What Are Mental Models?

A friend cancels plans. One person thinks: "They're overwhelmed — I hope they're okay." Disappointed, but okay. Another person thinks: "They don't care about me. I knew I wasn't a priority." Hurt, angry, spiralling.

Same event. Different reaction. Why?

Because they're running different mental models.

What Is a Mental Model?

Mental models are your brain's templates for answering three questions:

• What is this?
• What does it mean?
• What should I do next?

Your brain doesn't think from scratch every time something happens. That would be too slow and expensive. Instead, it runs models — shortcuts learned from experience that compress complexity into quick predictions.

Mental Models Create Emotion by Creating Prediction

Most of your emotional reactions come not from raw facts, but from the predicted meaning your brain assigns to those facts.

Heart racing in a meeting becomes "I'm going to mess up" — not because that's what heart racing means, but because your model links them.

A short text reply becomes "They're angry at me" — not because of the text, but because your model predicts rejection.

Change prediction, change emotion. Change emotion, change behaviour. This is therapy gold.

The Hidden Cost: Models Can Become Prisons

Models become rigid when they are never tested. If you avoid the situation your model predicts is dangerous, you never discover whether the model is accurate.

This is the engine behind chronic anxiety and avoidance cycles.

The Model Loop

Here's how it works:

The coping behaviour that gives relief also prevents learning. The model stays intact because it's never tested.

How Models Form

Models are learned from experience, family norms, culture, and reinforcement. They often made sense at the time they were formed.

If conflict in your family meant danger, "conflict = threat" becomes the model. If your early performance was criticised harshly, "imperfection = catastrophe" becomes the model.

These models were adaptive once. The problem is they persist into contexts where they no longer fit.

Three Common Outdated Models

Model 1: "Conflict Means Abandonment"

Leads to: appeasing, silence, resentment, indirect communication

Prevents: honest conversations that could actually strengthen relationships

Model 2: "Anxiety Means Danger"

Leads to: avoidance, safety behaviours, shrinking life

Prevents: learning that the feared outcome usually doesn't happen (or is survivable)

Model 3: "If I'm Not Perfect, I'm Unsafe"

Leads to: procrastination, overpreparing, burnout, imposter feelings

Prevents: learning that "good enough" is usually actually good enough

Helpful Models vs Harmful Models

A model is useful if it predicts reality accurately and helps you act effectively.

A model is harmful if it predicts disaster inaccurately and shrinks your life.

The question isn't whether you have models — you do, and you need them. The question is whether your models are calibrated to current reality.

Why Common Fixes Fail

People often try to change models with:

• Arguing with themselves while activated — rarely works; the emotional brain doesn't respond to logic in that moment
• Reassurance seeking — gives brief relief, strengthens the underlying doubt
• "Positive thinking" as denial — doesn't address the model, just overlays a slogan
• Trying to eliminate anxiety before acting — keeps you waiting forever

These approaches fail because they don't provide the one thing that actually updates models: corrective evidence.

How Models Actually Update

Models update when the brain experiences corrective evidence that is emotionally believable.

This means:

• You have to actually do the thing the model says is dangerous
• You have to do it without the safety behaviours that prevent learning
• The experience has to be emotionally registered, not just intellectually understood

This is why exposure works. Not because it's suffering, but because it provides data the model can't ignore.

Safety Behaviours Protect the Model

A key insight: your coping behaviours often prevent you from discovering that the feared outcome wouldn't happen — or would be survivable.

If you always escape the party early, you never learn you could have stayed and been fine. If you always check for reassurance, you never learn you could tolerate the uncertainty. If you always overprepare, you never learn that "enough" was actually enough.

Small Tests Beat Heroic Efforts

You don't need flooding. You don't need to face your biggest fear tomorrow. You need repeated manageable tests.

The brain updates through believable, repeated experiences — not through one-off suffering. Small wins compound. Each successful test weakens the old model and builds the new one.

Examples in Practice

The "Tone = Danger" Model

Partner speaks flatly. You assume anger. You become apologetic. Partner feels confused. You interpret confusion as anger. Loop strengthens.

Test: Ask one clarifying question before concluding. Track what happens.

The "Anxiety = Warning" Model

You feel anxious before a social event. You conclude "don't go." Relief. Model strengthens. Life shrinks.

Test: Go for 20 minutes with one micro-goal. Document actual outcome.

The "Certainty = Safety" Model

You seek reassurance to feel safe. Short relief. Long-term doubt increases.

Test: Delay reassurance by 30 minutes. Notice what happens to the urge.

Micro-Experiments for This Week

1. Write one "model sentence" about your most common spiral. Make it specific.
2. Identify one safety behaviour you rely on and reduce it by 10%.
3. Run one small test: Ask one clarifying question. Stay 5 minutes longer. Speak one sentence without overpreparing.

FAQs

"Isn't this just 'stories we tell ourselves'?"
Partly, yes. But models live in emotion and behaviour. You update them through evidence, not slogans.

"What if my model is true?"
Then testing will confirm it. The point is calibration, not naive optimism.

"I tried exposure and it made me worse."
Often because it was too big, or layered with safety behaviours that prevented real learning. The model didn't update — it got reinforced. Structured pacing matters.

"I can't even identify my model."
Start with the last time you spiralled. What did it mean in the moment? Write the sentence: "When ___ happens, it means ___, and I must ___."

Mental models underpin everything in this series. Survivorship bias, anchoring, availability — they're all specific model failures. Templates that misread reality. Understanding how models work gives you leverage over all of them.

Next: a common mental model error — copying visible winners as if one success proves a strategy.