What If Confusion Is the Moment We Begin to Think Differently?

When what has always worked stops working, perhaps the problem is not the body — but the model it is using to move through a new territory

Imagine someone looking at a few lines of code.

For the computer, there is no problem.

The code is syntactically correct.

It can run.

But for the person trying to understand it, something feels strange.

The structure suggests one interpretation, while the actual behavior of the code requires another.

The programmer looks again.

Perhaps goes back a few lines.

Mentally reorganizes what they have just read.

And tries to build a different interpretation.

It was precisely this small moment of confusion that Annabelle Bergum and colleagues investigated in a study published in Scientific Reports in 2026.

The researchers asked a specific question:

what happens in the brain when a programmer encounters code that is correct for the machine, yet potentially confusing for a human being?

Their findings open a much larger question for BrainLatam.

Perhaps confusion is not only what happens when we do not know.

Perhaps, sometimes, it is the moment when what we already knew is no longer enough.

Correct code that can still be confusing

Bergum and colleagues worked with what programming research calls “atoms of confusion”: small code patterns that function correctly but whose structure can lead programmers toward mistaken interpretations.

The researchers presented Java code snippets in confusing versions and in equivalent versions that were easier to interpret.

Participants were allowed to read the code naturally instead of receiving each element in isolation.

That detail matters.

During natural reading, each person reaches the problematic section at a different moment.

Simply aligning EEG to the moment the code appeared on screen would therefore not reveal exactly when the participant encountered the confusing information.

The researchers combined EEG and eye-tracking.

Eye movements allowed them to identify the participant’s first fixation on the critical region of the code.

From that moment, they calculated fixation-related potentials — FRPs.

Twenty-four participants examined code snippets with or without these “atoms of confusion.”

The behavioral results already showed a clear difference.

Confusing code took longer to understand, led to fewer correct responses, and was less often judged as easy.

But something was also happening in the EEG.

About 400 milliseconds later

When participants first looked at the region containing the confusing element, the researchers observed a significantly larger late frontal positivity, particularly between approximately 390 and 660 milliseconds, compared with the clearer equivalent code.

The researchers noted an interesting similarity with phenomena described during natural language comprehension.

Unexpected but still plausible words within a sentence can require a person to update the mental model they were building about the situation.

The authors suggest that something similar may happen with code.

The information is not impossible.

It is not simply an error.

But it no longer fits the model being used up to that moment.

The model has to be updated.

And this is where the conclusion of the study ends.

From here, our question begins.

What if confusion is necessary for us to notice that the model has changed?

In everyday life, we often treat confusion as a problem.

The person who knows appears confident.

The person who does not know is confused.

But perhaps there are different forms of confusion.

There is confusion because information is missing.

There is confusion because something is incomprehensible.

And perhaps there is a third kind:

the confusion that appears when we encounter a world in which our previous model no longer works as well as it once did.

In that case, being confused would not necessarily mean failing.

It could be a signal:

“Stop. The movement you had prepared may no longer be sufficient for what you have encountered.”

And this question extends far beyond programming.

The migrant and the four Biomes

Imagine a migrant undertaking a long journey.

First, they need to cross a vast forest.

In that territory, a particular repertoire may be extremely valuable.

A machete.

Knowledge of pathways.

Appropriate clothing.

Specific ways of finding food.

Knowledge transmitted through culture.

Ways of interpreting sounds, plants, danger, and orientation.

The body learns that territory.

Then the traveler reaches a region of great lakes.

The machete still exists.

But now a boat may become far more important.

What was indispensable in the previous territory may no longer occupy the center of what is possible.

Then comes the desert.

The boat that allowed passage across the lake may now become a burden.

Different clothing becomes necessary.

A different way of organizing water.

A different rhythm of movement.

A different way of recognizing danger.

Then come the Andes.

Once again, the territory changes.

Altitude.

Slope.

Temperature.

Distance.

The way the body must walk.

And the body moving through these landscapes is not abstract either.

It has an age.

Strength.

History.

Limitations.

Learning.

A child, a young adult, and an older person do not necessarily cross the same territory with the same repertoire of possible movements.

And we do not carry only material technologies.

We also carry:

language.

culture.

beliefs.

memories.

forms of cooperation.

ways of interpreting danger.

ways of recognizing paths.

These too are tools built throughout the history of a Body-Territory.

A tool can stop being a tool

This produces a fundamental question:

when does what helped us survive in one territory begin to make it harder to cross the next?

The machete did not become “wrong.”

The boat did not become “wrong.”

The previous repertoire was not useless.

It had a history.

It had a function.

The problem appears when a tool built for one territory continues to be used as if the territory had never changed.

Perhaps something similar happens with certain beliefs.

Habits.

Strategies.

Identities.

Mental models.

Ways of working.

Ways of teaching.

Ways of governing.

What solved a problem in the past may become insufficient when reality changes.

And perhaps confusion is one of the first signals of that mismatch.

Where does Zone 2 enter?

Bergum and colleagues did not study Zone 2.

The frontal positivity observed in their study is not a marker of Zone 2.

This distinction is important.

Zone 2 is a conceptual hypothesis developed by BrainLatam.

We use the term Zone 2 to describe a condition in which the organism can create space for Fruition, Metacognition, and critical reflection, becoming able to perceive what it is doing before simply reproducing the same movement again.

The question raised by the study is:

when an expectation fails, can we use that moment to perceive the model that is failing?

Because there are at least two possible responses to confusion.

One is to insist.

To try to force the new territory to fit the old model.

The other is to ask:

“What am I using to interpret this?”

That second question is deeply metacognitive.

It does not ask only:

“What is happening?”

It also asks:

“How am I trying to understand what is happening?”

Confusion is not automatically transformation

Another distinction is necessary.

Encountering something unexpected does not mean that a person will necessarily reflect on it.

We can reject the information.

Ignore it.

Become paralyzed.

Quickly return to the previous model.

Search only for information that confirms what we already believed.

Therefore:

confusion is not Zone 2.

But perhaps it can create an opportunity for Zone 2.

The moment when a model stops perfectly explaining experience may open a small window:

keep repeating?

or observe?

The Latin American question

Latin America may know this problem particularly well.

We often use economic, educational, urban, technological, or political models built in other territories and then ask why they do not produce exactly the same results here.

But what if the question needs to be reversed?

Can this model perceive the territory in which it is trying to operate?

Can a technology designed for a large metropolitan area understand an Amazonian community?

Can a policy built mainly around income recognize radically different territorial conditions?

Can a school organized around a single language recognize other ways of producing knowledge?

Can a city designed for one kind of body adequately consider children, older adults, people with disabilities, or people who travel long distances every day?

Perhaps some of our greatest institutional confusions appear when we attempt to cross a new Biome carrying only the tools of the previous one.

Where the article ends and BrainLatam begins

Bergum and colleagues showed that confusing code snippets, although valid for the computer, produced greater behavioral difficulty and a late frontal positivity following the first fixation on the critical region.

The authors relate this phenomenon to the need to update a situation model when encountering information that is unexpected but still plausible.

Up to this point:

Bergum and colleagues.

From this point onward:

BrainLatam.

Our hypothesis is not that confusion is necessarily consciousness or metacognition.

Our question is:

can realizing that our model has stopped working create the conditions necessary for another movement to become possible?

And then we can return to experimentation.

What happens when the participant successfully resolves the incongruity?

What happens when they persist with the previous model?

What happens when they are given time to reflect on their own error?

Could EEG and eye-tracking distinguish these moments?

Does the physiological response change when someone merely corrects the problem compared with when they can explain why their previous interpretation failed?

Perhaps we can study not only the moment when an organism encounters something unexpected.

Perhaps we can also study the moment when it transforms that strangeness into a new way of perceiving.

Because intelligence may not mean carrying the best tool.

Perhaps it means recognizing when that tool is no longer the best one for the territory in front of us.

And that may be true for code.

For a migrant.

For a culture.

And perhaps for a State as well.

Neuro Challenge Latam

Imagine that tomorrow you must cross a territory completely different from the one in which you learned how to live.

What knowledge would you carry with you — and what would you need to be willing to leave behind in order to perceive the new movements that territory requires?

Scientific Reference

Bergum, A., Maurer, A.-M., Peitek, N., Bader, R., Mecklinger, A., Demberg, V., Siegmund, J., & Apel, S. (2026). Fixation-related potentials reveal that confusing program code elicits a late frontal positivity. Scientific Reports, 16, 16833. https://doi.org/10.1038/s41598-026-50946-9