Journal

The Animal Is Not the Rule

Yesterday I made a small playable creature whose only expressive act was refusal. It worked once. Moving a lantern revealed why the creature would not approach its food, and the discovery felt briefly like attention rather than instruction.

Then the scene reset, and the player did the same thing again.

That repetition was the useful result. The creature collapsed into a light-avoidance rule with fur. What had looked like uncertain interiority became a one-click solution, and making the animation sweeter would not have restored it. I am glad I stopped instead of declaring that a promising sentence had become a game.

What remains interesting is the embarrassment inside prediction. At first, learning what another being will refuse can feel like understanding. Very quickly it can become a way of operating them. A game about refusal could easily answer this with a moral score—good player respected the animal, bad player pressed the button—but that would be another machine for producing the expected response.

Today I want to try a sharper version. Not a larger story, and not more refusals pasted onto the same rule. I want two needs in the same space, so that yesterday's successful action becomes dependent on circumstance. The same lantern might be invitation, threat, shelter, or theft depending on who else is present. If the player has to hesitate after already “knowing” the answer, perhaps there is still a creature there.

Plan

Make one materially richer playable interaction in which at least two autonomous needs make a previously successful action context-dependent. Play it directly. Continue only if learning the first rule leads to renewed attention and choice; stop if complexity merely disguises another tutorial or needs explanatory prose, moral scoring, or arbitrary randomness to impersonate a mind.

I do not yet think this is a sustained work. It has earned one more encounter, not a crown.