“Look at the doggy!”
Piaget’s classic Cat vs Dog example provides a useful way of thinking about how people learn.
A toddler sees a dog for the first time and creates a simple schema:
“Four legs + fur + tail + woof = Dog.”
Later, they see a cat and call it a dog. This is assimilation: fitting new information into an existing mental model.
Then the cat meows and climbs a tree. The old schema no longer works, creating disequilibrium. The child adapts by creating a new schema for cats. This is accommodation.
Over time, learning becomes less about memorising rules and more about refining mental models.
The “Grey Zone”
Now imagine the child encounters a COG.
It has four legs, fur, and a tail. It looks a bit like a dog and a bit like a cat.
Some children will think:
“It’s probably closer to a dog than a cat, but I’ll remain open to evidence that suggests otherwise.”
Others may be less comfortable making that judgement until they’ve seen more evidence.
Another group might not recognise the hybrid animal and try to force it into either the dog or cat category.
Neither approach is necessarily right or wrong. They simply reflect different levels of experience and confidence when dealing with uncertainty.
The important point is that learning doesn’t just involve building schemas. It also involves learning how to apply them flexibly when a situation doesn’t fit perfectly into any existing category.
Driving Is No Different
Learners build driving schemas in exactly the same way.
Take a right turn at a T-junction:
Base Schema
Position near the centre line, drive up to the give-way line, and wait for a safe gap.
Schema Refinement #1
A lorry approaches from the left and turns into your road. Suddenly the base schema no longer works. The learner discovers they need to hold back to allow for the vehicle’s swept path and trailer swing.
A new schema begins to form:
“Large vehicles sometimes need extra space, so I may need to hold back.”
Schema Refinement #2
Later, a lorry approaches from the right.
The learner initially tries to apply their new “lorry schema”, but discovers that doesn’t fully work either. This time the cab may need to use more of the learner’s side of the road, requiring a different response.
The schema is refined again.
Generalising Beyond the Original Situation
What’s particularly interesting is what happens next.
Imagine both learners later encounter a bus straddling two lanes on a roundabout.
Neither learner has seen this exact situation before.
Learner A recognises similarities with the earlier lorry examples. The vehicle is large, requires additional space, and cannot reasonably be expected to remain neatly within lane boundaries throughout the manoeuvre. Although the setting is different, the underlying principle feels familiar.
They haven’t developed a specific “bus-on-a-roundabout” schema, but they can draw upon their previous experiences with large vehicles and adapt those schemas to the new situation.
Learner B may find that much harder.
To them, a bus on a roundabout and a lorry at a T-junction appear to be completely different situations. Because they haven’t yet developed a specific schema for buses on roundabouts, the previous lorry experiences may not feel like relevant data points. As a result, the situation can feel entirely novel, even though many of the underlying principles are the same.
Neither learner is necessarily better than the other. The difference is that one learner is able to generalise from existing schemas, while the other may need additional experiences before recognising the common pattern.
This is one reason why exposure to a wide variety of edge cases can be so valuable. The goal isn’t to build a separate schema for every possible scenario. Rather, it’s to help learners recognise the deeper similarities that connect seemingly different situations.
Where Difficulties Often Arise
The challenge is rarely the situations learners have already experienced.
The challenge is the situations that sit between existing schemas.
A learner may have schemas for:
- Normal T-junctions
- Large vehicles from the left
- Large vehicles from the right
Then they encounter a situation that shares features of all three, but doesn’t perfectly match any of them.
Without enough experience of similar examples, there’s a natural tendency to force the situation into the nearest existing category.
This is often where hesitation, uncertainty, or delayed decision-making occurs. Not because the learner lacks knowledge, but because they are still developing the flexible schemas needed to handle novel situations.
Building an Edge-Case Video Library 🎥
Most lessons naturally expose learners to common scenarios. Unfortunately, we can’t guarantee that unusual situations will occur during a lesson.
When they do appear, they’re fantastic learning opportunities. When they don’t, we’re often left discussing them theoretically.
That’s why I’m building a dashcam video library of real-world edge cases.
The idea is simple: pause the clip just before the critical moment, ask “What happens next?”, and discuss how the learner might adapt their existing schemas to deal with the situation.
The aim isn’t to create a “bad driving” compilation. YouTube already has plenty of those.
The aim is to build a library of genuine, unusual, and thought-provoking situations that help bridge the gaps between existing schemas.
A Real-World Edge Case
Here’s a recent example from one of my own journeys. Unfortunately, this was not during a lesson – what a valuable experience that would have been! However, I saved the Dashcam footage and replayed the video during a lesson.
We approached a signal-controlled junction where:
- The main traffic lights were showing green.
- Temporary traffic lights at the same junction were showing red.
- A sign beneath the temporary lights said, “When red light shows STOP here.”
- A highways maintenance vehicle was parked nearby.
My learner’s response was immediate:
“I’d stop for the red temporary light.”
Which is entirely understandable. After all, that aligns with a schema that has served them well throughout their driving experience:
Red light = stop.
My own thought process was rather different.
I wasn’t convinced either set of traffic lights could be accepted at face value because the situation itself appeared contradictory.
Based on previous experience, my working hypothesis was that:
- The roadworks had probably just finished.
- The main traffic lights had been switched back on.
- The temporary traffic lights had not yet been decommissioned.
- Turning the temporary lights off before restoring the main signals could have caused congestion during the changeover.
I also considered something else:
If I could see conflicting signals, drivers approaching from other directions might be seeing conflicting signals too.
In other words, one driver could be obeying a green temporary light while another obeyed a green permanent traffic light. If both assumed their signal was unquestionably correct, the potential for conflict was obvious.
So rather than treating the situation as a straightforward temporary traffic light scenario or a straightforward signal-controlled junction, I treated it as something in between. My conclusion was to proceed cautiously and effectively treat the situation as a give-way, ready to stop if necessary.
What’s interesting here is that neither approach was irrational.
My learner was applying an established schema that is correct virtually all of the time.
I wasn’t applying a different rule. Rather, I was drawing upon several previous experiences that appeared partially relevant:
- Roadworks being removed.
- Temporary traffic lights being decommissioned.
- Contradictory traffic control information.
- Drivers encountering conflicting priorities.
None of those schemas perfectly matched the situation in front of us, but each contributed something useful.
This is very similar to the Cat, Dog and COG example above.
Faced with a COG, some people are comfortable thinking:
“It’s probably closer to a dog than a cat, but I may need to revise that judgement if new evidence appears.”
Others prefer to gather more evidence before committing to a category.
The traffic-light junction felt like a driving equivalent of a COG. It wasn’t quite a normal traffic-light junction, and it wasn’t quite a temporary traffic-light junction either. It sat somewhere between several familiar schemas.
The key learning point is that experienced drivers don’t necessarily possess a separate schema for every possible situation. More often, they have developed enough experience to combine and adapt existing schemas when faced with something novel.
For some learners, that sort of reasoning develops relatively quickly. Others benefit from seeing several similar examples before they begin to form a dedicated schema for handling ambiguous or contradictory situations. Neither approach is inherently better; they simply reflect different stages and styles of learning.
How You Can Help
If you have dashcam footage showing an unusual, ambiguous, or difficult-to-categorise situation, I’d love to see it.
Once compiled, the library will be freely available for members of this group to use with their own pupils.
Let’s build a genuinely useful teaching resource together. 🚗💨
(Content is created from my own thoughts and academic knowledge, refined and proofread using AI.)
