Wobble
A proper scaffold is earthquake retrofitting for Agency.
There is a very tempting sentence adults like to believe.
It sounds responsible. It sounds structured. It sounds like care.
It goes like this:
If the routine changes, the routine failed.
I understand why people reach for it.
Routines matter.
Predictability matters.
Transitions matter.
For many children, especially children who have been misunderstood, overcorrected, hurried, shamed, surprised, dragged, or made to pay for adult disorganization, routine can be the difference between movement and collapse.
A routine can reduce noise. A routine can preserve capacity. A routine can let a child know what comes next before the world starts shouting at them with a thousand invisible mouths.
So yes.
Routines matter.
But a routine is not a cage.
A scaffold is not a leash.
And a child leaving the expected route does not automatically mean the route failed. Sometimes the routine works so well the child no longer needs to stay inside it.
That’s harder to see — from the outside, it looks like drift:
Mimi was supposed to go to swing. Swing was the route. Swing was known. Swing was predictable. Swing had worked before.
Then Mimi saw Tessa.
Then Mimi waited.
Then Mimi hugged Tessa.
Then Mimi did not want swing anymore. She wanted her friend.
If we read that badly, the adult move seems obvious.
Not right now, Tessa.
It’s Mimi’s time for swing.
We can play later.
Come on, Mimi.
First swing, then washroom.
That move would sound safe. It would sound structured. It would sound like the adult was protecting the routine. But underneath the nice adult voice, something else would be happening.
The plan would be owning Mimi’s movement.
The routine would be treated as the payer.
The child’s new route would be treated as error.
That is the shadow.
Sometimes adults protect the plan so hard that they interrupt the learning the plan made possible.
Mimi’s routine did not fail because she left the swing route.
Mimi’s routine succeeded because she had enough agency to choose a new one.
That’s what I mean by wobble.
Not collapse.
Not chaos.
Not failure.
Wobble.
Change entering the field.
The Routine Is Not the Point
A routine is supposed to preserve movement.
It is not supposed to own movement.
Note that distinction.
If the routine owns the child, then success means the child stays inside the routine.
The adult asks:
Did she follow the schedule?
Did she complete the transition?
Did she go where we expected?
Did she avoid disruption?
Those aren’t useless questions. But they are also not the deepest questions.
Here’s something that scratches beneath the surface:
Did the routine preserve enough agency for the child to move?
Because if the answer is yes, then a changed route may be evidence of growth.
A scaffold succeeds when the child can leave it without falling through the floor.
That isn’t a metaphor I use lightly. A scaffold isn’t proof that the learner can’t move. A scaffold is the temporary structure that lets the learner move while the world is still too expensive to pay all at once.
It holds enough load for the child to try.
It reduces enough risk for the child to reach.
It preserves enough return for the child to wobble without being punished by gravity.
So, when Mimi turned away from swing and toward Tessa, the question shouldn’t be:
Why is Mimi off routine?
Instead, consider:
Is the old route still available if this new route becomes too expensive?
If yes, then the adult doesn’t actually need to panic.
The adult can offer the new route without burning the old bridge.
That’s the clean move:
“Wow, thank you for the invitation, Tessa.
What do you think, Mimi?
Ready to give it a shot?
We have swing ready if you decide, ‘not now.’”
That preserves three things at once.
Tessa’s invitation remains alive.
Mimi’s agency remains alive.
The original scaffold remains alive.
Nobody has to be wrong.
Nobody has to be erased.
Nobody has to be forced into the old plan just because the new route surprised the adults. That’s what good scaffolding does.
It doesn’t prevent wobble.
It prevents collapse when wobble arrives.
Wobble Is Not Failure
Wobble is change entering the field.
That’s all.
Sometimes wobble is big.
Sometimes wobble is tiny.
Sometimes wobble is a child crying.
Sometimes wobble is an adult being pulled away.
Sometimes wobble is a math problem no one in the room actually knows how to do yet.
Sometimes wobble is Owen appearing in the wrong hallway with the confidence of someone who has personally pinned fractions for the three-count.
Sometimes wobble is Mimi seeing Tessa and deciding that swing can wait.
Sometimes wobble is a happy little accident.
The adult field often treats wobble like a threat because wobble breaks prediction.
Learning breaks prediction, too.
Growth breaks prediction.
Agency breaks prediction.
A child who can only succeed by staying exactly where the adult placed them has not necessarily learned to carry.
They may have only learned not to disturb the plan.
That can look beautiful from a distance. It can look regulated. It can look successful. It can look like the adult knew what they were doing. It can look like the child knew what they were doing, too.
But if the child cannot change course, cannot refuse, cannot choose, cannot return, cannot try something new, cannot leave the scaffold without being treated as a problem, then the plan has not restored agency.
It has purchased appearance.
A strong plan is not one that never wobbles;
A strong plan knows what to do when wobble appears.
A routine is not built to prevent change.
It is built to prevent collapse when change arrives.
Learning-as-Flux
This is where I need to name the pattern.
I call it Learning-as-Flux.
Not because we need another fancy phrase in education.
We absolutely don’t. Just ask any veteran who’s watched enough colleagues and administrators come and go to manifest ‘revolving door’ as the school’s front entrance. And yes, maybe they’re the patch of moss and dandelions that keeps sprouting between the cracks in the concrete.
Say your words and be done with it.
I’m here to stay.
The cupboard is full. The labels are stacked to the ceiling and somehow still multiplying.
But sometimes a phrase helps us stop reaching for the wrong one.
Learning-as-Flux treats learning as movement through changing constraints toward an objective.
The learner succeeds when the objective is reached in a way the learner can own.
The learner remains in process while movement continues and the account stays open.
The learner may pause or return without shame because ownership does not require uninterrupted motion.
The learner may abandon or revise the objective when the field changes without forfeiting agency.
Agency isn’t just movement.
A body can be moved. A schedule can be completed. A transition can happen. A worksheet can be finished.
That doesn’t mean the learner owned the move.
Agency begins when the movement can return to the learner as theirs.
Notice what is not there.
Failure.
That is not an accident.
Failure is explicitly refused.
There is no failure-state inside Learning-as-Flux.
That does not mean nothing goes wrong.
That does not mean every attempt works.
That does not mean consequences disappear.
There can be missed attempts. There can be mess. There can be cost. There can be retreat. There can be repair. There can be radioactive sludge.
But none of that requires declaring the learner failed while the loop is still readable.
Failure is not a terminal state inside learning.
Failure is often the classification adults slap onto the loop after they stop reading.
This matters because once adults call failure, the field changes.
The adult posture changes.
The child’s possible movement changes.
The next move becomes correction. Or consequence. Or compliance. Or “try again later.” Or “not ready.” Or “too dysregulated.” Or “not there yet.”
Sometimes those labels may describe something useful.
But they can also become tombstones.
A living account gets declared dead because the adult stopped reading too soon.
Learning-as-Flux refuses that.
It asks:
What is the current objective?
Is the learner still in motion?
Did the field change?
Did the objective change?
Is return still available?
What became payable? What became too expensive?
What is the next smallest move the learner can actually own?
That informs a very different read.
Objective Abandoned
One of the states inside Learning-as-Flux is objective abandoned.
I know.
It sounds cold.
|It sounds like accountant language.
Good.
Sometimes we need accountant language because sentimental language gets slippery and starts laundering the cost.
Objective abandoned does not mean the learner failed.
It means the old objective is no longer the active objective.
That can happen in a second.
I was going that way.
Nope.
This way.
Actually, never mind.
Back.
That is not moral collapse.
That is live navigation.
Mimi’s old objective was swing.
Swing mattered.
Swing was real.
Swing didn’t become fake just because it changed.
The field changed. Tessa appeared. Friend appeared.
A new route opened.
The old objective was abandoned.
A new objective came online.
Try something with friend.
That is not failure.
That is a live update.
Mimi was not off routine.
She refreshed the read.
If it resonated with the notion “the new objective’s cooler/shinier/better, let’s do that instead”, that’s exactly what it should feel like. The guiding question then becomes whether the new objective remained payable.
Could Mimi still return to swing if the new route became too much?
Could Tessa refuse?
Could the adult keep both routes alive long enough to see what the children did? Could the field allow the new route to test itself without turning the child into a problem?
That is where growth lives.
Not inside the plan. Not outside the plan.
At the edge where the plan has done enough work that the learner can risk something new.
Risky Play
This is why Risky Play matters.
Not fake danger.
Not reckless adult neglect.
Not “let children do whatever and call it development.”
Risky Play means the child gets to take meaningful, measured risks in a field that still preserves return.
The child reaches beyond the known-good route.
The adult does not eliminate all risk. The adult preserves enough structure that risk can become learning instead of catastrophe.
That’s a very different thing.
In school language, scaffolds often get discussed around academic work, self-regulation, social-emotional learning, transitions, and access.
All useful.
But I think one of the best tests of a scaffold is whether it can support Risky Play.
Because Risky Play tends to be agent-led.
It is meaningful. It is alive. It often involves peers, novelty, uncertainty, negotiation, refusal, repair, and consequence.
It looks a lot more like life outside the worksheet.
A scaffold that only works when the child stays inside adult-authored tasks may be useful, but limited. A scaffold that lets a child safely enter meaningful risk may actually restore agency.
The move is small. It has to be.
ONE reversible step beyond the known-good.
That’s the key.
A proper scaffold doesn’t grant you permission to do anything you want, nor does it scream at you to stay within its territory forever.
It should tell you:
You can try that.
And if it becomes too much, return is still available.
That is how the child learns that the edge is survivable.
That is how the child learns that wobble is not the same as doom.
That is how the child learns that choice can be costly without being catastrophic.
Success gives the child evidence:
I moved.
I tried.
I carried something new.
A miss gives different evidence:
I moved.
It was too much.
I came back.
The route still held.
I can try again later.
Both matter.
Both are learning.
No failure required.
Happy Little Formation
In the self-reg room, something happened that wasn’t planned.
Mimi had a thread.
Tessa had a thread.
The EA had a thread.
The room had a thread.
The math had a thread, too.
Swing was still somewhere in the background, alive but no longer dominant.
The adult did not design that formation ahead of time.
Nobody sat in a meeting and wrote:
At 1:14 p.m., Mimi will abandon swing to sit beside Tessa, who will be holding math, while an EA who cannot do the math models how to seek help through video-based instruction and co-regulated breathing.
That would be an insane plan. Everybody would laugh and spread receipts across the table.
Each one would scream:
“This is why that cannot be.”
And yet, when the threads met, something scaffold-like appeared.
A temporary support structure formed out of live material already present in the field.
I call that an emergent scaffold.
Story-facing, I like “happy little formation.”
Because sometimes Bob Ross is more useful than policy language.
A happy little formation is not imposed.
It emerges.
That is where Currere and Complexity finally shake hands.
The story carries the lived thread.
The field carries the emergence.
That is why this is not implementation.
It is formation-reading.
The child carries one thread. Another child carries another.
The adult does not murder the moment with procedural correctness. The adult reads the formation and protects the conditions long enough to see whether it carries.
That is what happened.
Mimi was not assigned to Tessa.
Tessa was not made responsible for Mimi.
The EA did not make the girls perform friendship.
The adult did not turn the moment into a social objective.
The formation appeared because the children had live threads and the adult field did not flatten them.
That is delicate.
Because adults love converting emergence into programs.
A child shows interest once and suddenly there is a goal, a chart, a prompt, a reinforcer, a forced repetition, and a clipboard standing in the corner wearing spontaneity as a skin suit.
Please do not do that.
Emergence is not a vending machine.
If the moment becomes adult-owned, the child’s movement changes.
The happy little formation works only while the movement remains owned by the children inside it.
Adult support can protect the edge.
It cannot own the route.
The Whetstone
This is also how I use AI.
Not as authority.
Not as author.
Not as oracle.
As a whetstone.
A whetstone does not own the blade.
It does not choose the cut.
It does not decide what needs to be carved.
It creates friction against an edge so the human can feel where the edge is dull, where it catches, where it slips, and where it sharpens.
That is the only reason this arrangement stays clean.
AI produces wobble.
Sometimes that wobble is useful.
Sometimes it opens two or three unexpected lanes.
Sometimes it gives me the phrase I did not know I was reaching for.
Sometimes it misreads the whole room and starts confidently seasoning the soup with drywall.
That is fine. Not because the output is automatically good.
Because the process has a scaffold.
I do not ask AI to replace the read.
I ask it to move against a constraint map dense enough that drift becomes visible.
AI will create wobble.
It’s true.
The question then becomes whether the human has a scaffold strong enough to use wobble without surrendering the work.
If the scaffold is weak, AI drift becomes authority.
If the scaffold is strong, AI drift becomes information.
That is Whetstone Protocol.
One move.
Read.
Check drift.
Return to the map.
Revise.
Try again.
The tool may sharpen the blade.
It does not hold the knife.
The Constraint Map
This is where the constraint map matters.
A constraint map is not a prompt trick.
It isn’t “make the AI sound like me.”
It isn’t vibes.
It isn’t a brand voice document wearing a trench coat and pretending to be doctrine.
A constraint map is a dense set of boundaries, refusals, routing checks, language locks, audience limits, and proof conditions that lets me see when the work is drifting.
It tells the system what not to do. It tells me what to watch for. It keeps certain moves expensive. It makes certain errors visible.
It protects the work from becoming easier in the wrong direction.
That last one matters.
Easy is dangerous.
A smooth answer can be wrong.
A satisfying answer can be wrong.
A beautiful answer can steal the child.
A compelling explanation can flatten the field.
So, my constraint map keeps asking:
Are we still tied to observables?
Did we make an interior claim?
Did we turn support into rescue?
Did we let the adult become the hidden battery?
Did we route cost onto the child?
Did we accidentally make refusal more expensive?
Did we preserve authorship?
Did we preserve autonomy?
Did we preserve agency?
Did we protect the reader from mistaking resonance for receipt?
Did we make the writing clearer without making the doctrine cheaper?
That is why the map needs density. The constraint map has enough tangent lines that when a curve appears, the read starts becoming difficult to miss.
Not impossible to miss.
Difficult.
The human still owns confirmation.
The human still owns judgment.
The human still owns movement.
The human still presses the green button.
The map does not remove responsibility.
It makes responsibility more visible.
That is the whole point.
Refusal Comes First
Refusal comes first.
Not because refusal is negativity.
Not because refusal is resistance for its own sake.
Refusal comes first because refusal is the purest form of agency.
A tool may suggest.
A tool may sharpen.
A tool may create wobble.
A tool may not make refusal more expensive.
That is the line.
If a scaffold makes refusal harder, it is not restoring agency.
If a routine makes refusal impossible, it is not preserving safety.
If an accommodation makes movement possible only by trapping the learner inside someone else’s plan, it needs to be reread.
If AI makes the writing faster but makes the human less able to say no, the tool is no longer a whetstone.
It is steering.
That is why refusal has to sit underneath the whole stack.
Only after that comes the Triple A stack:
Authorship. The work remains mine.
Autonomy. The direction remains mine to choose.
Agency. The movement remains mine to own, carry, and return to.
That applies to writing. It applies to AI. It applies to adults. It applies to children.
Because return to matters. It says the move doesn’t merely happen through me once; it wasn’t a fluke. Nor was it given to me after I pressed a button.
It remains available as mine after the move is completed.
It remains mine after the scaffold fades.
When adult facilitators protect a child’s authorship, autonomy, and agency, they protect the conditions from which independence can emerge.
Independence does not emerge because adults withdraw support.
Independence emerges when support protects authorship, autonomy, and agency long enough for the child to begin carrying movement themselves.
That is the difference between scaffold and dependency.
A controlling scaffold may produce compliance.
An agency-preserving scaffold can produce independence.
Because the child gets to experience:
I chose.
I tried.
I could return.
I could change the objective.
I carried the next move.
That is what we are protecting.
Not the plan.
Not adult smoothness.
Not the appearance of success.
Movement that can eventually belong to the child.
Imagine a child in a wheelchair. Let’s make it a power wheelchair because we’re in the business of being kind to children. Let’s put Owen in the chair because that’s the cool thing about stories: we can test the arrangement as long as we keep treating our characters with love, affection, and care.
We love you, Owen.
Imagine him bumping into things in the classroom and taking longer to do everything.
Imagine him causing a scene every time he needed to go to the bathroom. Or get water. Or transition to a different space.
Imagine him derailing your carefully crafted classroom experience because he keeps getting reminded of all the things he wants to do but can’t.
Imagine him participating on Sports Day and earning points for his team.
Imagine him channeling Steve Nash through no-look passes and bouncing the ball intentionally off his wheels to Jimmy in the pocket who’s wide open.
Imagine him learning about Terry Fox and then deciding he wants to know more about marathons.
Imagine him wanting to do a lap around the school every day in honour of Terry Fox.
Now look Owen in the eye.
Can you affirm his agency?
Or did the chair make the movement too scaffolded for you to call it his?
Earthquake Retrofitting
A scaffold does not stop the world from moving.
It cannot.
Reality wobbles.
Plans wobble.
Relationships wobble.
Learning wobbles.
The point is not to prevent movement.
The point is to keep the load-bearing beams from cracking when movement arrives.
That is what a good routine does.
That is what a good support plan does.
That is what a good adult does.
That is what a good constraint map does.
That is what Whetstone Protocol is trying to do with AI.
Not prevent wobble.
Not worship wobble.
Not call every wobble growth.
Hold the structure long enough to read what the wobble is.
Some wobble is noise.
Some wobble is risk.
Some wobble is warning.
Some wobble is opportunity.
Some wobble is the child reaching beyond the last version of the plan.
If adults cannot tell the difference between different kinds of wobble, they may suppress the very movement they were trying to support.
That is why “failure” is far oo blunt.
Failure can name a result after a loop is truly closed. Yes, failure is definitely a possible read.
But inside live learning, failure often arrives too early.
It can become a hammer. It can become a ceiling.
It can become the adult’s excuse to stop reading.
We need better discipline than that.
Read.
Move once.
Reread.
Settle what actually settled. Keep return available. Let the child own what the child can own.
Protect refusal.
Protect authorship.
Protect autonomy.
Protect agency.
And when wobble comes, do not panic just because the plan moved.
Ask what is carrying.
Ask who is paying.
Ask whether return remains available.
Ask whether the child is disappearing or becoming more visible.
Because sometimes the scaffold works so well the learner no longer needs to stay inside it.
Sometimes the routine succeeds by being left.
Sometimes the old objective is abandoned because a better one came online.
Sometimes the child was not failing the plan.
Sometimes the child was outgrowing that version of it in real time.
A proper scaffold is earthquake retrofitting for agency.
It does not stop the world from moving.
It keeps the child from being crushed when it does.