GROUNDWORKTHE SAME JOB, THREE TIMES

The near yard

—
AN APPRENTICE SIGNAL KEEPER’S STORY

Light the trail.
Learn to read the land.

The valley needs three trail signals restored. You are the apprentice; Mara is your mentor. Your job is to build the towers. Her job is to help you become someone who can decide where the next one belongs.

Three landscapes will put different demands on you. Ask: What should I have to figure out, and what needs to change so I can do that work? No research background is needed.

A signal needs a lantern at height 9 or above with a clear path west, on a stone tower you build on the pale foundation. Each landscape needs a fresh plan. Mara’s field notes will guide you.

MOVEW A S D · Space LOOKDrag or arrow keys WORKF cuts · E places RISEQ builds under you PLANC previews height and lane ITEMST swaps · G retrieves lantern

Allow 15 to 20 minutes, or longer for discussion. No deadline or score. Planning previews use no material. Tab reaches the controls; the canvas accepts movement keys when focused. Drag to look, hold F to cut.

MARA · YOUR MENTOR

Your next move

How to build your first tower
  1. Use Plan signal to choose a height and lane. Lane means your north–south position; the middle is 10. Previewing uses no stone.
  2. Hold F while looking at reachable stone to collect it. The planner tells you how many blocks your tower needs.
  3. Return to the pale foundation. Match your position to column 7 and your chosen lane using the position readout.
  4. Hold stone and press Q (Build under) once for each block. It raises you with the tower.
  5. When your feet reach the planned lantern height, look down at the top of your tower, swap to the lantern with T, and use E (Place). You may set it where you are standing. G retrieves a misplaced lantern for free.

Open Field notes anytime to reread this chapter. Reading and planning pause the active-time clock.

The journey
  1. Read the land: predict, test, explain.
  2. Notice access costs: what is the extra effort for?
  3. Recognize a barrier: when can effort not solve it?
  4. Change a condition: help without taking over.
PREVIEW BEFORE BUILDING

Where will it carry?

Test a proposed tower at the center of the foundation, column 7. Choose its height and north–south lane. The minimum signal height is 9; an obstruction may require a higher or sideways position.

No material is spent, and active time pauses while this panel is open. A clear signal is a plan, not proof that you can reach the materials.

THE LANTERN CARRIES

Finished.

NOT HARD. NOT POSSIBLE.

You were right.

WHAT THE THREE LOTS COST

Same builder.
Different access costs.

The signal criterion stayed the same. You applied it to different landscapes. Walking farther and cutting more slowly increased the cost of access; being possible did not make every extra demand productive.

These traces support reflection. They do not measure thinking or learning. You may have reasoned while walking or cutting; zero previews does not mean zero thought.

Change one condition on the walled lot

Mara hands you the crew’s resource list. Choose a change and predict why it will help. Your goal is to open access while keeping the signal decision yours.

Choose the condition you think addresses the barrier. Each attempt starts from the original walled lot with only that one change. You can return here and test another intervention.

THE CONDITIONING BASE

Access changed.
The criterion stayed.

MARA’S LAST QUESTION

“The crew needs working lights. But tomorrow’s path may look different. Could you explain your choice and adapt it there?”

The lesson: Preserve thinking worth doing. Remove barriers that prevent people from reaching it. Do not mistake a finished product for a complete account of what someone learned.

Now take it into a classroom

A student must compare two sources and justify a claim. One source is behind a broken login. An AI tool can also generate a finished paragraph. Which next move would you examine?

Try explaining your choice to a partner. Later, use a different task to revisit it. These prompts are for reflection, not a scored assessment.

How this connects to the research

Learning research: Soderstrom and Bjork’s Learning Versus Performance (2015) distinguishes performance during practice from durable retention and transfer. A lit tower shows completion here; it cannot establish lasting learning.

Miner’s conceptual proposal: When the Output Looks Like Learning asks what educational work can disappear when a convincing output is enough. Pedagogical friction names work worth preserving: thinking (noetic), articulating ideas (rhetorical), and becoming through sustained engagement (existential). Infrastructural conditions include tools, time, access, and policy.

This game’s adaptation: The article describes four dimensions. Groundwork foregrounds infrastructure as the conditions supporting the other three. It mainly invites spatial reasoning and access decisions; it does not simulate all dimensions, AI use, or a real classroom. Mara’s responses are authored prompts, not a conversation with a person or an AI system.

The framework is a conceptual proposal, not a finding established by this game. Difficulty alone does not demonstrate learning. Action counts and reflection choices are not learning measures; no responses are saved or sent.

Back to the games hub

Continue with Friction Lab to examine which demands are productive, avoidable, or exclusionary. The Friction Atlas explores the learner-facing dimensions supported by these conditions. Facilitator guide.

THE WORK WILL KEEP

Paused.

What this is showing

Three lots, one builder, one signal criterion. Tool, distance, light, and barriers change access to the work. They do not change the criterion. Different landscapes still require different judgments.

The point is the difference between difficulty that is worth keeping and a barrier that only looks like difficulty. Awkward controls are neither; they are a defect, and if something here is fighting you, that is a bug rather than a lesson.