core disability, rather than a multifaceted entity. The basis for such an assertion is provided through science that ultimately maintains the professional expert voice as the authority with which to explain autism, marginalizing the voices of people who have autism. Such representations Smukler argued, present autism as a deficiency rather than a difference. Such “damaged” individuals therefore require specialist help to function in society. Smukler argued that these depictions of people with autism have become the dominant perspective and, hence, normalized and not questioned.
Much work has examined the concept of theory of mind, and Costall and Leudar (2004) commented that rather than appearing as a new and dominant theory, many of the assumptions of the theory of mind hypothesis were visible in psychological research several years before the hypothesis came to adopt its dominant position. Indeed, Costall and Leudar (2004) argued that the hypothesis amalgamated these assumptions and came along when developmental psychologists needed a theory that would hold up to close experimental scrutiny. Within experimental psychology, the accessibility of the hypothesis to experimental testing contributed to the acceptance that psychological experiments could represent the situations in which people have to understand other's minds generally (Costall & Leudar, 2004). Smukler (2005) argued that it is this acceptance that maintains the position of scientific experimentation as a powerful tool for examining human beings. Smukler, in agreement with other critical theorists, however, warned that decisions about which aspects of individuals, and society more broadly, to consider are never neutral; in other words, scientists are never value-free. This is a point that is evident in some writings by people with autism, such as Klein (2002), who asserted that the ideals of the theory of mind hypothesis value non–autistic/neurologically typical traits. The valuing of such non–autistic/neurologically typical traits rejects the possibilities for presenting autism as a difference rather than a deficit and, hence, prevents an ethos of neurodiversity.
Dyck, Ferguson, and Shochet (2001) questioned the specific role that a theory of mind may have in the explanation of autism. They argued that, because of the varying abilities on theory of mind tasks by children with autism, it was not necessarily specific to such a spectrum of disorders. Dyck et al. therefore proposed that theory of mind ability should be seen as a disability dimension affected by a range of factors and not necessarily as a marker for a specific disorder. Dyck et al. argued that participants in studies have emotional empathy impairments in addition to theory of mind impairments and that both emotional empathy and theory of mind deficits are not unique to autism spectrum disorders but can occur in a variety of children. However, despite such concerns, the theory of mind hypothesis remains influential in current research approaches to autism.
Purpose of This Article
In this article, we critically examine the theory of mind hypothesis using a small number of online sources from people with autism. We chose to focus on such materials due to previous research proposing the benefits of computer-mediated communication for people with autism and the ways in which the medium has been used by people with autism to challenge understandings of autism through the development of more empowering identities (see, e.g., Brownlow & O'Dell, 2006; Ward & Meyer, 1999). The materials we used are from Web sites that primarily focused on the celebration of neurodiversity and were selected because they discuss theory of mind from a perspective outside of the dominant professional discourse. The articles hold particular importance in the discussions for several reasons: first, because they reflect a voice of people with autism on a topic about which discussions are normally dominated by professionals and, second, because they provide an important challenge to the scientific concept of theory of mind as a means of constructing people with autism as deviant or deficient rather than different. The contributions are cited here in conventional academic style, as we feel they are documents written by experts in the field of autism.
Klein (2002) proposed that theory of mind explanations reflect the dominant way of thinking in society in that they favor neurotypicality (i.e., nonautistic traits). Klein argued that, implicit in discussions of theory of mind, is the assumption that the neurotypical way is the only way, and, as such, people with autism are at fault because they are not like neurotypical individuals (NTs). Klein reflected on the theory of mind hypothesis and the way that it characterises autistic and nonautistic individuals:
If one of my kind cannot figure out what a normal person is thinking, it is a theory of mind error…it is the fault of the autistic for not being like the NT. If a normal person cannot figure out what one of my kind is thinking, it is because we are not using the proper means to tell you. Again it is the fault of the autistic for not being like the NT. (Klein, 2002)
Klein (2002) therefore argued that within the theory of mind hypothesis, individuals, and specifically people with autism, cannot be different. If they do not behave in a similar manner to NTs, they are ultimately classified as being impaired and failing the test. Smukler (2005) echoed this point and argued that theory of mind theorists ultimately define autism in terms of an insufficiency in skill that constitutes a problem that must be fixed. Such individuals are rarely constructed as being different by such theorists and, hence, accommodated; instead, traditional literature constructs such individuals as being impaired and requiring specific intervention.
Alternative Theories of Mind
Perkons (1998) presented a theory of social delusion and, within it, questioned whether there is really such as thing as a theory of mind. Although Perkons adopted a humorous approach to the identification of NT deficits and included a page explaining this for those who take things more literally, the author made some important points with regards to our understandings of and representations of autism. Perkons notable observed that
A theory is usually something that can be written down or explained to another person, yet the people who supposedly have it, don't ever explain it. This makes me wonder if it exists at all…before I buy into a “Theory of Mind” as something other people have, but I don't, I want to know what it is. (Perkons, 1998)
Perkons (1998) developed the argument that people without autism, or “normal” people, do not tend to consider the possibility of people living in separate worlds and having very different outlooks on life. Perkons argued that people with autism have a very good theory of mind, but that it is a different theory from that of nonautistic people.
Perkons (1998) also reflected on the specific tests devised by professionals to test the concept of theory of mind and described a classic theory of mind test where sweets are put in Box A and, then, when a person is out of the room, they are transferred to Box B. The question posed to participants is, where will the person look for the sweets on their return? Perkons noted that it is generally observed that both autistic and nonautistic children will say Box B at around 3 years but that their rationale for doing so might be very different:
I think the NT kids were assuming that the “person” was the “same” as they were, an extension of their own mind and emotions, and therefore would think the candy was in box B, just as they knew it was…The autistic kids would have been baffled, as I would have been. From my experience adults and older kids knew a lot of things that I didn't. In fact, they could do a lot of things that were “magical” to me—they could drive cars, they knew where the cereal was without looking for it… etc. So how can a sensible child possibly imagine what on earth the adult or older kid might or might not know? For all he knows adults might have x-ray vision. (Perkons, 1998)
Perkons summarized the findings from the assessment tests by highlighting important differences between people with autism and NTs:
The basic difference seems to be:
NT Theory of Mind = Everyone thinks like me, except when shown to be otherwise.
Autistic Theory of Mind = Everyone thinks differently from me—vastly and mysteriously—except when shown to be otherwise. (Perkons, 1998)
However, despite the different strategies and possible theories of mind as proposed by Perkons (1998), the findings from theory of mind research typically characterize the performance of people with autism on such tests as deviant or impaired rather than different. This view of proposing challenges to the dominant theory of mind has become characteristic of a more general move within academia and by autistic people.
Discussion: Implications for Understandings of Autism
In the discussions above, we have considered come of the key ways in which traditional assumptions about theory of mind have been challenged by people with autism, both through their questioning of methods used for testing the theory of mind hypothesis and the assumptions about individuals based on the outcomes of tests.
For example, Klein (2002) proposed that theory of mind explanations reflect the dominant way of thinking in society in that it favors neurotypicality. Klein argued that, implicit in discussions of theory of mind, is the assumption that the neurotypical way is the only way, and, as such, people with autism have a deficit because they are not like NT individuals. By accepting the traditional theory of mind hypothesis, a certain construction of people with autism is assumed. Smukler (2005) argued that such representations present autism as a deficiency rather than a difference. Such “deficient” or “damaged” individuals therefore require specialist help to function in society. Smukler argued that these depictions of people with autism have become the dominant perspective and, hence, normalized and not questioned. However, this professional discourse of difficulty and deficiency is difficult to sustain when reading the Web site entries, such as those cited earlier, and postings to online discussion lists where such discussions frequently present sophisticated challenges to expert knowledge bases (see, e.g., Brownlow, 2007; Brownlow & O'Dell, 2006)
The impairment in theory of mind of people with autism as reflected in an inability to communicate effectively in social situations may therefore be a reflection on the method of interaction rather than the abilities of people with autism. After an appropriate channel of communication has been established, complex reflections and discussions can be seen both in online discussion list exchanges and Web site contributions. Some contributors critically discuss the concept of NT in ways that not only appreciate that others may have thoughts different from their own but present these thoughts by reflecting on complex constructions of the characteristics of autism and neurotypicality.
The online exchanges between such groups frequently call for the conceptualization of alternative theories of mind. The use of theory of mind by professionals and academics is dominated by scientific thinking that presents theories as objectively researched facts, and, hence, although theory of mind remains a hypothesis, it is frequently presented by writers as an explanatory “thing”. Although there is no acknowledgment in dominant traditional psychological literature of the concept of an alternative theory of mind, a body of research is emerging that discusses alternative theories of mind (see, e.g., Costall & Leudar, 2004; O'Dell & Brownlow, 2004). Such alternative conceptualizations present challenges to dominant constructions of autism and question the deficit model in favor of a position that celebrates neurodiversity. Embracing neurodiversity would have important social and cultural advantages by drawing on the talents and skills of a population who have traditionally been marginalized through the framing of abilities as impairments.
The method that we selected for this investigation drew solely on online sources that reflected on issues concerning the theory of mind hypothesis. We acknowledge that these texts may not be representative of all people with autism but form part of our general stance of representing the important voices of people with autism within such a debate. This work is also similar to previous work that we have carried out in online discussion lists that challenged the deficit model of autism (see, e.g., Brownlow, 2007; Brownlow & O'Dell, 2006). Furthermore, the theory of mind hypothesis represents people with autism as one homogenous group, and we challenge this assertion, instead calling for a focus on differences within the group labeled autistic as well as discussing commonalities. Therefore, we do not believe that the voices prioritized in this article represent people with autism in general, as there is not a single voice of people with autism but many voices that can contribute to this rich debate and have significant positive influences on both academic theorizations and society more widely.
Charlotte Brownlow, PhD (email@example.com), Lecturer, The Open University, Milton Keynes, MK7 6AA United Kingdom. Lindsay O'Dell, PhD, Lecturer, The Open University, UK.