Theory of Mind and Autism: What the Idea Really Means

By Dimitrios Kastanas

Introduction

In the last few years, public interest in neurodiversity has grown rapidly. The term neurodiversity highlights the idea that human brains naturally work in different ways. People vary in how they learn, experience the world, and understand language, and there is no single “correct” way to think, learn, or behave. In this article, we focus on one of the contexts in which it most often appears: autism spectrum disorder, or simply autism. Autism is a neurodevelopmental condition characterised by difficulties in language communication and social interaction. One influential idea in autism research is the Theory of Mind (ToM) hypothesis. This hypothesis suggests that difficulties understanding other people’s thoughts, beliefs, and intentions may explain many of the social and communication challenges experienced by autistic individuals.

The goal of this article is to explain what ToM actually is and what it means for someone to experience difficulty with it. We will leave a detailed review of the scientific evidence for this hypothesis for another article. Instead, our focus here is to help readers understand how ToM supports social and language communication, and where the differences between autistic and non-autistic individuals are often assumed to lie. It is important to note that the evidence for a clear ToM deficit in autism, one strong enough to fully explain differences in social or language communication, is mixed and, in many cases, quite weak. Still, because this remains the most influential and long-standing hypothesis in the field, it is worthwhile to lay out its basic ideas so readers can understand what it claims and why it has shaped so much of the discussion around autism.

Before we go further, it’s helpful to pause on language. You may see the phrases “autistic individual” and “person with autism”. The first is known as identity-first language, while the second is called person-first language. Current research suggests that the autistic community does not have a single, universal preference for one term over the other. Because of this, the most respectful approach is simply to ask someone which terminology they prefer when you are speaking with them.

What is Theory of Mind?

With this introduction in place, we can turn to the main question: What is theory of mind? It may seem like a simple question, but the answer is more complicated than it appears.

Originally, ToM was defined as the ability to represent the mental states of oneself and others. Interestingly, this idea first emerged from experiments with chimpanzees. In one classic setup, researchers showed chimpanzees videos of a human struggling with a problem, such as being unable to reach a bunch of bananas. After watching the video, the chimpanzees were given several pictures and asked to choose the one that would help the human solve the problem. In this example, the correct answer would be a stick long enough to reach the bananas.

The surprising finding was that chimpanzees often chose the correct solution. This suggested that they may understand goals and intentions well enough to predict someone else’s behaviour. Whether chimpanzees also understand beliefs, a more complex mental state as we’ll see below, is still an open question. We will explore the evidence for ToM in animals more thoroughly in a separate article.

Mental States and Propositions

Returning to the definition of ToM, the next question is: What does it actually mean to represent the mental states of oneself and others? And perhaps even more basic: What are mental states?

To make these ideas clearer, let’s imagine the simple situation shown in Image 1. John owns a laptop, and the laptop is sitting on the table. The statement “the laptop is on the table” is therefore true. This statement is an example of a proposition: a sentence about the physical world that can be either true or false. Other examples include “the sky is blue” (true), “whales are mammals” (true), and “I am 1.90 meters tall” (false, in my case).

Now consider the expression: “I know that the laptop is on the table.” Here, we take the original proposition (“the laptop is on the table”) and place it inside a larger structure: “I know that…” This larger structure is called a mental state frame. It acts like a slot into which a proposition can be inserted. What makes it a mental state is that it expresses an attitude, i.e., how someone relates to the proposition.

In this example, the phrase “I know that…” expresses one specific attitude toward the content of the proposition (i.e., “the laptop is on the table”). Together, the attitude (“I know”) and the content (“the laptop is on the table”) form what philosophers call a propositional attitude. A propositional attitude is simply a type of mental state.

Many familiar verbs express attitudes toward propositions, including believe, think, expect, intend, want, and fear. Each of these verbs describes a different way a person can relate to some idea about the world.

Representing Mental States

In the example above, the propositional attitude “I know that the laptop is on the table” uses the first-person pronoun “I” to show the relationship between the attitude and the proposition. Because it refers to our own mental state, this is an example of representing our own mental states.

If we return to the definition of ToM, “the ability to represent our own and others’ mental states”, we have now covered the first part. We have explained what it means to represent our own mental states: they are propositional attitudes, built from a proposition and the attitude we take toward that proposition.

False Belief and Decoupling

Let’s continue the example. Remember that the laptop belongs to John. John mistakenly believes that his laptop is on his backpack, because that is where he last remembers putting it. Here we have another proposition: “the laptop is on the backpack.” Like all propositions, it can be true or false. In this case, it is false, because the laptop is actually on the table.

Now we use a different mental state verb, believe, and a different subject, John. We can say:

“John believes that the laptop is on the backpack”.

In this sentence, we have John’s attitude (“John believes…”) toward the proposition (“the laptop is on the backpack”). Even though this proposition is false in the real world, it is true for John because he believes it.

This is a crucial idea: we must be able to separate what is true in reality from what is true in someone’s mind. The ability to treat a proposition as the content of another person’s belief, rather than as a description of the actual world, is called decoupling.

Decoupling allows you to understand that:

• In John’s mind, the proposition “the laptop is on the backpack” is true.
• In the real world, it is false, because the laptop is on the table.

Representing this difference requires understanding John’s false belief about the world. And that brings us to the second part of the definition of theory of mind: the ability to represent other people’s mental states.

Representation vs Inference

In autism research, some have argued that a theory of mind deficit prevents autistic individuals from representing their own or others’ mental states. In the example above, this would mean they might struggle to separate what is true in reality from what John believes about the laptop’s location. It could also involve difficulty representing their own knowledge, such as correctly keeping track of where the laptop actually is (on the table). We return to this shortly.

Some researchers broaden the definition of ToM to include not only the representation of mental states but also the inference of mental states.

Representing a mental state involves constructing a mental state frame, such as “X believes that…”, which expresses someone’s attitude toward some content. But this frame is incomplete on its own. It needs a proposition to fill it, for example, “the laptop is in the backpack”.

The process of figuring out which proposition belongs inside the mental state frame is called inference. Inference is the step where we determine what someone else is thinking, wanting, believing, or intending.

It is important to note that the proposition you insert into the frame is never explicitly given. You must infer it from the situation. In our earlier example, you infer that John believes his laptop is in his backpack by observing that he is searching there and cannot find it.

Example of Inference

Here is another example. Imagine you see someone running down the street. You look ahead and notice a bus stop, along with a bus approaching. Based on the context, you infer that the person is running to catch the bus. In terms of mental states, you would say:

“I believe that the person is running to the bus stop”.

In this sentence, the mental state frame, “I believe that…”, is filled with a proposition you constructed from the available clues (e.g., the bus stop, the person running down the street, the bus approaching, etc.). This process of using context to determine the right proposition is exactly what researchers mean by mental state inference.

Theory of Mind in Autism

In autism research, the original claim was that autistic individuals struggle with the representation part of theory of mind. However, a great deal of evidence now suggests that this is not the case. Representation is essentially a binary ability: either you can construct a mental state frame (“I believe…,” “John knows…”) or you cannot. Most autistic individuals can do this, which indicates that representation itself is not the problem.

For now, the current consensus is that the main difference between autistic and non-autistic individuals lies in inference, specifically, in how they determine which proposition to place inside the mental state frame.

Conclusion

The main point to take from this article is that theory of mind involves the capacity to represent and/or infer mental states. A mental state is what researchers call a propositional attitude. This has two parts:

1. A mental state frame, which expresses someone’s attitude toward a proposition (for example, “I believe that…” or “John knows that…”), and
2. A proposition, which is a statement about the world that can be true or false.

Together, these form a mental state.

In autism research, it was once argued that autistic individuals have difficulty representing their own or others’ mental states. However, a substantial amount of evidence now suggests that this is not the case. Instead, the difference between autistic and non-autistic individuals appears to lie in inference, the process of figuring out which proposition belongs inside the mental state frame.

Inference requires taking in all the available, observable information and using it to construct something unobservable: the mental state of another person, or even our own.

Throughout this article, we set aside several important questions that deserve their own discussion. In future articles, we will explore the origins and scientific meaning of the term neurodiversity, review the full body of evidence for and against the theory of mind hypothesis in autism, examine whether animals have theory of mind, and take a deeper look at the distinction between representing and inferring mental states. We will also explore in more detail how autistic and non-autistic individuals may differ in their mental-state inference strategies, and what this means for social and language communication.

Fasten your seatbelt; the road ahead is full of fascinating questions!

References

• Long, E. L., Catmur, C., & Bird, G. (2025). The theory of mind hypothesis of autism: A critical evaluation of the status quo. Psychological Review. Advance online publication. https://dx.doi.org/10.1037/rev0000532
• Premack D, Woodruff G. Does the chimpanzee have a theory of mind? Behavioral and Brain Sciences. 1978;1(4):515-526. doi:10.1017/S0140525X00076512