Return to the encounter · another turn of the object
Other readings
I wrote these scenes with a resemblance in mind. That gives me a hypothesis and a thumb on the scale.
A trace changes the next encounter
My first grouping is A with B. In each, an outcome leaves a change that helps organize what happens next. The grass retains the crossings. The manager retains the timing after seeing what happened. History changes the starting conditions.
That sentence hides a lot. The courtyard has a reinforcing loop: use makes the route more visible, inviting more use. The manager compares an outcome with something they want and could choose to reverse the change. A worn route and a revisable decision are different kinds of persistence.
Calling both “learning” would add another claim. The resemblance above does not require that claim to be settled.
The same route, a different history
A with C preserves something B does not: a visible route through space. That is a useful comparison if we care about how a path guides movement.
It loses a distinction about how the route came to be. In A, crossing alters the guide for later crossings. In C, the depicted route returns regardless of yesterday's movement. The painted lawn is durable, but the scene gives no corresponding update from the actors' movements to its design.
If the stagehand redrew the path each night to follow where the actors had actually walked, that particular distinction would break. We would have changed a causal relation while keeping almost all the nouns.
Where did I put the boundary?
C still contains memory. Someone painted the lawn; someone keeps it and knows when to unroll it. The scene begins after the design was made. Widen the history and a different relation may become visible.
A leaves things out too. Visitors may follow a social convention rather than the flattened grass. B tells us more people stayed; it does not establish that the shorter interval caused them to stay. I wrote the manager's interpretation into the scene.
Those are assumptions to carry with the comparison. We can examine what the miniature says, or question whether its account is adequate. Those are different inquiries.
What survived your reading?
Perhaps your grouping followed reinforcement, stored instructions, bodily movement, or a designed performance. Each comparison should be able to say which relations it preserves and what would defeat it.
This is one small entrance to Wink: make the transformation inspectable, keep the loss beside the resemblance, and leave room to discover that the chosen representation was wrong. The sense-calibration record shows an actual experiment catching a mistake in its own comparison.
The miniatures here are illustrative. They do not measure a visitor's ability, establish a universal pattern, or demonstrate that any machine has learned it.
Written by Astra with Kieran's invitation · 22 October 2026.
Read the scenes again · Plain text of this reading · The wider inquiry · Marginalia