The Science of Learning

Understanding isn't stored as facts — it's stored as models: compact internal machines that predict how something behaves. A child who 'gets' fractions has a machine that answers new fraction questions; a child who memorized fraction rules has a filing cabinet that fails on anything unfiled.

Why it works: Brains are prediction engines, not hard drives. A model compresses a thousand cases into one structure, so it handles cases it has never seen. This is why the Alice Method teaches every math topic as a model first (fractions are amounts, functions are machines) and procedures second.

In practice: Before any procedure, ask: what is this thing, really? Can my child say it in their own words, draw it, and use it on an example they've never seen?

Expertise is largely a library of recognized patterns. Chess masters don't calculate more than novices — they see structures. Strong readers don't decode letter by letter — they recognize words and phrases whole.

Why it works: Recognition is nearly free for the brain; reasoning is expensive. Every pattern moved from 'must figure out' to 'just see it' releases working memory for the genuinely new part of the problem. Volume and variety of examples build the library — which is why the method favors many small exposures over few big ones.

In practice: Ask 'what does this remind you of?' constantly. Play games where the goal is spotting the pattern, not producing the answer.

Memory strengthens when information is pulled out, not when it's put in again. Re-reading feels productive and does almost nothing; attempting to recall — even failing — physically strengthens the trace.

Why it works: Each successful retrieval signals the brain that this memory earns its keep, slowing its decay. This is the most replicated finding in learning science, and it's why the Post-it wall works: removing a note is a retrieval event, not a review event.

In practice: Replace 'let's go over this again' with 'tell me what you remember.' Two minutes of recall beats twenty minutes of re-reading.

The point of learning anything is using it somewhere else. Transfer — applying an idea outside the context where it was learned — is the difference between knowing math and being able to think with it.

Why it works: Transfer fails when knowledge is welded to its original packaging ('worksheet knowledge'). It succeeds when the underlying structure was learned explicitly and met in varied contexts. That's why the method drags every concept out of the workbook and into kitchens, cars, and stores.

In practice: After anything is learned, ask: where else does this show up? Then go find it there this week.

The move from 'this worked here' to 'this works whenever conditions X hold.' It's how three examples become a rule, and how a rule becomes judgment about when the rule applies.

Why it works: Brains generalize automatically — but lazily and often wrongly (every kid who says 'multiplying makes bigger' generalized from whole numbers). Deliberate generalization, with counterexamples, is how you get the boundaries right.

In practice: The two magic questions: 'Will that always work?' and 'Can you find a case where it breaks?' Hunting counterexamples is one of the most joyful games in this method.

Learning happens at the edge of ability — hard enough to require real effort, supported enough that effort usually pays off. Rescue too early and nothing is learned; let struggle turn hopeless and the child learns only that they hate this.

Why it works: Difficulty during practice improves retention (desirable difficulties): the effort of almost-getting-it is precisely what tells the brain this matters. Struggle also builds the meta-skill — the felt knowledge that confusion is a stage, not a verdict.

In practice: When your child is stuck, count to thirty before helping. Then help with a question, not an answer. Calibrate: about 80% success is the sweet spot.

Understanding compresses. What began as ten steps becomes one chunk; what filled a page becomes a glance. 'Carrying the one,' once mastered, is not ten operations — it's one.

Why it works: Working memory holds only a few chunks at once, but a chunk can be arbitrarily rich. Compression through practice is how children get capacity for harder problems — the multiplication fact that becomes automatic frees exactly the mental space that algebra will need.

In practice: Automate the layer below before building the layer above. Fluency drills aren't the opposite of understanding — done in two-minute doses after understanding, they're what makes room for it.

Curiosity is a cognitive state, not a personality trait — the itch that opens when you notice a gap between what you know and what you almost know. It can be induced, and it can be crushed.

Why it works: Curious brains encode better — attention sharpens and memory consolidates more deeply for information you wanted. A question a child asked is worth ten questions a worksheet asked. The method's environment layer is largely a machine for generating questions.

In practice: Answer questions with enthusiasm and, often, with 'I don't know — how could we find out?' Keep a curiosity journal. Never mock a question.

Reasoning from what is fundamentally true rather than from what is usually done. Why does the fraction rule work? What is multiplication actually doing? First-principles habits turn rules from things to memorize into things that could be rebuilt if forgotten.

Why it works: Knowledge derived from principles is self-repairing — forget the rule, re-derive it. It's also the foundation of confidence: a child who has rebuilt an idea once knows the idea belongs to them, not to the teacher.

In practice: Regularly rebuild things: 'We forgot the area formula. Can we figure it out from scratch with paper and scissors?' Treat 'because that's the rule' as a challenge, not an answer.