Asperger's Syndrome and the Understanding of "Others' Information" - Theory of Mind and Conditional Probability

Asperger's Syndrome and the Understanding of "Others' Information" - Theory of Mind and Conditional Probability

Theory of Mind (ToM) is an important concept in the discussion of Asperger's Syndrome (or Autism Spectrum Disorder (ASD) more broadly) and the differences in information held by others. Theory of Mind refers to the ability to understand the possibility that others may have different knowledge, beliefs, and intentions than oneself, and to use this understanding to predict others' behavior and to imagine others' positions.

False-belief tasks, such as the Sally-Anne test, are representative methods of measuring this ability and have been widely used to observe development in early childhood. For example, after placing marbles in a basket, one child leaves the room while the other transfers them to a box. If the first child to return is asked where to look--if he or she can answer "where the returning child believes" instead of the reality he or she sees, the child is considered to understand "the false beliefs of others.

Developmental psychology research reports that most typically developing children are able to pass this false belief task at about 4 to 5 years of age. However, a high percentage of children with ASD are unable to answer this task correctly, and have been described as having difficulty understanding that their information and beliefs differ from those of others, and as having a weak ability to take others' perspectives and predict their behavior. In fact, one classic study reported that many children with ASD failed the False Belief Task.

In recent research, however, the situation is somewhat more complex. Among the false belief tasks, there have been reports of children with ASD and Asperger's syndrome having difficulty with "explicit tasks (question-answer type)," which are susceptible to language ability and the nature of the task, but being able to pass "nonverbal/implicit" type tasks (those that require guessing others' beliefs through images or naturalistic interactions). However, there are still some cases in which children with ASD and Asperger's Syndrome are able to pass the task. Nevertheless, many findings indicate that people with ASD still have difficulty in situations that require more advanced and natural social cognition, such as "secondary theories of mind (beliefs about the beliefs of others)" such as lying, secrets, and sarcasm, and detecting social gaffes.

So why do we talk about this "inability to assume differences in information about others" in connection with the context of conditional probability and judgment theory? The parable in the lecture about the child who hides his toys illustrates the problem of "thinking only in terms of the 'truth' one knows," i.e., that "simple probability without considering information conditions" is not enough to correctly predict the behavior and risks of others. In other words, we cannot correctly predict the behavior and risks of others based on "simple probability without considering information conditions. In a situation such as cyber security, where there are human tactics and information asymmetry, who knows what and which information is shared can make the difference between success and failure. However, if we unintentionally treat others as having the same information as ourselves, then we lose the original meaning of using conditional probability. In other words, the failure in th
e false belief task was taken as an analogy for the logician's lack of understanding of probability = subjective probability.

Certainly, people with ASD and Asperger's syndrome are not uniformly "incapable of understanding others' information," and their abilities vary widely from person to person, with examples such as "weak in explicit questioning but behaving well in implicit social interactions. Nevertheless, it is reasonable to point out that in social risk judgments and complex human relationships, the ability to recognize "gaps between one's own and others' information" and to consider conditional probabilities based on such gaps is important for stable decision making and safe design.

In short, the "toy stash problem" is not just a psychological test for children, but serves as a powerful symbol of the need for metacognition that "people do not necessarily have the same information as others" and the importance of conditional probability that "probability (or prediction) changes if information is different. In situations such as social, security, AI, and policy design, making decisions based on such "information asymmetry" and "gaps in perception" is more important than mere mathematical optimization. I believe that such "information asymmetry" and "gaps in perception" are more important than mere mathematical optimization.