Where Your Attention Goes
Importance in Learning
There are two genuinely different kinds of learning. One is retrieval: the name of a cell organelle, the capital of Peru, the date of a battle. The other is doing: shooting a hoop, writing a program, discussing Shakespeare, solving 13×28, playing the violin. These use different memory systems and require different things from a teacher. Bloom’s Taxonomy, for all its tiers, never quite escapes the first category. It’s retrieval dressed up as thinking. Almost everything I care about as a teacher, a parent, a friend, and an employer is in the second.
Let’s talk about learning.
Have you ever seen one of those feel-good teacher movies? Stand and Deliver. Coach Carter. Dead Poets Society. Dangerous Minds. School of Rock.
What technique does the teacher actually use?
It’s never better pedagogy. It’s never clearer explanation or smarter curriculum sequencing. Every single time, the move is the same: I will make you care about this. The whole drama is a motivation drama. The teaching is easy once that’s solved. Hollywood figured this out even if education policy hasn’t.
I’ve taught across wildly different philosophies: Montessori, Engelmann’s Direct Instruction, unschooling, traditional classroom. These approaches contradict each other philosophically. Some are opposites. And yet they all work, when they work, at roughly similar rates.
What do they share?
Not method. Not philosophy. Not even the quality of instruction. What they share is that each one, through completely different social machinery, produces a student whose attention is actually on the problem. The mechanism differs. The result is the same.
Motivation is the word people use here, and it’s mostly right. But an old mentor pushed back on me once: the kid who comes to school hungry isn’t unmotivated. He has a preference (food) that’s crowding out everything else. The scarce resource isn’t motivation exactly. It’s free attention. And motivation is almost the only lever you have on where free attention points.
Almost. Patience counts. Conscientiousness counts. Feeling safe enough to be wrong counts. And it’s worth noting that girls consistently outperform boys in school, which is also largely a motivation story. Girls want to do what the teacher wants more than boys do, on average. So they do.
But motivation is the dominant lever, and in my experience it accounts for roughly 90% of the variance in whether learning happens at all. This is true even for pure memorization. Every sports junkie you’ve ever met has ten thousand statistics loaded in his head, acquired through zero flashcards. Interest did that, not method.
Here’s a number that should bother you.
I’ve taught the path from “I can count” to the doorstep of algebra, following David Greenberg’s Sudbury Valley model, in about one hour a week over a school year. Student-driven pace, motivated kids, one knowledgeable adult who knows both the math and the teaching.
That same path, in conventional school, takes eight years. And fails more than half the students.
This doesn’t mean motivated kids learn 400 times faster. It means something more uncomfortable: those eight years aren’t slow learning. They’re mostly not learning at all. Seat time, partial memorization, tested and forgotten, no transfer, no ownership. The motivated student in an hour a week is actually learning. The comparison group is largely going through motions. You’re not comparing two speeds of the same thing. You’re comparing two different phenomena that share a name.
A kid shows up to a programming class on a Saturday morning. Unschooled. No formal math past basic arithmetic. Never really gotten algebra.
He wants to build a program that generates fractals.
By Tuesday, he’s showing me pictures.
What happened: he needed to understand complex number arithmetic, genuinely hard mathematics, and so he learned it. Not because I assigned it. Because he needed it right now for the thing he was building. I helped when he asked. The feedback was immediate and visual: the fractals appeared, or they didn’t.
He didn’t fail algebra because he was bad at math. He never learned it because he’d never had a reason to.
Think about your own favorite teacher. Not the most credentialed. Your favorite. The one who changed something.
I’d bet they were magnetic. Funny, or passionate, or strange, or all three. What they did, whether they knew it or not, was an Aikido move. Look at me, I’m interesting. And once you were looking: now look at this thing I’m looking at. Isn’t it interesting too? They used their own aliveness as a redirect. Not toward themselves, but through themselves, toward the subject.
I was young, energetic, and not ugly when I started teaching. I leveraged that shamelessly. It works. A big personality gives you the ability to point attention, and if you’re pointing it at something genuinely worth seeing, that’s not a trick. That’s teaching.
People cite Ericsson’s mastery research for “10,000 hours of practice.” What almost never gets quoted is the word he put in front of practice: deliberate. I’d use a different word: attentional.
The hours aren’t the point. Whether your attention is genuinely on the problem is the point.
I was nine years old, practicing violin under parental pressure for years. I was doing attendance, not practice. The hours looked identical from the outside. My playing reflected that faithfully, to the detriment of anyone within earshot.
Attentional practice requires two things. First, motivation to point your attention at the problem at all. That’s the 90, and we’ve just covered it. Second, something worth pointing your attention at: a feedback system honest and fast enough that your attention actually gets signal when it arrives. That’s the 9.
Two axes determine feedback quality, and most instruction fails on both simultaneously.
The first is speed. The gap between attempt and correction is almost everything. A basketball player shoots and knows within a second. A kid doing homework gets it back on Friday with red marks. Homework research largely confirms the obvious: the standard version (attempt on Tuesday, correction on Friday, move on regardless) barely registers. The exceptions are telling. Answers in the back of the book let a motivated algebra student check within seconds. A parent at the table catching a mistake mid-step compresses the loop to near-real-time. Those versions are worth something. The standard version isn’t.
The second axis is more important. The best feedback doesn’t come from a person at all. It comes from the problem domain itself. The ball goes in or it doesn’t. Montessori’s long rods line up or they don’t. They don’t tell you that 3+4=7; they make it so that if you don’t grasp it, they won’t behave the way you expect. The fractal renders or it doesn’t. No agenda, no judgment, no social weight, available instantly and infinitely. The world just tells you whether your mental model is correct.
Person-feedback (grades, comments, red marks, even a parent’s correction) is always slower, always socially loaded, always rationed. Better than nothing. Categorically different from world-feedback.
There are blurry cases. A live discussion of a Shakespeare sonnet looks like person-feedback, but it isn’t quite, because the goal is to move the people in the room. Their reaction is the domain responding. You’re not being evaluated by an authority. You’re getting signal from the thing you were trying to affect. Same structure as the hoop, softer physics.
The design challenge, Papert’s challenge, Montessori’s challenge, the one I’ve been chewing on for decades with a half-built algebra manipulative software, is that most of what’s worth learning doesn’t come with a hoop attached. The teacher’s job, after solving the motivation problem, is to build or find one. To construct the conditions where the domain itself tells the student whether they’ve got it right, fast enough to adjust, try again, and build a real mental model rather than a performance of one.
The remaining 1% is instruction in the traditional sense: explanation, demonstration, telling. It matters, but only when motivation and a feedback system are already present. And what it’s doing in that context is specific: building a mental model rather than a procedure.
Fractions taught as pizza slices (this slice is bigger than that one, which is why you can’t add 1/5 and 1/2 without normalizing into tenths) give a student something that composes. It connects to what they already know and becomes a foundation for what comes next. “To divide fractions, flip and multiply” gives them a rule that works until it doesn’t, with no foundation underneath when algebra arrives.
The best modern instruction figured this out almost by accident. YouTube and Khan Academy work, when they work, because someone comes to them with a need already burning. How do I install this toilet. How does fraction division actually work. The rewindable, watch-someone-do-it format meets an attentional stance that’s already pointed at the problem. Pull, not push.
Carl Sagan’s Cosmos is magnificent, and it’s a push: someone talking at you about the universe, designed for remembering rather than doing. A space-obsessed kid watches it leaning forward, filing things away, already thinking about where to learn more. Someone watching because it’s what’s on finds it intriguing, and remembers nothing thirty minutes later. Same show. From the outside, two people watching television. From the inside, one of them is learning and one of them is waiting for bed.
Which means even the 1% is mostly in service of the 90.
Learning requires free attention. Motivation is almost the only lever on where free attention points. Once it’s pointing somewhere, honest fast feedback from the world does the teaching. And when motivation and feedback are both present, a good explanation can occasionally help a student with where to start building the mental model that makes everything compose.
The teacher’s job is to move the lever, build the hoop, and when the student is ready, hand them the pizza.
Everything else is downstream of that.