the warrant-enforcement-semantics round lands its three dispatch rulings: two forms of normative content (M1, landed subject to confirmation), the settlement-semantics fork unruled and run as a fork (M2), and legitimacy, the Normative Inductor and the force layer consumed rather than revised (M3)

Landed as the round's first commit, per the queue rule above (prompts/2026-10-20-warrant-enforcement-semantics/PROMPT.md).

M1 — two forms. Normative content enters the market in two forms: obligations as credal constraints on priced sentences, and values as securities traded directly. Stated by the maintainer in conversation on 2026-08-23 and not previously in this file. The dispatch marks it confirm, amend, or strike; it is landed here so that the round may cite it, and its confirmation is queued under Awaiting the author. If struck, the round's own report says which of its readings survives (Part B, semantics (iii)).

M2 — the fork is unruled. What a warrant's security settles against — deductive settlement, a later legitimate evaluation, or nothing — is not ruled. The round runs the three readings as a fork on one exact fixture and reports; the maintainer rules on the result. The round's finding and the decision it leaves are queued under Awaiting the author when the round closes.

M3 — consumed, not revised. Nothing in the round redefines legitimacy (wiki/Legitimacy.md and the 2026-09-25 entries above), the Normative Inductor architecture (lean/Workspace/Normativity/Contrib/NormativeInductionInterface.lean), or the force layer (projects/normativity/notes/TRADERIZED_FORCE_INTERFACE.md). The round discharges a hypothesis those objects already name — liability in Evaluation.conditional_normative_inductor — and adds nothing to their signatures.