When I was at secondary school, I quickly discovered that I didn’t have a natural flair for mathematics. Finding the subject incredibly abstract, I frequently struggled to keep up. This was the exact opposite of my experience in other subjects, making the struggle with maths all the more painful.
Each year I had a different teacher. To them, everything was clear and simple. They would stand at the blackboard, chalk in hand describing the steps you had to work through. But not one of them ever properly explained the concepts or the thinking behind those steps; or the possible mis-steps that might catch you out; or the variations or alternatives that you might consider.
No doubt all that thinking was happening in the teacher’s head but it was never shared with us poor students.
Last week, I was reminded of those frustrating moments in the maths class when I came across a paper from the early ‘90s, which was all about something called cognitive apprenticeships – a term that I don’t think I’d heard before.
The strange invisibility of thinking
Most people, I think, are familiar with the concept of an apprenticeship. As an apprentice, you watch someone do the job. You try parts of it yourself. You get help when you get stuck. Gradually, you take on more responsibility.
Crucially, you don’t just learn the steps and the outcomes—but also the thinking behind them. How decisions are made. How problems are approached. How judgement is applied.
Over time, this apprenticeship path to learning a skill has become less common. Not surprising when you consider that many of the important workplace tasks people are required to do these days are largely cognitive - meaning that much of the work people do is no longer visible.
And when something isn’t visible… it’s very easy not to teach it. As I discovered in my maths lessons.
What we tend to teach instead
A well-designed piece of learning is going to focus on key tasks and procedures and the supporting information that sits around them.
But there’s another layer underneath all of that. The layer that actually determines whether someone can really use what they’ve learned. I’m talking about things like how to:
approach a new problem
decide what to try first
recognise when something isn’t working
adjust and try again
There’s nothing difficult or mysterious about any of this. In practice, these are just things an expert would do.
But if those ‘things an expert would do’ are never made explicit, expert application can end up feeling mysterious and complex. Always out of reach unless you possess some kind of ill-defined magic quality.
The inert knowledge problem
And this invisibility of thinking tends to create a pattern that’s probably very familiar to you. Learners complete a programme. They perform well in structured exercises. But when they face a slightly different situation after the training, they flounder.
Not because they can’t do anything. But because they feel unsure about what to do outside of the specific parameters that are already familiar.
One way to start fixing that pattern is to provide more extensive and more varied practice activities in the learning environment. This provides the opportunity to make some of that expert thinking more visible and more explicit.
What experts actually do (but rarely show)
If you look closely at expert performance, a few things start to stand out. Experts don’t just apply knowledge. They:
try simple cases to get a feel for a problem
generate multiple possible approaches
evaluate which path is worth pursuing
notice when they’re not making progress
step back and reframe
None of this is particularly dramatic. But it is something they do constantly (and consistently.) Most significant. It’s usually internal. Which means unless someone deliberately brings it to the surface, learners never really see it.
The illusion of smooth expertise
Which can lead to another subtle outcome. When we cover the acquisition of skills and knowledge in a learning environment, we often present it as something that is clear-cut and easy to apply.
A worked example. A simple explanation. A finished answer. Which can create the slightly misleading impression that to be an expert you must:
know what to do immediately
follow a clear path
arrive at the answer efficiently
But if you watch real experts at work, you see something else entirely. False starts. Dead ends. Moments of uncertainty. In other words, thinking that looks much closer to struggle than certainty.
And that matters. Because if learners never see that part, they’re likely to interpret their own uncertainty as failure, rather than as part of the process.
Bringing thinking into the open
So, what does it actually mean to make thinking visible? At a very practical level, it involves doing things like:
Thinking out loud while solving a problem
Showing how you decide between different options
Explaining why you abandon one approach and try another
Highlighting the moments where you’re unsure
Not just: Here’s how it works. But: Here’s how I’m figuring this out.That shift is small but quite significant. Because it gives learners something they rarely get: a model of how to think, not just what to do.
Support that adapts (and then disappears)
The other important piece to this, of course is if thinking is made visible, learners can begin to try it themselves. And as it’s unlikely that they’ll be able to do all of it from the word go, some targeted, responsive support will definitely be needed. Things like hints, prompts, partial solutions and questions that guide attention.
Over time, that support can be reduced as the learner does more of that thinking independently.
Why context matters more than we think
There’s also a broader issue at play. In the traditional apprenticeship model described earlier, tasks are naturally situated in context. You can see why something matters. You can see where it’s used.
In many learning environments that connection is weaker. Tasks are abstracted, over-simplified and detached from real use. Which makes it harder for learners to:
understand relevance
see how parts connect
transfer what they’ve learned
So even when thinking is made visible it also needs to be anchored in something that feels real.
This isn’t a rejection of teaching
At this point, it might sound like an argument against structured teaching. It isn’t. Teaching still matters enormously. But in many cases, it won’t be sufficient on its own.
Because understanding something conceptually is not the same as being able to use it in a variety of situations and contexts. And the gap between those two is often where learning breaks down.
We are always thinking about how to design better learning. But there’s another way to look at it. How do we help people see and practise the thinking that sits behind performance? That’s a slightly different problem. And it leads to slightly different design decisions.
A subtle shift in emphasis
If you start to look at things more from this perspective, a few things start to change. You might, for example:
Spend less time polishing explanations
Spend more time highlighting decision-making
Design activities that require judgement, not just recall
Build in moments where learners have to articulate their thinking
Show imperfect examples, not just ideal ones
None of these are dramatic changes. But together, they shift more focus on to building learner capability.
Where this leaves us
If you step back, there’s a slightly uncomfortable realisation here. Many of the things we care about most in learning design, depend heavily on processes that are internal, tacit and rarely taught directly.
Which means we can design excellent learning and still leave a crucial part of performance untouched.
A different question to end on
We always ask: “What do learners need to do and know?” But perhaps a more useful question is: “What do they need to see—that we’re currently keeping hidden?”
Because once thinking becomes visible it becomes something we can teach and something learners can actually use.
Until next time…
Andrew