Educate not Train

Educate, Not Train

Why understanding the “why” outperforms instruction — and why, in the end, it comes back to energy

Joseph P. McFadden Sr.

Over the years I’ve watched a puzzle repeat itself across continents and industries. I’ve stood on molding floors in China and walked through our own labs here in the States, and I’ve seen the same thing in both places: well-trained people not following the very procedures they were trained on. Not out of laziness, and not out of defiance. These were capable, conscientious people who had sat through the training, passed the checks, signed the forms — and then, when the moment came, quietly did it their own way.

For a long time that bothered me. We’d invested in the training. The procedure was right there — written down, mandated, unambiguous. So why the gap between what people were taught and what people actually did?

Being an engineer, and being someone who can’t leave a “why” alone, I went looking for the mechanism. And, as tends to happen with me, the trail led back to energy.

It comes back to energy

I’ve argued before that all roads lead back to energy — how it’s stored, how it’s moved, how it’s let go. It turns out that’s just as true inside our heads as it is inside a stressed polymer.

Your brain is metabolically outrageous. It’s a couple of percent of your body weight and it burns something like a fifth of your energy budget. Evolution does not tolerate that kind of expense without demanding efficiency in return — and the efficiency it found is prediction. Rather than laboriously computing the world from scratch each moment, the brain runs an internal model and predicts what’s coming, correcting only when reality disagrees. Predicting is cheaper than perceiving from zero. A brain that predicts well is a brain that spends less.

Sit with what that means. At the hardware level, we are machines built to minimize effort. When a challenge lands in front of us, we don’t reach first for the rulebook. We reach first for our own internal model — the one assembled from a lifetime of lived experience — and we ask, mostly below the level of awareness, “what’s the least-effort path through this that I already trust?”

Rules land on models

Now put a training protocol into that picture.

When we train someone, we hand them a set of instructions. But that person did not arrive empty. Long before your training class, they had years — decades — of lived experience that already built and tuned their predictive model. That model is theirs. It’s earned. It’s fast, it’s cheap to run, and it has worked for them before.

So the protocol you’re handing over isn’t landing on a blank page. It’s landing in direct competition with a deeply grooved internal model that the brain, by design, prefers — because that model costs less energy to follow. Ask someone to do a task in a way that contradicts what their own experience predicts should work, and you’ve set up a contest your training usually loses. They know the rule. They ignore it anyway. Not because they’re bad at their jobs, but because that is exactly what a predictive, energy-minimizing brain is built to do.

That, I’ve come to believe, is the real reason “trained” so often fails to become “does.”

Reach the model: explain the why

Here’s what I’ve found actually changes behavior. You have to reach the model, not just the hands.

If you help a person understand the why behind the ask — the actual mechanism, the reason the step exists — something shifts. The instruction stops being an arbitrary rule competing against their experience and becomes part of their model. Now their own predictive machinery is working for the protocol instead of against it, because they can see, for themselves, why the low-effort shortcut they’d otherwise reach for leads somewhere bad.

And I want to be honest about a condition that matters: this only works if the protocol actually makes sense. Understanding is not a trick for manufacturing compliance. When you explain the why, you also expose the reasoning to daylight — and if the reasoning is thin, the person will see that too. That’s a feature, not a bug. A rule that can’t survive being understood probably shouldn’t be followed on faith either.

Let me give you a concrete one from my own world. Tell a technician “don’t wipe this area with that solvent,” and you’ve handed them a rule to weigh against their experience — and they’ve wiped a thousand parts with a thousand solvents, so their model says it’s fine. But explain why: that the part is under stress, that the solvent lowers the surface energy right at the crack tip, that the two together can split the part wide open in seconds — and now they’re not obeying a rule, they’re avoiding a failure they can picture. The next time they reach for that cloth, their own brain stops their hand. That is the whole difference between training and education.

Educated people improve the work

And there’s a dividend I didn’t anticipate when I first started down this path.

When people understand the why, they don’t just comply better — they start improving the process. Education fosters collaboration, not blind obedience. Someone who grasps the reason behind a step can see when the step is clumsy, or when there’s a cleaner way to reach the same goal. That makes them a partner in making the protocol better rather than just a follower of it — precisely because they understand what the protocol is for.

But that dividend only pays out if you build the room for it. You can’t ask people to understand and then punish them for asking questions. Education needs an environment where questioning is welcome, where the conversation runs both directions, where “why do we do it this way?” is treated as a contribution rather than a challenge to authority. Take that away and you’re right back to training — rules landing on models that quietly reject them.

That’s education

So I don’t train service personnel. I educate them. I explain the why, I stay open to the questions that come back, and I let the protocol earn its place in their understanding. It costs more up front — more time, more conversation, more willingness to have my own reasoning examined. But it’s the only approach I’ve found that actually closes the gap between what people are taught and what people do.

And it comes back, as it always seems to, to energy. We are built to spend as little of it as we can. You can fight that with rules, and lose. Or you can work with it — give the predictive brain a reason good enough to adopt as its own — and win. That is the same holistic instinct I bring to a cracked part or a failed weld: don’t just document the what. Understand the why. Then act.

A collaborative effort — decades of my own work, packaged with the assistance of Claude, Anthropic’s AI.

Combating engineering mind blindness · McFaddenCAE.com