What I Kept Rejecting —
and Why That Matters

The first version worked. An educator who had never heard of Reggio Emilia would have found it impressive. I almost shipped it.

The anchor questions looked right. Poetic, open-ended, thoughtful. But they came from the topic — not from the child. There's a difference, and it's the whole thing.

Early Spark would generate: "What does the wind carry — and what does it leave behind?" That's a beautiful question. It's also mine. It belongs to an adult who has read about wind and thought about loss and transience. It does not belong to the nursery child who looked at leaves swaying and said "Look, they're dancing" — and meant it as a theory about aliveness, not about wind science at all.

The question that follows that moment has to live inside what the child said. Not above it. So I kept rebuilding until the anchor question could start from a child's exact words rather than from a topic I typed in.

The materials had the same problem. Everything Spark suggested was something that moves when acted upon — leaves, feathers, organza in the breeze. Beautiful. Also incomplete. Reggio practice asks children to be agents, not just observers. I needed materials that put the child on the other side of the equation — their own breath, a fan they control, their hands moving through the air. The experience of being the wind, not just watching what wind does.

The inquiry questions stayed inside metaphor. All three, every time. Poetic questions are not wrong — but a set of three that never asks why or what would happen if becomes a ceiling rather than an opening. I kept rejecting outputs until wonder and curiosity about cause appeared together, reliably.

There was a version of the documentation prompt that asked teachers to write down "a moment that surprised you."

I liked the language. It felt soft and inviting. I pulled it because it was too open. It let the teacher decide what counted as worth noticing — which means a less experienced educator would write down the loudest moment, the most photogenic one, the one that confirmed what they already believed.

Experienced Reggio educators know that documentation is not about what surprised you. It is about what the child said, what the child's body did, and whether the child came back. Those three things are not interchangeable. They tell you different things about where the inquiry is alive.

I split the documentation prompt into three separate fields — words, body, return — and I shipped it. Used it. Reflected on it. Reformed it.

The same thinking shaped how Spark understands participation. Children communicate through drawing, construction, movement, gesture, sound, silence, and the way they arrange objects in space — not verbal language alone. If the tool only helped educators notice what children said out loud, it was missing most of the inquiry.

An experienced Reggio educator might read Spark and say: this is still adult-designed.

You're generating the provocation before you've met the children. That's not emergent curriculum. They'd be right — partially.

Spark is not a replacement for observation. The observation has to happen first — and it has to be yours. What Spark helps with is the translation: you've seen something, you know it matters, but you can't yet find the shape of it. That gap between noticing and knowing what to do next is where most educators get stuck. That's what the tool is actually for.

What Spark holds is not pedagogy. Pedagogy is relational, contextual, lived — it can't be built into software. What Spark holds is ten years of accumulated knowledge about what a useful provocation looks like, and a structure that helps educators think more carefully with that knowledge rather than against a blank page.

The sequencing is still entirely the teacher's. What the children do on day one shapes what happens on day two — and nothing Spark generates can anticipate that. The educated eye remains the most important thing in the room. Spark just gives it somewhere to begin.

And sometimes that beginning goes nowhere. Some children will ignore the provocation entirely. The inquiry may end in five minutes. That is information too — and a thinking companion worth its name should help you notice that, not paper over it.

What Spark gives you is a door. What happens when the children walk through it — that part is entirely yours.

Expertise shows up most clearly in what you choose to reject — not what you produce.

I rejected anchor questions that belonged to adults. I rejected materials that kept children as observers. I rejected documentation prompts that were too open to be useful. I rejected teacher positioning guidance that told educators to "be present but not intrusive" — because that phrase means nothing to someone who has never tried to practise it in a real classroom with real children who are testing every boundary you have.

Each rejection came from the same place: ten years of knowing what a meaningful provocation actually feels like when you're standing in front of children — and being unwilling to accept an output that didn't carry that weight.

What Spark generates now is the result of that process. Not pedagogy packaged into software. A thinking companion that has absorbed enough to ask better questions — and knows when to get out of the way.