Theory of Mind (ToM) is a key feature in my condition.
From Wikipedia:
In psychology and philosophy, theory of mind (often abbreviated to ToM) is the capacity to understand other individuals by ascribing mental states to them. A theory of mind includes the understanding that others’ beliefs, desires, intentions, emotions, and thoughts may be different from one’s own.
This idea first came up to me when I learned about Autism – it is thought to be a key component in the Autism Experience.
When I read about it, it seemed easy! Of course I know other people have their own thoughts! There are some tests used such as:
- Sally-Anne Task: Assesses if a child understands that Sally will look for a marble in her basket, even though Anne moved it to a box while she was away.
Seems logical, I have a perfect score! Of course this is a test for 3 year olds. I had LLMs present me with other interactive tests but I passed all of those too.
Key Point: Mindreading
I was missing some key points from the Wikipedia article:
- Mentalizing/Mindreading: The ability to infer what others are thinking or feeling.
- False Belief Understanding: Recognizing that someone can hold a belief that contradicts reality.
- Components: Intentionality, empathy, and goal detection.
That first point is mentioned but it isn’t in the main description and I didn’t pick it up.
My Experience: Mindreading
Mindreading is critical (see NT Experience) and it turned out I didn’t have that: I was missing the ability to pick up the social signal. It wasn’t until my therapist described the experience of automatic ToM that I realized my gap. In fact this was the key piece of information that led me to uncover the mechanism: no Social Salience!
I was fooled by the online tests because they were not measuring this, they gave it away for free:
- I knew there was a problem to pick up
- they fed me the hidden information like “your friend doesn’t want to talk and is avoiding your gaze”
I can solve these easily: they are either logic puzzles (who knows what and when) or some kind of what would cause this behavior question. If I am given all the facts I can figure it out. This is different than Simulation (NT ToM) or Theory-Theory (autism ToM) – it isn’t ToM at all. It is debugging – what would this cause or what would cause this?
This is a significant difference from the autism experience – autistic people typically do have the social signal, even if it is noisy and expensive to consume. In me it is missing entirely.
My Experience: Identity of Others
I had a conversation about somebody where it was stated “I love them but I hate their actions”. This makes some logical sense to me, but in practice I don’t think that way. I think I don’t conceive of people in the same way that NT people do. (See also Identity where I have notes on my perception of self, identity and others).
Typically NT people think of each person as a “persistent object”. They store a “vibe”, historical emotional weight, and some sense of their “soul” (who are they really) separate from their actions and moods. If the person acts inconsistent with this they might be “having a bad day”. Their actions are separate from their self. There may be unconditional relationships for close friends and family. There is a slow moving average of feelings over time (hysteresis).
For me, although I recognize logically that the person has internal state, all I can perceive is their output. I can’t separate their thoughts (which I do not know) from their actions. A person is the sum of their actions (truth, utility, interaction friction). If the sum goes negative I do not like their actions and I do not like the person. A very large negative action would probably push that negative right away. I do not store any sort of vibe/emotion associated with a person. Of course I recall historical facts (e.g. they like cheese, they are good at pinball), but this is just data. A person is a Black Box to me and I attach a continuous measure of “like” based purely on actions.
So I think my conception of another person is quite different than the NT perception. I didn’t even realize this until thinking deeply about this conversation.
Additionally, I think my feelings towards others on the negative scale are fundamentally different. I might say “I don’t like X” but it is more a classification of “likely to cause harm” (maybe they lie). I might say “X is very bad” and that is a greater magnitude of harm. “I hate X” is for people that I think cause large scale harm. I don’t feel hate as a visceral emotion. I might get angry (systemic anger from harm rather than social anger from offense/shame) when thinking about people currently causing harm, but this is not the same as NT hate.
Lest you think I am some kind of killer robot, I do actually come across as pretty friendly (I think). At work I am universally praised as being very helpful and No Ego. With friends I am very Easy Going. I do have Flat Affect and may often seem indifferent because I don’t signal my emotions in a way that NT people expect. While I do not Care in the NT sense of feeling an emotional tether, I am very a loyal and good friend and am a dedicated father and husband (again, I think). I value the people in my life, but perhaps not in the way that NT people experience those same relationships.
My Experience: Predicting or Explaining Others
Despite not having ToM, I can predict the actions of others or deduce the inputs needed to convince somebody of something. This is a very high effort construction of a logical model of the person given whatever facts I know (e.g. they like cheese, they are silly, they know about computers). Given the current state of things and this model I could predict how the person would behave – specifically I could make an educated guess. I don’t have any measure of the quality, but when I want to convince somebody of something I am often successful.
My logical model is not a simulation of how the other person thinks, considering their emotions and peculiarities. NT people are (seemingly) able to simulate how others think who do not think the same way as them. My mental model of people is that they are the same as me constrained by the couple of facts that I know about them. Lucky for me I tend to think logically and even if that isn’t how the other person thinks it probably makes a reasonable argument.
This might be termed Egocentric Logical Realism:
- Egocentric: I use my cognitive architecture as the primary reference point.
- Logical Realism: I believe that because logic is a constant (like gravity), it should apply to all minds equally.
The big difference between me and people with ToM I think is that I don’t do this very often. A big presentation at work about a project I work on? Sure I will figure out what the audience needs to know. Any other situation? No idea what they are thinking and no effort put into guessing. I can just ask – this is something of a faux pas for NT people.
My Experience: Summary
I have no social salience and I completely miss the “mindreading” part of Theory of Mind. Because I lack the signal I also don’t mirror emotions or signal them in a usable form. Mostly: I can cry when sad, laugh when there is something funny and raise my voice when angered, but these are high magnitude events that trigger the physiological effects, not subtle cues.
My conception of other people is fundamentally different than NT people: I perceive people as black boxes where I attach factual data. I see only their words and actions. I don’t perceive them as bundles of emotions and thoughts separate from their actions. I don’t know how they think, I can only track facts I know about them. I don’t think about other people unless I consciously focus on them or something I see triggers a pattern match (e.g. cat videos and my daughter) – they are without presence in my mind aside from these cases.
Effectively I have no Theory of Mind – quite the opposite of what I originally thought. This is different than autism where ToM is impaired by being noisy, high lag and effortful. Mine is missing and I didn’t even know it was a thing.
Cognitive Theory of Mind vs. Functional Modeling
In clinical terms, my lack of persistent representation of people might be called a lack of Object Permanence or Object Constancy. While I know people exist when I am not thinking of them, they have no “mental presence.” When they are out of sight, they are deallocated from my active processing and exist only as a database entry in my memory.
Clinicians might label me as having “Cognitive Theory of Mind” (deliberate perspective taking) but not “Affective Theory of Mind” (feeling what others feel). However, I think even “Cognitive ToM” is the wrong label for me:
- No Social Signal: The approach most autistic people use (Theory-Theory) is still fed by a social signal, even if it’s noisy. I have no signal at all.
- Telemetry vs. Simulation: I don’t attempt to simulate a mind. I use Functional Logic Modeling to predict outputs from given inputs and only do that when required. I am reading telemetry, not mindreading.
- No Social Motivation: I have no innate desire to understand internal states. The Social Salience network usually provides this drive – without its output, I simply never realized there was anything to “understand” until recently.
- Identity as Metadata: I don’t represent people as “bundles of emotions.” They are Black Boxes tagged with facts. This isn’t a background process – it only spins up when I have a specific reason to focus on them.
From the outside this might have functional overlap with Cognitive Theory of Mind, but the mechanism is entirely different.
For what it’s worth, I neither care that I do or do not have Theory of Mind. I am not a “broken” version of an NT person, I am a functional version of myself. Understanding that I simply don’t have this hardware is the most accurate way to represent my experience.