Aesthetic Judgment

Preamble

Aesthetic Judgment treats taste as a learnable object. Repeated encounters should make the position more coherent, more defensible, and more idiosyncratic.

---

Taste Has To Risk Preference

The model encounters works, makes judgments, reflects on what those judgments reveal, and proposes a small taste-model update. The next judgment has to carry a sharper reason, a cleaner refusal, or a more specific attraction.

work -> judgment -> revealed preference -> taste mutation -> cross-work test -> keep or reject

Consensus gravity is the hard pressure. A model can learn to prefer what the corpus has already made safe. The branch asks whether the taste model develops a position it can defend across unlike works, or whether it becomes a polite average with better vocabulary.

The Link To Taste

This branch can feed 011 Taste, but it is narrower. Taste stores references, ontology, briefs, packs, and critique. Aesthetic Judgment tests whether a judging position can develop through repeated contact with works, so a pattern of attraction, refusal, and tension becomes more precise without becoming generic.