Groundwork: what gets us to the work?

A guided story about access, effort, and the intellectual work worth preserving. Allow fifteen to twenty minutes, or use a projected demonstration for a shorter session.

Setup: One desktop browser per pair, or a projected demonstration. One person operates; the other explains the plan. Swap roles between lots. Allow additional time for unfamiliar controls. Use a projected or partner-operated version if navigating a 3D scene is uncomfortable.

0–2 minutes: frame and demonstrate

Introduce the story: the player is an apprentice restoring trail signals with Mara. The crew needs both working lights and a keeper who can choose positions on unfamiliar ground. Ask: “What should the learner have to figure out, and what must be available for that to happen?” Demonstrate Plan signal, T to swap, Q to rise, G to retrieve the lantern, and R to return. Field notes reopens the chapter and step-by-step building help. Previewing spends no material.

2–8 minutes: compare the lots

Near yard: Choose a prediction before previewing. Change one value and predict again. Ask what changed the judgment. Would remembered coordinates work on a new landscape?
Far field: Which additional actions support that judgment? Which merely make the material available? Do not label all extra effort productive.
Walled lot: What evidence supports “This cannot be done here”? Keep the explanation about conditions on the lot.

If navigation consumes the session, demonstrate the next step and continue the discussion. Finishing within the activity slot is not a test of ability.

After each lot, choose a response to discuss with Mara or continue without one. Her fixed replies offer perspectives, not scores. Invite disagreement and a counterexample: when might a worked example help a novice, or transporting supplies become the learning goal?

8–11 minutes: change one condition

Each pair predicts what a steel pick, delivered stone, or brighter light will change. Choose one and replay. If it leaves the barrier intact, use Try another condition. Each attempt starts fresh, so interventions do not accumulate.

11–15 minutes: transfer to one real lesson

Start with the closing case: a student must compare sources, one source has a broken login, and an AI tool can supply a paragraph. Which support makes the learning task reachable? Which might model or replace it? A generated example can be useful; a completed paragraph alone does not settle what the student understands.

The table records active time, blocks cut, travel, and requested previews. It does not separate thinking from nonthinking; participants may reason while moving or cutting.

Name the intellectual work students should practice in a real lesson.
Identify one condition that adds avoidable effort or blocks participation.
Choose one change that preserves the intended intellectual work.

Leave with: “Keep students responsible for ____. Change ____ so they can reach that work. Look for ____ in their explanation or performance.”

Research grounding and limits

Soderstrom and Bjork's (2015) Learning Versus Performance distinguishes practice performance from durable retention and transfer. This game does not test either retention or transfer.

Miner's When the Output Looks Like Learning is a conceptual framework with four dimensions. Groundwork adapts it by foregrounding infrastructure as conditions supporting noetic, rhetorical, and existential work. The game mainly invites spatial reasoning and access decisions, not all three forms of development. Mara is an authored character; the AI case is a thought experiment.

Questions for a pilot: Can a novice explain the distinction without using the labels? Does the same learner apply it to a different classroom example later? Where does the simple wall metaphor fail to represent cumulative or changing barriers? Observe usability separately from conceptual understanding.

The same signal criterion does not mean equal cognitive difficulty. This teaching model supplies no evidence of learning or dissertation findings. Continue with Friction Lab or The Friction Atlas.