Mindreading

Dorcas Chung

 

The purpose of my paper is to explore how we come to predict and explain the actions or behaviors of other people, and further to examine how this prediction and explanation enables us to socially interact and communicate with one another.   I specifically want to explicate the theory developed by Simon Baron-Cohen known as mindreading, and examine how this theory applies to people with autism.  He uses his model of autism to argue that people with autism are deficient in their development of a theory of other minds.  However, I will argue that his position on autism does not take into account the difference between the way one interprets an autistic subject’s actions, and the way that that autistic subject interprets her own actions.

 

Evolutionary Advantages of Behavior Prediction:

Before I begin, I would first like to mention that the prediction of actions and the communication amongst members inside and outside of one’s own species is so important because it is said to enhance survival and reproduction.  Simon Baron-Cohen compares this ability to the game of chess.  First, each individual member (chess piece) needs to be aware of each of its own members, know how to interact with one another, and strategize in order to survive and progress [1] .  An example of this can be seen in apes, who like humans have a social hierarchy that enables them to live harmoniously with one another, thus ensuring better results in gathering food, raising young, protecting from predators, and navigating through various territories. [2]    Second, each player needs to be aware of the position of its individual members in relation to her opponent’s positions.   Deception, for instance, has been shown to be beneficial for survival purposes.  When a bird sees a predator approaching its nest, it will act as if it is injured so as to divert attention from the nest to the bird itself.  By deceiving the predator into thinking that the bird is an easy kill, the bird can better ensure the survival of its young [3] .  

Humans, however, seem to have this ability of predicting behavior and of communicating to a greater extent.  We are able to know more than that we have such an ability, or that we are able to survive by predicting and communicating in this manner or that manner.  We possess the complex language system of speech, and a capacity to introspect that allows us to further explore specifically why and how this ability comes about.  Because of that, we are not only able to survive; we are able to advance in ways unique to humankind: socially, technologically, artistically, philosophically, and so on.  My hope is to utilize this skill of introspection and language to examine the mechanism behind the ability to predict behavior and communicate.  

 

Three Theories for Behavior Explanation and Behavior Prediction:

Baron-Cohen describes three theories used to examine how we are able to predict and explain the actions of another person.  I will first present the theories he rejects, and then try to explicate the theory that he endorses.  The first theory is what he calls the Physical Stance.  According to this stance, one attempts to understand human actions in terms of the functions of its biological and chemical mechanisms.  For example, we can say that an individual observes an action, and then makes logical deductions that six particular brain states were activated in a particular order, which influenced this person’s particular action. [4]   We include in this stance such connectionist models as the Hopfield Net, which is a model that describes what is going on inside the brain by way of connecting nodes which yield either positive activities – the recognition of particular traits or concepts – or negative activities – the absence of recognition of particular traits or concepts.  This in turn allows an individual to act according to whatever part of the memory is stimulated or unstimulated [5] .  

Baron-Cohen argues that the Physical Stance has an epistemological problem, namely, that it assumes humans will have the ability and the knowledge to understand those biological and chemical mechanisms that cause one to act and behave in a particular manner.  While this model may be accurate, it is impossible to understand the functionality of such non-observable parts of the human system.  Baron-Cohen writes that “neither ordinary folk nor scientists have a ‘brainscope’ with which to ‘see’ what brain states are active in another person at any given time and then to predict that person’s next action.” [6]   In effect, this model is too complex for most people to recognize or understand. 

The second alternative, known as the Design Stance, attempts to provide a simpler model of behavior and communication by focusing on the functions of a person’s observable parts.  Just as the delete key on a computer signifies the erasing of a program or words on a document, so certain external detectors on a human signify a particular action or behavior.  The behaviorist approach to psychology is a way of understanding via the Design Stance, whereby communication and behavior are recognized by way of reinforcement [7] .  For example, if you poked my eyes, my body will react in a particular way.  I will immediately move so as to avoid the pain of being poked, and I will blink, all presumably in order to protect my eyes.

While the Design approach can also be successful in inferring unobservable processes, it is still lacking.  Mechanical devices such as boom boxes, computers, or cameras have clear observable parts that enable one to more easily understand its functions; however, humans and other sentient beings do not have such clear observable parts.  Our observable parts are limited to things like reflexes or reinforcement behavior, which under this stance thus limits from adequately predicting and intuiting those behaviors that provide no clear indication of a particular action or lack of action. In other words, while the Design Stance is good at describing or explaining behavior, it is insufficient in predicting behavior. 

The final approach, called the Intentional Stance, is the alternative that states that humans will naturally attribute purposeful attitudes to other people, and even anthropomorphize purposeful attitudes to animals or inanimate objects.  For example, if you observe me walk from my bedroom to my computer, you will assume that I have a purpose in walking to my computer.  You may guess that I want to check my email, or I intend to work on my paper.  Also, I will often times open up a Word document and begin typing when a little paperclip animation appears and says, “You seem to be writing a letter.  Would you like some help?”  In this case, I may say that my computer thinks I want to do something I don’t really want to do.   I can similarly apply a purposeful state about my thermostat in saying that it wants to keep my apartment at 70°F during the winter time and will therefore shut off when it thinks my apartment is getting too hot. [8]  

Purposeful action is consistent with the assumption that mental states have what are called propositional attitudes.  These attitudes provide some sort of causal link between a person’s thoughts and a person’s words and actions.  To explain, let us consider the words “intend”, “desire”, “believe”, or “fear.”  We can say that such words signify attitudes that are directed towards particular propositions.  While these kinds of attitudes are considered internal or inside a person’s mind, the propositions they are directed towards are normally external or outside the person [9] .   In other words, something is being said about somebody’s mental state, namely, that he has an attitude directed towards something, and that something is the proposition.  For example, I can say: “Dorcas intends to finish this paper,” where the word “intend” is the attitude, directed towards the proposition of finishing this paper.  The same goes for “David Baxter desires to speak to me,” “Jill Cavanaugh believes that I’m a dork,” and “Samuel Chung fears confrontation.” 

The notion of propositional attitudes is important because it is believed that an individual will perform some action such that the action will signify an intent to actualize that which is related to what that individual has an attitude about.  Consider the following example: I walk to my kitchen counter at 7AM, pour cereal into a bowl, and proceed to open the refrigerator door to take out a carton labeled “Milk.”  What these actions signify are the propositional attitudes I have about certain things, and the intent to bring about circumstances such that I am able to react accordingly to these attitudes.  Hence, I act in this way because I believe that the carton labeled “Milk” indeed contains milk; I take out the carton because I desire to pour milk into my bowl of cereal; and I want to do this because I am hungry, and I believe that this cereal will satisfy my hunger. 

The assumption that another person’s actions signify such mental states is extended to other people by way of empathy [10] .  By empathy, we mean that an individual has the natural ability to understand that another person is another person – one with mental states and propositional attitudes – just like the individual himself.   I can make the inference that, since my actions are causally connected in some way to my propositional attitudes, the actions of other people will also be causally connected to their propositional attitudes.  When I perceive what is external (a person’s actions), I make a prediction about what is internal (a person’s mind) [11] .  Hence, if I observed another person pour cereal in a bowl and proceed to take out a carton labeled “Milk” from the refrigerator, I will likewise presume that she believes that the carton contains milk, and that the milk will accompany the cereal that she will eat to satisfy her hunger. 

 

Theory of Mind and Theory of Theory of Mind:

In order to avoid confusion, I should make the note here that I am exploring this issue from two levels of analysis.  I will call the first level theory of mind, and the second level theory of theory of mind.  The first level, also called mindreading by Baron-Cohen and mentalization by Uta Frith, describes strictly the kinds of mental states an individual attributes to another person through his observing this other person’s actions and emotions.   We may place in this category of theory of mind the notion of folk psychology.  Folk psychology is a basic way of describing the general, everyday commonsensical understanding that an individual has of other people.  There are two main assumptions underlying folk psychology: First that I have some kind of privileged access to my own mind; and second, that from such access I am able to infer that the minds of other people work in the same ways that mine does.   If I threw a baseball through a window, and then see my father running towards me while yelling at the top of his lungs, my folk psychology would make me suspect that he is angry with me.  If a man tells his girlfriend that she has gained a lot of weight, the girlfriend may form the theory of mind about her boyfriend that he intends to insult her. 

When we consider this kind of theory of mind, we should note that folk psychology involves cross-cultural similarities in how one reads attitudes (i.e., in understanding that a particular facial expression represents a particular feeling) [12] .  We should also be aware that there are similarly cross-cultural differences in a person’s theory of mind about another person (i.e., due to how one is raised, or which culture one is born into).  Consider the action of slurping one’s soup or burping during dinner.   While an Asian may formulate a theory of mind about the slurper or burper that she is complimenting the chef on his great cooking, an American may formulate a theory of mind about the same slurper or burper that she is being rude and inconsiderate to her dinner party. 

The second level – theory of theory of mind – tries to understand how an individual comes to interpret and infer from the actions or observed emotions of another individual that she possesses that particular propositional attitude.  Though we may say generally that every normal-functioning human being has a folk psychology of somebody else (a theory of mind), psychologists or philosophers may question the assumption of folk psychology itself.  Is this the best way to understand how we attribute mental states to other people [13] ?  In this situation, one may say that while the theory of mind operates under a certain paradigm, the theory of theory of mind evaluates the paradigm itself, and questions whether it is the best way to understand how minds work.  Thus we can see that my presenting the Physical Stance, Design Stance, and Intentional Stance above is an examination of this issue from the level called theory of theory of mind, rather than theory of mind.   I will now continue by further exploring the theory of mind that Baron-Cohen calls mindreading. 

 

Evidence for the Intentional Stance:

Baron-Cohen claims that we can test mindreading or the Intentional Stance.  He begins with a model that includes four mechanisms for mindreading capabilities [14] .  He then presents studies that test this model by observing the actions and expressions of babies and toddlers.  If humans have a natural ability to mindread, this should be demonstrated in its most easily observed form.  In other words, it should be demonstrated in babies and toddlers.   

Simon Baron-Cohen’s Mindreading System [15]

 

The first mechanism, the Intentionality Detector (ID), utilizes the senses (such as sight and sound) in order to detect the most basic or primitive forms of goal-directedness, such as want and desire.   One also may say that this primitive goal directedness is a more primitive form of propositional attitudes.  We should note that this is not a trait unique to human beings.  Other mammals, amphibians, insects, and even amoebas, for example, are goal-directed organisms, because they have been selected for by evolution to seek pleasure or avoid pain [16] .  As a result, we can know to a greater extent that humans have ID.  Studies done on babies have shown this by their sensitivity to the changes in the behavior that adults have towards them, such as recognizing things like the difference between giving and teasing. 

The second mechanism known as the Eye-Detection Detector (EDD) utilizes the sense of sight, and signifies the ability of a person to recognize eye contact or eye movement.  Studies find that 2-month old infants have a fascination for the eyes, as if they recognize that the other person looking at them is another sentient being.  Not only do they notice the eyes more than any other part of the face, they also seem to distinguish between somebody looking at them and somebody looking elsewhere.  There is also a tendency in six-month old babies to look at somebody longer if that somebody is looking back at them, and a tendency in three-year-old children to be able to identify whether other faces are looking at them.  The game of peekaboo is a very good confirmation of this, since babies and children seem to be quite thoroughly stimulated by the sudden visual contact of eyes. 

Eye contact is important because it gives cues into a person’s mental states.  For example, it is often the case that a parent can tell when a child lies by watching the movement of his eyes.  One can say that EDD provides the means by which a person can have joint-attentional functions, or the third mechanism, called the Shared-Attention Mechanism (SAM).  Incidentally, while ID and EDD build dyadic representations of mindreading (a relationship between Agent and Object, or Agent and Self), joint-attention or SAM builds triadic representations of mindreading (a relationship between Agent, Self, and Object).   Triadic representations require that the Self concur, or at least partially concur, with the Object that is perceived by the Agent.  To better explain this, we can say that Belia’s seeing a woodchuck run into the bushes presents a dyadic representation – the relationship here is between Belia (the Agent) and the woodchuck (the Object).  However, if we considered a third person named Rep observing what Belia observes, we can say that Rep provides the triadic representation.  Hence, Rep (the Self) is able to perceive and report on what Belia (the Agent) likewise perceives, namely, the woodchuck (the Object).  Studies on children to show this triadic representation can be understood similarly in this way: by reporting the perceptions of others and, where applicable, attributing goal-directed states to them.  Through utilizing EDD (i.e., eye-contact), children are able to sense a goal (ID) in others and report a prediction of what those other people or animals will do.   By way of SAM, 3-4-year-old children have recognized that, if they perceived a person looking at Candy Bar A rather than Candy Bar B or Candy Bar C, it is likely that that person wants what she is looking at, that is, Candy Bar A. 

The fourth and final mechanism described by Baron-Cohen is called the Theory-of-Mind Mechanism (ToMM).    ToMM synthesizes the representations of the aforementioned epistemic states with those more complex forms of propositional attitudes, such as thinking, believing, knowing, and recognizing.  The unique aspect of ToMM are M-Representations, which allow for referential opacity, or the suspension of the truth values of propositional attitudes.  We can say for example, that when Belia saw the woodchuck run into the bushes, she believed that the animal that ran into the bushes was a woodchuck, even if the animal was actually a skunk.  In this case, it didn’t matter whether the animal that Belia saw was a woodchuck or a skunk; what mattered was that she believed it was a woodchuck, or that her M-Representation of the animal was a woodchuck.  This is important because it assumes that, though both Rep and Belia perceived the same object run into the bushes, Rep can predict that Belia’s attitude about that proposition is one way, even if Rep’s attitude about that proposition is another way.  In other words, by understanding that Belia can hold false beliefs, Rep is able to distinguish his own mind from Belia’s mind, and thus form an M-Representation of her mental state.  Consequently, he can mindread. 

It would therefore not be surprising that one of the most useful studies to test ToMM in children is to determine whether they are able to understand that other people can have false beliefs, and whether they are able to deceive [17] .  In one such study to test an understanding of false beliefs, children – I’ll singularize and give them the generic name Child A – are given a tube labeled “Smarties” and asked what is contained inside.  Because this tube is named after the Smarties candy, Child A originally believes (justifiably) that candy is contained inside the tube.  The experimenter then opens the tube and reveals that there was not candy, but rather pencils inside.  Later on, another child – who I will give the generic name Child B – enters the room.   Child B presumably has never been shown what is inside the tube.  To see whether Child A is able to form a theory of mind about Child B, the experimenter asks Child A what Child B thinks is inside the tube.  If Child A states that Child B thinks there is candy inside the tube, it will show that Child A can understand that Child B has an M-Representation about the tube that candy is contained therein.  Conversely, if Child A states that Child B thinks there are pencils inside the tube, it will show that Child A fails to mindread Child B.  In this case, Child A would have reported about reality rather than the belief state of Child B [18] . 

Another study testing the ability of children to deceive involves a rabbit puppet and a wolf puppet playing a game with a child.  While the rabbit puppet played the role as friend, the wolf puppet played the role as foe.  The purpose of the game was for the child to allow the rabbit puppet access to candy inside a box, and prevent the wolf puppet access to the candy by way of deception.  In order to test how the child would respond to either puppet, there was a padlock to signify that the box could be locked.  The box was purposely unlocked, so that both puppets could open the box and get the candy.  However, the puppets would not approach the box unless the child told them whether the box was locked or unlocked.  If the rabbit puppet inquired about the box, the child was to give an affirmative response and allow the rabbit access.  If the wolf puppet inquired about the box, the child was to lie and give a negative response, in order to prevent the wolf from approaching the box and thus opening it.   If these children were able to lie and deceive the wolf puppet, it would be evidence for the child’s ability to mindread [19] .

What these studies found is that the ability to ToMM is not innate and was not as clear-cut as ID, EDD, and SAM seem to be.   The Smarties tube test, in particular, found that children under the age of 3 are generally unable to recognize that others have false beliefs.  Interestingly enough, they are also generally unable to acknowledge that, before their beliefs were falsified, they themselves had previously thought that candy rather than pencils were contained inside the tube.  On the other hand, those who are older than 3 are generally able to state that this other person holds a false belief about the box that it contained candy rather than pencils.   The study seems to support a position that most normal-functioning human beings do gain ToMM naturally; however it is a developmental process in thinking that does not surface until children are around 4 years of age.  Psychologist Peter Mitchell writes, “Gopnik and Astington concluded that there is a single cognitive deficit in the young children that explains their poor performance across a variety of tests of false belief.  They asserted that at approximately 4 years of age, clinically normal children benefit from a radical conceptual shift in their thought that equips them with a representational theory of mind.” [20]

 

Analysis of Autism:

Autistics present a unique look into the ability (or lack thereof) to mindread.  People with autism suffer from a developmental disorder that affects their social and cognitive functions, which further prevents them from interacting with others in the way that normal-functioning human beings would.   For the purposes of this paper, I will not delve too deeply in explicating the details of this disorder.  I will, however, note several aspects of autism that I think are relevant to Baron-Cohen’s mindreading project.  

Baron-Cohen uses his mindreading model as the paradigm for pinpointing the source of the autistic disorder.  He finds that autistics are indeed capable of ID, citing evidence for their communicating wants and desires.   Kristen Ong, who worked as a behavioral therapist for two years at the former Verdugo Hills Psychotherapy Center (now called Pacific Child and Family Associates) in the Southern California area, supports this position by providing the following example:

I was working with a child (I’ll call him Johnny).  Johnny loved chips but wouldn’t ask for them.  He would put his hand out and whine.  I would bring chips to the session and in the middle while he was working, I would take out a chip and eat one.  Of course, he would put his hand out and as I would ignore him, he would move his hand closer and closer until it was right next to my mouth. [21]   

 

Though he had trouble communicating his desires through language, Johnny’s love for chips clearly shows that he is able to utilize his senses and act accordingly to achieve that which his wants are directed towards.  

Baron-Cohen finds that autistics can understand and interpret eye direction, and thus are not deficient in EDD [22] .  However, there are some peculiarities in the way they make eye contact with other people.  As we have already seen, eye contact and eye direction is important because it gives cues into a person’s mental states.   In her book, Autism: Explaining the Enigma, Frith mentions that a normal-functioning child tends to communicate with others by intentional pointing, for the purpose of seeking a mutual understanding about whatever the child is pointing to.  For example, a child may point to an animal in a zoo and seek his dad’s affirmation of its existence.  His dad in turn would probably reply, “Yes, honey, that’s a giraffe!  Look at the long neck!”  When children seek mutual understanding in this way, they are already assuming that these other people are other minds.  This, however, is not the case with children who have autism.  The eye gaze of autistics is often interpreted as blank, as if any person they are near is an inanimate object rather than another subject.  When autistics utilize their eyes, they do it by way of instrumental pointing rather than intentional pointing.  In other words, they point to something only if they desire it; they do not point in order to affirm a mutual understanding of it. This in turn leads to the hypothesis that children with autism initially cannot make a distinction between their own minds and the mind of another [23] .  Using his model, Baron-Cohen claims that because the level of psychological stimulation through EDD in autistics is questionable, the joint-attentional functions – SAM – in turn will be questionable. 

The speculation that autistics have a deficiency in comprehending the triadic representationalism of SAM is verified by several observations.  Two such observations are: an autistic’s speech, which tends to be too loud, too soft, or too monotone; and what Leo Kanner calls an autistic’s insistence on sameness, i.e., a resistance to changes in everyday routines.  If their routine is in some way disturbed, it is not unusual for the autistic to throw a tantrum.   These two characteristics are attributed to an oversensitivity and/or undersensitivity to multiple stimulation of the senses (the oversensitivity and/or undersensitivity varies from autistic to autistic, and from situation to situation).  Temple Grandin, a woman diagnosed with autism, seems to reiterate this by writing that the insistence is due to difficulties in synthesizing multiple sensory input, and then choosing which input needs to be attended to [24] .    Baron-Cohen argues that this difficulty in turn affects the output from SAM to ToMM. 

Like SAM, there are also a number of instances where the lack of ToMM is observed in people with autism.  However, I will only describe two.  One instance is found in another characteristic observed by Kanner, that is, autistic aloneness.  By this, he does not mean that autistics are shy and prefer to shrink back from contact with people.  Rather, he means that autistics are unresponsive to people, and likewise do not seem to care about being in contact with other humans.  Mitchell interprets this to mean that autistics appear to fail in connecting with other minds, preferring instead their own inner world [25] . 

A second instance is found in the aforementioned studies involving the reporting of false beliefs and deception.  The Smarties tube study and wolf puppet deception study, for example, indicate that children over 4 years old with autism have significantly failed to attribute a false belief to other people.  What is intriguing, however, is that blind people – who obviously would have no ability to EDD – are still able to deceive and formulate referentially opaque M-Representations about others, whereas autistic people – who do have sight and are past the mental age of 4 –consistently fail in reporting the false beliefs of others. These results in turn provide further confirmation of the hypothesis that people with autism are mindblind in much the same way that colorblind people are colorblind [26] . 

 

Anomaly to the Paradigm:

            Having explored the evidence given to support the hypothesis that autistics over the mental age of 4 initially are deficient in SAM and ToMM, I want to now examine an anomaly that may be used to refine or refute Baron-Cohen’s model.  The specific anomaly that I would like to examine is found in the autobiography of Temple Grandin.  Grandin provides a unique perspective into the mind of an autistic, because she herself was diagnosed with autism.  She has since overcome many of the difficulties faced by people with autism, and now holds a Ph.D. in animal behavior.  My intent here is to see how her experiences as an autistic child hold up to the Intentional Stance and Baron-Cohen’s mindreading model.  I do realize that Grandin is one person who provides a single anomaly in comparison to the long list of observations done by various researchers that support the autistics’ initial lack of mindreading abilities.  However, because she is one of very few people who are able to provide an inside perspective of the mind of an autistic, I believe that Grandin’s story is worth considering. 

            In her book, Emergence: Labeled Autistic, Grandin recalls a number of experiences that appear similar to those who suffer from autism.  Like most others, she had a monotone voice, avoided making eye contact with other people, and participated in obsessive behavior such as excessive spinning.  She also appeared very adverse to physical affection, such as hugging or hand-shaking.  In addition, Grandin writes that she had greatly preferred the world of her inner mind over those external to her [27] . 

Grandin’s difficulties in socializing in turn caused a variety of social problems, which made it easy for others to misconstrue her outward actions.  In grade school, she was frequently disciplined for misconstrued behavior.  When a teacher admonished her about her lack of concern for schoolwork, Grandin writes, “I did care, but my inability to express my feelings and thoughts rhythmically sabotaged my desires.” [28]   In another instance, Grandin – then in third grade – had taken a summer camp trip.  Through a variety of situations, children had introduced Grandin to a new set of vocabulary words, which she found rather fascinating.  She wrote that she enjoyed the way these words rolled on her tongue, and soon found herself obsessed with saying them, even though she did not know what they meant.  Unfortunately, these words were: “boobs”; seeing a boy’s “peter”; and “I’m ripe.”  Needless to say, she was misunderstood as “making a pass” at the other boys, and being “sexually advanced, over-sexed and unnaturally obsessed with sex.”  Within a week of the camp, Grandin had found herself in the infirmary with a urinary infection, with the nurse claiming that the infection was due to her masturbating too much [29] .  

Despite these experiences, Grandin does remember instances where her actions and intentions were dissimilar to those who suffer from autism.  She reports that she had the ability to deceive, and like any other devious child, she did it frequently.  She recalls finding creative ways to trick the “seeker” during games of hide and seek, and also did her share of damaging property.  When she was around 9 or 10 years old, she and her friend Sue decided to throw whiskey bottles at their teacher’s garden.  This, of course, caused a great amount of wreckage.  When the teacher lamented about her damaged garden to the class the next day, Grandin looked her teacher straight in the eye for the first time, and told her that she did not know who ruined the garden, but saw two other boys near her house on the same day.   When the teacher escorted these boys to the principal’s office, Grandin said she was thoroughly satisfied to see these two boys – who regularly taunted her – get in trouble.  After all, “they might have done it if they’d thought of it.”  In another instance, Grandin recalls assisting her cousin Peter in wrecking his neighbor’s lawn.  When Peter mentions that they would no doubt be caught, Temple giggled and replied, “Who’s to blame?  The dogs did it.” [30]  

One may say that Grandin’s deception examples that are anomalies do not refute the position that autistics are deficient in SAM and ToMM.  Grandin’s ability to deceive may have been because she is a high-functioning autistic, and was able to develop a theory of mind by the age of 9 or 10.  However, Mitchell also offers examples of autistic people who are able to deceive [31] . 

Moreover, it is important to consider Grandin’s explanation for her behavior by going back to her earliest childhood memories [32] .  As I had previously mentioned, one of the problems with autism is the failure to synthesize multiple sensory inputs, which in turn prevents the individual with autism from interacting with other minds in the “normal” way.  This indeed seems consistent with an autistic’s deficiency of SAM.   On the other hand, while Grandin’s outward symptoms previously led us to speculate on her deficient theory of mind; the explanations she gives for those symptoms indicate otherwise.  Grandin’s account does not imply that the sensory overload prevented her from making triadic representations; rather, it implies that the sensory overload prevented her from acting upon the propositional attitudes that were based on those triadic representations.   As a young child, she knew that her mother was there, and was talking to her.  Nevertheless, she found it really hard to make eye contact at the time because there was too much sensory stimulation for her brain to handle.  She writes, “I ached to be loved – hugged.  At the same time I withdrew from over-touch as from my overweight, overly affectionate, ‘marshmallow’ aunt,” and further, “It was as if a sliding glass door separated me from the world of love and human understanding.” [33]

In addition, the aforementioned examples regarding her misconstrued behavior do show that there were instances where she not only had difficulty expressing her propositional attitudes, but other “normal-functioning” people had difficulty interpreting her actions as representing a true set of propositional attitudes.  This interpretation of her propositional attitudes were not false in the way of reporting referentially opaque M-Representations.  Instead, those who misinterpreted her behavior had made referentially false reports of Grandin’s M-Representations.  Of course, it is not unusual for children to be misunderstood, just as it is not unusual for an American to form a theory of mind about an Asian who slurps her soup that she is being rude and inconsiderate.   We could say that we recognize that the American had failed to make a veridical prediction in her mindreading of the Asian slurper due to various cultural biases.  Just as the American had assumed falsely that the actions of the Asian represent intentions similar to Americans, we can say that these children and adults had formed a theory of mind about Grandin that her actions signified particular mental states. 

 

 

 

Decoding Difficulties:

What I want to propose is the possibility that a theory of theory of mind, like the one proposed by Baron-Cohen, may fail to understand an autistic’s mind, when theory of theory of mind theorists do not try to put aside their own beliefs that actions signify a purposeful state.  To explain, Frith considers in the second edition of her book that her original notion of a triadic representation of mind, namely, the synthesizing of multiple sensory inputs, is misguided.  She writes that while autistics lack a central coherence, or an ability to understand things in terms of a larger context or ultimate purpose, mindreading – or mentalizing as Frith calls it – does not require the large amounts of information integration that central coherence requires.  “Many able people with autism manage to acquire a theory of mind, even without the facilitating effects of a mechanism that allows intuitive mentalizing.” [34]   Frith and Baron-Cohen himself hypothesize that people with autism are much more capable of systematizing information rather than empathizing with others.  By systematizing, he means that autistics have an amazing ability to focus and even obsess over the intricate details of something, rather than the ability to look at something in terms of a “bigger picture” or greater purpose.  This is likened to the modern-day geek, who often has a very narrowly-focused interest in a particular field, yet may have little or no interest in how that narrowly-focused interest connects with other more practical areas of life, and may have a very poor understanding in social norms or social interactions.  

A good example of this may be seen in a picture located on the front cover of Frith’s book.   Here you see four people playing cards.  While the person on the left is hiding his cards, the two in the middle are looking suspiciously at that player and suspecting that he is conniving something.  The player on the right, however, seems to be oblivious to what is going on, and seems to be thinking about his next hand [35] .  According to Frith, a normal-functioning human being would interpret this picture in terms of the kind of intentional states that I have named.  We look at this and try to understand the bigger picture.   When Frith showed this picture to a high-functioning autistic named A.C., she writes:

On the cover of your book there’s a picture of some people playing cards.  I remember looking at the picture for something like an hour, figuring out how smooth the pigments of the paints the artist [used] had to be, and the quality of brushes, and how greatly developed the sub-economy of artists at that time must have been to demand that quality of painting and of reproduction of the actual textures of the fabrics in the characters’ clothes, and of course this is the most obvious thing about the painting, the high realism and the skill of the artist, and then I read inside the book, and I was like, What the hell?  There’s this whole “soap opera” that the “normal” person is supposed to pick on first, and this person cheating, and that person knows, and that other person doesn’t, etc. it’s nuts! [36]

 

The perspective given by both A.C. and Grandin thus lead us to consider whether the normal-functioning individual falsely assumes that the actions of the autistic individual represent intentions or a lack of intentions similar to normal-functioning individuals.  Any study offering evidence for a deficiency in SAM or ToMM must be considered in light of the possibility that the observer misinterpreted the actions of the autistic subjects.  In other words, we should take into account a potential difference between the way the observer interprets the actions of the autistic, and the way the autistic subject interprets his own actions. 

 

Conclusion:

            In this paper I have offered an explication of a theory into how we are able to explain the actions of other people and attribute mental states to them.  I have provided a model presented by Simon Baron-Cohen to describe how normal-functioning human beings mindread.  I have also given an account of autism, and attempted to show how there may be problems in how Baron-Cohen uses his model to explain autistics’ deficiencies in mindreading.  While the problems presented do not refute his theory, they do call into question the evidence he offers for his theory.