What Mentalism in ABA Means (and When the Label Becomes the Problem)

Here is the part nobody warned you about in grad school: calling something mentalistic can itself be a stop-think. Brian Middleton, a BCBA who is also autistic and ADHD, made that point in a CEU on translation fluency, and it lands harder the longer you sit with it. Most of us were trained to spot mentalism in other people's language. Parents say a kid is "seeking attention." A teacher says a student is "manipulative." A coworker says someone is "lazy." We hear those words and we react. We label them mentalistic. And in that exact moment, we stop watching the behavior in front of us. We are doing the very thing we accuse them of doing. That same reflex is what turns into calling autistic kids "rigid," when the rigid one in the room might be the adult.

The textbook definition of mentalism in 30 seconds#

Mentalism is the habit of explaining behavior by pointing to something inside the person's head. A wish. A want. A trait. A mood. A diagnosis used as a cause. If a kid hits and you say, "He is aggressive," you just used aggression to explain the hitting. That is circular. The hitting is the only proof you have that he is aggressive, and the only reason you called him aggressive is that he hit. Nothing was learned.

Applied behavior analysis, or ABA, is the science of changing behavior by looking at what happens before and after it. ABA asks what the environment was doing when the behavior showed up. Mentalism skips that step and points at the person's insides instead. That is the textbook answer. It is correct. It is also incomplete.

Why mentalism is a problem for an observation science#

Middleton credits Ryan Sane of PsychCore for the cleanest version of this idea, and it is worth quoting him in full.

Behavior analysis is a science of observation, not a science of explanation. Now in experimental behavior analysis, there is a science of explanation component. And so there is a component of behavior analysis that is an explanation piece, but behavior analysis for applied, for applied in practice, for conceptual, for most of experimental is just about observation. From the talk — Brian Middleton

Read that twice. The applied side of ABA is about watching and describing. It is not about explaining what is happening inside a person. The moment you say "He hit because he was angry," you smuggled an explanation into a science whose whole job was to describe. That is what makes mentalism a problem. It is not that feelings are bad or that minds do not matter. It is that a story about insides cannot be tested by watching. And if you cannot test it by watching, you cannot use it to plan treatment.

So the rule of thumb stays simple. Describe what you saw. Describe what came before it. Describe what came after it. That is the data. Leave the inside-the-head story for a different science.

The catch: calling something mentalistic can itself be a stop-think#

Here is where Middleton flips the script.

So instead of engaging in the mentalism of dismissing something that somebody describes as being mentalistic. Yes. Saying something is mentalism can be mentalistic because it's a stop think. It's a, it prevents you from being able to actually observe. From the talk — Brian Middleton

Think about what happens in your body when a parent says, "He just wants attention." You feel the urge to correct them. You think the word "mentalism." You file the parent under "does not get it." You move on. You stopped watching. You stopped listening. You stopped asking what the parent actually saw that made them land on that word.

A stop-think is anything you say or think that ends the work of observation. "Mentalism" can be one of those words. So can "non-compliance." So can "manipulative." Once the label drops, the looking stops. The label feels like a conclusion. It is not. It is a shortcut your brain took so it would not have to keep paying attention.

You can be a strict behaviorist and still fall into this. The shape of the trap is the same. A word arrived and saved you from observing.

What to do instead of throwing the M-word at a parent or teacher#

Two moves.

First, translate before you correct. When a parent says "he wants attention," repeat the part that is observable. "So when he does X, you come over. And it sounds like that pattern is happening a lot." You just turned a story about insides into a description of a pattern. You did not call them wrong. You also did not agree with the cause they assigned. You showed them another way to look at it.

Second, model the function-based version instead of policing the language. Middleton puts it this way.

We're shaping culture to be more function-based rather than mentalistic. We need to stop with the, like, that person is crazy and all that other stuff. But how do we stop that? Well, we don't stop it. We create differential reinforcement. We show people that, hey, this is better. This is a much more effective approach. From the talk — Brian Middleton

Differential reinforcement here means you reward the talk you want to hear and let the talk you do not want to hear go nowhere. When a teacher says, "He was just being defiant today," you do not lecture. You ask, "Walk me through what happened right before." When she tells you, you respond with curiosity. When she lands on a function-shaped observation, you light up. That is the reinforcement. Over time, that is what changes how she talks about the kid.

How this shows up when staff describe a kid as rigid or attention-seeking#

This is the part where the page stops being academic and starts being personal.

When a behavior technician says, "He is rigid," she usually means he got upset when something changed. She used a trait word to wrap a moment. The trait word makes the moment feel like the kid's fault. It also makes the moment feel permanent. Neither of those is data. The data is that something changed, and the kid responded in a way that was hard to be around.

When a teacher says, "He is just seeking attention," she usually means he did something loud and then someone came over. The seeking word adds a motive. The motive is a story. The pattern is the data.

You can hold both sides of this. You can take the staff member's observation seriously and still not import the explanation she taped to it. The skill is hearing the part she saw, dropping the part she made up, and feeding the part she saw back to her in a way that helps her see it next time without the story attached.

Middleton closes the loop on what that staff-side rigidity language does when it gets pointed at the kid.

I find it absolutely hilarious that a lot of times people talk about how autistics are rigid. And yet I noticed that it's not autistics are rigid. It tends to be other people who are rigid. And the only time we're rigid is when we're under threat or struggling. From the talk — Brian Middleton

That reframe is the whole social-validity payoff of this page. Rigidity is not a trait you find in a person. It is a pattern that shows up when a person is under pressure. If the kid is rigid every Tuesday at 1pm, the question is what shows up on Tuesday at 1pm. The kid is telling you something about the room. You just have to be willing to keep watching after the label lands.

A two-sentence script for the next time it comes up in a team meeting#

You do not need a speech. You need two sentences you can say without getting your back up.

One: "That is a pattern I want to look at more closely. What were you seeing in the five minutes before that?" Two: "Let me write down what happened, and then we can talk about what might be keeping it going."

That is it. No "that is mentalistic." No correction. You moved the room from inside-the-head talk to outside-the-head talk, and you did it without making anyone wrong. You also kept yourself honest. You did not get to short-circuit observation either.

Frequently asked questions#

Is talking about feelings always mentalistic?

No. You can talk about feelings as observable patterns. "She started crying, her shoulders dropped, and she stopped responding to questions" is a description. "She was sad" is a guess about insides. The first one is usable. The second one is a story. Use feeling words when they help a parent or a coworker hear you. Just do not let those words replace the watching.

Can a BCBA use the word "want" in a treatment plan without being mentalistic?

Yes, with care. If "want" is shorthand for a mand, which is a request, say that. "When the learner mands for the iPad" is observable. "When the learner wants the iPad" leans on insides. In a parent-facing summary, "want" is often the right word because it helps the parent track what is happening. In the technical write-up, the request is what you saw, so write the request.

How do you redirect a teacher who keeps describing behavior as defiant?

Do not start with the word. Start with the moment. Ask her to walk you through the last time it happened. As she talks, write down what came before, what the kid did, and what came after. Read it back. Most of the time, the teacher will see the pattern on her own. That sticks better than a definition would.

One more thing before you close this tab#

The hardest part of this skill is the part where you have to keep watching after you already have a word for what you are seeing. The word is not the work. The watching is the work. Middleton's whole talk is an hour of him modeling that, with seatbelt beeps and Jurassic Park and a baby in a back seat. It is worth your hour.