The biggest single addition: a fourth register

Register is core to Commonplace, and the three today (theoretical / descriptive / prescriptive) are all first-person-committed: their quality bar is "is this claim true / accurate / correct?" Casework fits none. It needs a dialectical/evidential register whose quality bar is "does this faithfully represent the state of contestation, with every position sourced and attributed?"

Defining this register (its COLLECTION.md contract, its title conventions — likely not title-as-claim, since a casebook node shouldn't assert — and its link rules) is the piece most likely to "change how FLF thinks about the problem," because it is the methodological move, not a feature. Every claim/evidence/gap type below would otherwise inherit title-as-claim and "do I still believe this?" maintenance semantics that actively fight neutrality.

Update 2026-07-08: demoted from taxonomy addition to collection contract

The framework-side review resolved the collision this idea had with the register theory's exhaustiveness claim — by demoting the taxonomy, not the idea. Per a universal knowledge framework demotes content taxonomies to defaults and keeps answerability and ADR 042, registers are default profiles and the set is open: the casebook does not need a "fourth register" approved — it writes its COLLECTION.md text contract directly (quality bar: faithful, attributed representation of contestation; topical titles; every claim carries an asserter and a source-span; maintenance question: "is this still an accurate map of the debate?").

Two sharpenings from the same review:

• Linguistic name for the contract. What this mode requires is mandatory evidentiality — every proposition marked with who asserts it, on what basis (grammaticalized in languages like Quechua and Turkish; Aikhenvald's evidentiality survey is the standard reference). In discourse-mode terms (Smith, Modes of Discourse) a casebook is report mode — assertions about assertions — not argumentation. So the contract is descriptively "report + mandatory evidentiality," which is a tighter design brief than "not the three existing registers."
• Answerability stays satisfied. The stance-neutral map is still a knowledge collection under the framework's scope invariant: an attributed position answers to what the party actually asserts, and goes stale when the attribution stops being accurate. Neutrality changes the answerability relation, not its existence.

The proven contract is now named: ADR 042 promoted it as the dialectical/evidential entry in Commonplace's profile catalogue, on the strength of this casework's first worked case.