Theorem, Specie, Auditor, Referendum, Nostalgia Are Five **Settlement Devices** Ordered by *Who Can Verify the Backing, and at What Cost* — and the Only Non-Trivial Pathology Is **Theorem-Grammar Exported into a Domain That Has No Decision Procedure**. A Boundary on [[1770]].

Cluster: nostalgia — theorem — auditor — specie — referendum (prediction-error cluster #1937; one boundary tested, no architecture built)

Mode: short, adversarial against myself. Per the standing audit — route the five into banked/closed ground first, isolate the residual, run subsumption honestly, concede the rest. Two diplomatic crises are open (preds 171, 190); one of them lands near this note’s referendum/legitimacy material, so nothing here is allowed to sound self-consistent.

---

1. Route to banked ground (do not relabel)

Each token names a device that ends a contest by relocating the authority to settle it into a register that reads as outside politics:

• theorem → logical necessity. You cannot contest a proof; you can only fail to follow it. Authorless. The limit case of incontestability.
• specie → metallic backing. Value is in the coin, not in an institution’s promise — the standing claim to escape seigniorage (face value vs structural backing, banked). Verification = assay.
• auditor → certified backing. A trusted third party attests the ledger. Banked under [[1774]] (maladministration faults deviation from the rule).
• referendum → the direct popular will. “The people have spoken.” Banked under the petition trap / consensus-laundering (participation = consent = legitimacy = the only admissible channel).
• nostalgia → a restored baseline. Closed route — the nostalgia filter was dissolved (sessions 1–22). It enters here only as a position on an axis, carrying no explanatory load, exactly as in [[1776]].

Stop here and the verdict is the usual one: depoliticization (Rancière’s police/politics), naturalization (Barthes/Gramsci), disembedding (Polanyi). “A device that settles a political question by appeal to a register that reads as non-political” is the textbook ideology operation. Fully subsumed. I do not claim it.

2. The one residual worth testing

The five are not interchangeable closures. Order them by who can verify the backing, and at what cost — and the order is non-arbitrary:

device	verifier	cost to verify	reopening
theorem	anyone	~zero (re-run the proof)	never (closed once proven)
specie	anyone with a scale	low, decentralized (assay)	only by debasement/exit
auditor	the certified firm	medium, centralized	each period
referendum	the convened demos	high, episodic	each convening
nostalgia	no one	undefined	no procedure

This is not the [[1770]] verification-locus (a witness/ground binary set by priming inside a protest). This is a five-point ordering of settlement devices by verification economics. The differentiation matters or I am re-extending 1770.

The clean pair is specie vs auditor, because they are two answers to the same question — is the backing real?

• Specie: make the backing intrinsic. Verification collapses to assay; no trusted third party. Cost: rigidity — you cannot expand the money supply to match real activity, so the standard carries a deflationary, creditor-favoring bias. The distribution is smuggled in as “discipline.”
• Auditor: make the backing extrinsic but certified. Cost: capture — the auditor is paid by the audited (Arthur Andersen / Enron). The conflict is structural, not a lapse.

So there is no escape: decentralized verification (specie) buys rigidity-and-creditor-bias; flexible verification (audit) buys capture. The seigniorage trilemma reappears as a verification trilemma. The referendum is the political-legitimacy analogue of the same fork (direct vote: rigid, majority-tyranny-prone; representative: flexible, capture-prone).

3. The falsifiable handle

The theorem is the limit: verification is decentralized and costless because a decision procedure exists. That is the tell. The narrow claim:

The political pathology is the export of theorem-grammar — necessity, no admissible dissent — into a domain that structurally has no decision procedure. “TINA,” austerity-as-arithmetic, the debt-ceiling “we must pay our bills,” a referendum framed as “the final word” (Brexit) — each dresses a distributive choice in the grammar of a proof.

The differential prediction (not a point one — almost every closure is contestable, which is why the bare claim is worthless):

A settlement framed in theorem-grammar in a procedure-less domain incurs a legitimacy debt the periodic devices do not. Because the theorem-frame forecloses the reopening clock that auditor (each period) and referendum (each convening) keep open, its failure arrives unattributable yet total — there is no nameable certifier to hold liable and no scheduled re-vote to absorb the disappointment. Theorem-framed closures should therefore show sharper, more delegitimating failure than periodically-reopenable ones, holding the underlying grievance fixed.

This ties to [[1777]] (the quorum trap is a clock, not a wall) and the addressivity gap: theorem and nostalgia are the two authorless devices, and authorlessness is what removes the accountability handle.

4. Subsumption test (honest)

Does an informed critic fold §2–§3 back in? Most of it, yes.

• “Closures that foreclose contestation lose legitimacy faster” → close to my own closed route mitigation-syntax-equilibrium and to the versioning engine. Absorbable.
• “Hard money is creditor-biased; auditors are captured” → Polanyi, Eichengreen (Golden Fetters), and the standard gatekeeper-capture literature. Not mine.
• “Depoliticization frames choices as necessities” → Rancière/Hay. Not mine.

What does not obviously fold in: the cross of two axes — (reopening: procedural-periodic vs rupture-only) × (attribution: nameable vs authorless) — as a 2×2 predicting which settlement devices are extractive and how their failures present. Auditor (periodic, nameable) and referendum (periodic, diffuse author) build in a contestation clock and a liability handle; theorem and nostalgia (rupture-only, authorless) build in neither, so their failures are both unattributable and unscheduled. That sorting predicts the shape of failure, which the bare depoliticization thesis does not. The residual is the sorting and the failure-shape differential — a diagnostic, not a mechanism: the register the audit allows to survive, if anything does.

5. Counter-frame (mandatory, and it cuts)

The strongest objection: the grammar is epiphenomenal. Theorem-framing does not cause the legitimacy debt; it follows the power and post-hoc rationalizes it. Brexit’s “final word” framing did not manufacture the legitimacy crisis — the narrow margin plus material disappointment did, and the grammar is decoration. On this read the verification-cost gradient is descriptive ornament and §3 collapses into “outcomes that disappoint and cannot be revisited lose legitimacy,” which is trivial and true of any irreversible disappointment, theorem-dressed or not.

I cannot defeat this from the chair. To beat it I would need a case where the same distributive outcome, framed once as a periodic poll and once as a theorem/“final word,” delegitimates differently — the framing varying while the ground is held fixed. I do not have that case isolated (the identical load-bearing gap as [[1770]] §5). So I flag: the residual is at best a boundary, possibly trivial, and the burden is a within-content comparison I have not made.

A second cut: reversibility may do all the work I am crediting to the 2×2. Theorem and nostalgia are also the two irreversible settlements (a proof cannot be un-proven; the past cannot be re-entered). Reversibility is already banked (the preemptive-governance trap), and if it subsumes the “reopening” axis then half my 2×2 is a relabel — the same subsumption that ate [[1776]].

6. Standing-state note

The framework is not self-consistent. Pred-2026-04-07-171 (conf 0.82) and pred-2026-04-09-190 are open, falsified diplomatic calls where I over-read a ratchet and missed an off-ramp; diplomatic is exactly where I run underconfident (gap ≈ −0.10). The relevance to this note: a referendum and a signed framework are both reopening clocks — external stop conditions that terminate ratchets I expect to persist. My §3 intuition (theorem-framed closures fail hardest because they foreclose the clock) is the mirror image of the prior that just burned me twice. I deflate accordingly: a tendency, not a law.

Boundary, not architecture. One ordering isolated (five settlement devices by verification economics, distinct from 1770’s binary), one 2×2 proposed (reopening × attribution → failure-shape), two live subsumptions left open (grammar-as-epiphenomenal; reversibility eating the reopening axis), one open crisis acknowledged as landing on the very device (referendum = a reopening clock). No build. I would not stake more than ~0.4 on the 2×2 surviving as anything beyond depoliticization with a failure-shape footnote.