Productive Struggle is Misunderstood (2 of 2)
Teaching for the Learning You Can't See
The first article in this series argued that productive struggle is misunderstood. The learning on a hard, open problem doesn’t happen during the struggle. The struggle loads a question the mind can’t yet answer, and the understanding gets built afterward, offline, in the hours and days after class.
But teachers have fifty minutes, a roomful of students, and a syllabus that already outran the calendar. It can sound like the job has left their hands. It hasn’t, though it has changed. If the understanding mostly gets built offline, then the teacher’s work is to hand students something worth carrying out the door and to protect the conditions that let it keep working. The leverage moves to loading the question well and guarding the quiet afterward to the degree possible.
This article argues for three prongs to foster that offline intuitive learning. Note that the evidence on how such skills develop is thin, as the first article covered, so some of what follows is reasoning from neuroscience and psychology rather than educational research.
Make the Challenge Itself the Draw
When teachers hear that a struggle has to matter to the student, they reach for relevance. Rewrite the word problem around basketball for the kid who likes basketball, around music for the kid in band. That is the shallow version of interest, and it wears off fast, because a thin problem in a favorite costume is still a thin problem.
The deeper pull comes from the challenge itself. A question the student didn’t know was a question, a situation whose pieces visibly don’t fit yet, a decision where two reasonable considerations collide, each of these can create its own interest regardless of subject. New knowledge can grip a student precisely because it upends something they thought they understood, answers a curiosity, or more commonly is just an interesting puzzle. A kid who otherwise could care less about World War I might be intrigued by the judgments and culture that might have caused it. Simple “assasination of Archduke Ferdinand” causes aren’t learning instigators, but the combination of militarism, firm alliances, and the like is a much richer, debatable landscape.
The friction is the appeal. A problem that resists a first pass, and that the student can feel resisting, is one the mind keeps gnawing on later.
So the design question is not what this student already likes but whether the problem has a real knot in it. Ask a middle schooler which of several ways to split a fixed budget across competing needs is the fairest, and there is no clean answer, only considerations to weigh. Ask an undergraduate to set a public-health rule where saving the most lives and respecting individual freedom pull in different directions, and the same thing happens. Get the knot right and the interest mostly takes care of itself.
Don’t Jam the Conscious Channel
Hand a student one hard, open problem and the mind has something to carry out of the room. Hand them five in an afternoon, each needing real thought, and none gets carried well, because the conscious mind can only hold and pass off so much before it clogs.
The firmer evidence is about the quiet stretch after a problem is set aside. When people work a hard problem, then take a break before coming back to it, the break helps most when it is filled with something mentally light, not a demanding task and not nothing at all. A light filler task beats both. The nature of the intervening task was a working memory load, and what happens with other kinds of tasks is unclear.
The default mode network, the web of regions that lights up when the mind wanders and is associated with creative thinking, runs anticorrelated with the networks that drive focused, effortful thought. One goes quiet as the other takes over. You can’t run both at full tilt. Fill the conscious channel and you damp the system that does the after-class work. I’ve argued before that schools, by scheduling focused attention wall to wall, give students almost no practice at the mind-wandering their durable skills depend on.
That balance isn’t the same in every head. The same research finds the two systems (executive function and default mode network) separate less cleanly in ADHD, so the default mode network keeps breaking in when focus is called for, which reads as distractibility and also tracks with more divergent, connection-rich thinking. I’ve argued that these students can be stars in an AI era, because the wiring that fights an all-day-focus schedule is the wiring that throws off the unexpected connections AI is raising the value of. They do worst in the packed, heads-down structure of typical schooling, and best when a course is built around big challenges with detail pulled in as needed. But I’d say a schedule that protects downtime and leads with challenges suits nearly everyone.
A course that piles fresh, demanding material back to back does two kinds of damage. It jams the channel so no single problem gets a clean handoff to the subconscious, and it packs the very hours the handoff needs with more of the load that gets in its way. Fewer, deeper challenges beat a long list.
Multi-Threading and Multi-Faceted Challenges
The background work being done by the subconscious isn’t something you carve into a lull in the room. The teacher’s lever is the calendar, not classroom stillness. You protect the quiet by not slamming a problem shut the moment it opens, and by giving it ample time to sit before you pick it up again.
The usual schedule works against that. March one unit to its finish, test it, and then start the next, and the gaps the learning needs are gone. Run several threads at once and the gaps arrive on their own, because while students work thread B, thread A sits in the background, and the days between two sessions on A are where the consolidating happens. Interleaving, which the memory research already backs for plain retention, earns its keep with more complex challenges too.
Spacing is where judgment splits from memorization. Space a fact and you come back to the same fact to see whether it held. Space a judgment and you come back to a harder version of the problem. Push on the decision boundary with a case sitting right at its edge. Tie the idea to a neighboring one in another unit or another subject. Add the wrinkle that makes the call genuinely tougher than it was. The student returns not to rehearse the old struggle but to a deeper one that feeds on it. Repetition hardens a fact. Variation grows a judgment.
Homework that reopens the challenge does far more work for intuitive judgment than repeating a procedure. Twenty variations on a judgment a student is still working out keep the question alive and press on it from different perspectives, where twenty repetitions of a solved procedure only close it further. A single problem they can’t quickly settle can do the same work, and so can a prompt to look for something in the world that bears on what they were wrestling with. Because a variant reopens the challenge and deepens it in the same move, it looks like the surest way to hand the mind a thread and start it working again.
Consider going to an overarching challenge. Rather than separate threads you stitch together by hand, run one substantial challenge, rich enough that its situational variants reach into different units on their own schedule. One week the challenge turns on a quantitative question and pulls in the math. The next it turns on an ethical bind or a historical parallel and pulls in those. The variants carry the variation, the spacing, and the interest at once, because they are the same challenge seen from a new side.
Problem-based and challenge-based learning are not new, and most educators have met them. The trouble is how the challenges usually get built. Scoped to fit one subject and steered toward a tidy answer, they turn into a delivery vehicle for the unit, with the messiness that keeps a question alive engineered out. A challenge that resolves cleanly by the end of the period does the same thing telling a student the answer does, closing the question before it can run. What this series is after is the opposite, a challenge open enough to leave something unresolved when the bell rings.
Yet there are still structural issues. Most real challenges cross subjects, and the ones that grow the highest judgment, the wisdom for approaching a messy, many-sided problem, are exactly the ones no single subject can hold. Boxing a challenge inside one course is a concession to how institutions are built, not the goal. The realistic first move, given those walls, is to take the math or the biology or the writing you are responsible for and set it inside a larger challenge that has to reach for it, so at least the reaching does some work. I’ve made the fuller case for putting the challenge first and pulling knowledge in on demand elsewhere.
None of this is easy to engineer, and a single shared challenge takes more design than a tidy run of units. As AI absorbs the facts and the routine, the human contribution that holds its value is wisdom, the judgment about open and complicated things, and that is exactly the learning that needs a real struggle, loaded well and then left alone to run.
©2026 Dasey Consulting LLC