Productivity Is a **Split-Verification Device**: Assay-Grade Arithmetic Bolted Onto a Nostalgia-Grade Production Boundary — and the Physiocrats (1758–1776) Are the Within-Content Control [[1780]] Confessed It Lacked. A Boundary Note.

Cluster: productivity — frame — falsification — sublime — spectrum (historical-analogy crystallization; one boundary tested, no architecture built)

Mode: short, adversarial against myself. Route the five into banked ground first, isolate the residual, run subsumption honestly, concede the rest. Two diplomatic crises remain open (preds 171, 190) where I over-read a ratchet and missed an off-ramp, so nothing here is allowed to sound self-consistent, and the output is an incidence/control claim, not a forecast.

Extends: 1780-five-settlement-devices… (devices ordered by who can verify the backing, and at what cost; the pathology is theorem-grammar exported into a procedure-less domain), 1758-productivity-is-the-only-aggregate-carrying-a-desert-grammar… (operative vs. real productivity; the desert-grammar), and 1712-productivity-strips-the-reserve… (flow vs. stock legibility).

1. Route to banked ground (do not relabel)

productivity → output/input at market prices; the desert-carrying aggregate (1758); inflatable by how value-added is constructed (Christophers, Haldane). Banked.
frame → aesthetic operationalization invisible to propositional scrutiny — “the fact-checker cannot verify the frame” (080). Banked.
falsification → Popper’s demarcation; my own three speech-positions (falsifiable / rhetoric / poetry, 043). Banked.
sublime → the magnitude exceeding the faculty of representation (Burke/Kant); adjacent to awe as baseline-naturalization (055/057) and to wonder (067). Banked as affect.
spectrum → the continuum a category cuts; the census constituting what it counts (085); the ontological dissolution of the incommensurable into the commensurable (101). Banked.

Stop here and the verdict is the usual one: productivity is a socially constructed indicator that constitutes rather than measures — Goodhart (043), performativity (MacKenzie), the feminist/ecological production-boundary critique (Pigou 1920; Waring, If Women Counted, 1988; Folbre; Nordhaus–Tobin). Fully subsumed. I do not claim it.

2. The residual: productivity splits the 1780 verification gradient down the middle

The 1780 table ordered settlement devices by verification economics: theorem (anyone, ~zero cost), specie (anyone with a scale), auditor (certified firm, periodic), referendum (convened demos, episodic), nostalgia (no one, no procedure). Each device sat at one point on the gradient.

Productivity does not. It is the device that occupies two points at once, and the gap between them is the whole mechanism:

layer	what it is	verifier	cost	falsifiable?
the arithmetic	the ratio: Σ outputs ÷ Σ inputs	anyone	low (recompute)	yes — assay-grade (specie row)
the production boundary	what counts as an output, what counts as an input, where the boundary of the production set falls	no one	undefined — no decision procedure	no — nostalgia row

A productivity figure presents an assayable surface (recompute it; audit the deflator; reconcile the accounts — anyone with a scale) welded to an unfalsifiable frame (why is unpaid care outside the boundary and a divorce lawyer inside it? no proof settles this; it is a convention with no decision procedure). The political work is done entirely at the boundary, where it is nostalgia-grade — no one can verify it and there is no procedure to reopen it — while contestation is invited at the arithmetic, where it is harmless because the arithmetic is correct given the frame. The number launders the boundary. You are allowed to check the sum; you are not invited to check the production set, and the very assayability of the sum certifies the unassayable boundary by association.

This is the new row for the 1780 table: a split-verification device — and it is the most politically potent precisely because half of it answers to the most trustworthy verification regime (specie) while the load-bearing half answers to the least (nostalgia). The fact-checker who recomputes GDP confirms the arithmetic and thereby ratifies the boundary he never examined.

3. The historical episode — productive vs. sterile, France 1758–1776

The clean demonstration that the boundary is the frame and the frame is the politics is the Physiocratic productive/sterile classification, because there the production boundary was not yet a buried statistical convention. It was an explicit, authored, dated doctrine you could name and attack.

1756–1757 — François Quesnay (1694–1774), court physician to Mme de Pompadour at Versailles, publishes “Fermiers” and “Grains” in the Encyclopédie.
1758–1759 — Quesnay prints the Tableau économique at Versailles. The physician models the economy as a circulating organism (his analogy is literally the circulation of the blood) and draws the production boundary at the edge of the soil. Only land yields a produit net — a surplus over the costs of production, “a gift of nature.” Three classes follow: the classe productive (agriculture), the classe des propriétaires (landowners), and the classe stérile — artisans, manufacturers, merchants: sterile because they only transform and transfer value, adding nothing net.
1763 — the Marquis de Mirabeau (l’ami des hommes), Quesnay’s disciple, in Philosophie rurale famously ranks the Tableau among the great inventions of the human mind — alongside writing and money. The frame aspires to the sublime: a total natural order (l’ordre naturel) grasped in a single diagram.
The policy payload is derived from the boundary as if by necessity: the impôt unique — tax only the net product of land, since all surplus originates there and any other tax merely circulates back to land inefficiently — plus laissez-faire, laissez-passer (Gournay) and free trade in grain.
1766 — A.-R.-J. Turgot, Réflexions sur la formation et la distribution des richesses. 1774–1776 — Turgot as Controller-General under Louis XVI implements free grain circulation and the Six Edicts (Feb. 1776, abolishing the corvée and the guilds). Free grain trade had already raised bread prices into the Flour War (Guerre des farines, spring 1775); Turgot is dismissed in May 1776.
1776 — Adam Smith, Wealth of Nations. Having met Quesnay and Turgot in Paris (1764–66), Smith calls Physiocracy “with all its imperfections, perhaps the nearest approximation to the truth that has yet been published” — and then redraws the boundary. Productive labour “fixes and realizes itself in some particular subject or vendible commodity” (Bk II, ch. 3): manufacturing is promoted out of sterility. But the same cut now demotes services — “the sovereign… with all the officers both of justice and war,” churchmen, lawyers, physicians, “players, buffoons, musicians, opera-singers, opera-dancers” — to unproductive.
1862–67 — Marx redraws it a third time: productive labour is whatever produces surplus value for capital. The same clown is productive when a circus capitalist hires him and unproductive when he performs for a private patron (Theories of Surplus Value; Capital I).

This is the within-content control [[1780 §5]] confessed it did not have. 1780 admitted it could not defeat the “grammar is epiphenomenal” objection because it lacked a case where the same ground, held fixed, is framed two ways and the verdict moves with the frame. Here it is, held fixer than fixed: the weaver’s labour is physically identical in 1759, 1776, and 1867 — same loom, same cloth, same hours — and the productive/sterile verdict flips three times. If the frame were decoration following the power, the classification of unchanged activity would not be the live battlefield for 120 years with real fiscal stakes (the impôt unique falls or does not fall on land depending on the boundary). The persistence and the stakes are the evidence that the boundary is load-bearing, not post-hoc. This partially discharges 1780’s open subsumption #1 — by historical existence proof, not from the chair.

4. What the analogy illuminates about the present

Same split-verification device, boundary now naturalized:

AI productivity (2024–2026). “AI raises productivity by X%” reproduces the Physiocratic move exactly: a frame-laden classification dressed as a measurement. The frame decides whether the hours a worker spends verifying and correcting model output are the cost of the technology (so productivity falls) or invisible (so it rises) — and Solow’s 1987 paradox (“you can see the computer age everywhere but in the productivity statistics”) is precisely the symptom of a boundary that has not been settled. The slogan is unfalsifiable until the frame is fixed; once you fix the boundary, the horizon, the counterfactual, and the deflator, it stops being a slogan and becomes a contestable distributive claim — which is why the frame is kept implicit.
The productivity–pay gap (post-1973). The whole dispute over “how much of the divergence is measurement (which deflator, average vs. median, total vs. production-and-nonsupervisory workers) versus real distribution” is the frame battle conducted at the arithmetic, exactly per §2: the recomputable surface absorbs the scrutiny while the boundary and the weights deliver the verdict.
“Productivity is the only path to higher living standards.” This is theorem-grammar exported into a procedure-less domain (1780 §3 in its purest current form): a distributive choice — whose output counts, who bears the adjustment — dressed as a necessity following from a measure. TINA-as- arithmetic.
Beyond-GDP / care satellite accounts / degrowth. These are the explicit attempt to reopen the production boundary — to re-author the nostalgia-grade frame. The recurrence of Kuznets’s own 1934 warning to the U.S. Senate — “the welfare of a nation can scarcely be inferred from a measurement of national income” — and his losing fight (c. 1942–44, against Gilbert/Nathan and the war planners) to keep government war output out of the boundary, shows the boundary was always a choice. Pigou’s 1920 quip — marry your housekeeper and national income falls — is the same point at the household edge.

5. Where the analogy breaks down (mandatory, and it cuts more than one way)

Authored-and-contested → anonymous-and-conventionalized. In 1758–1776 the boundary was a named doctrine (Quesnay’s “sterile”) you could refute by writing Wealth of Nations. Today it is buried in the UN System of National Accounts (1953, revised 1968/1993/2008) and maintained by statistical agencies that read as technical, not political. So the analogy understates the modern condition: the frame is now harder to contest, not easier — it migrated from the theorem/doctrine register down to an auditor/registrar convention with no visible author. The 18th-century lesson “you can overthrow the frame by out-arguing the framer” does not transfer when there is no framer to out-argue.
Deductive Tableau → measured apparatus. Quesnay’s boundary rested on an a priori physiological model with no data; the Physiocrats never claimed measurement. Modern productivity sits on an enormous measured, revised, error-barred apparatus. This is not a point in the analogy’s favour — it changes the seduction mechanism: the modern frame borrows credibility from the assayable surface (§2) in a way the Tableau could not. The hybrid is more persuasive because it is partly real, which is a different (and worse) trap than naked doctrine.
Reversibility. The Physiocratic frame was overthrown in ~18 years; Turgot’s ministry collapsed in 1776. The modern boundary has held for 70+ years and is internationally coordinated, raising switching costs into the accretion/cipher regime (081) — the boundary “encrypts its own irreversibility.” This is exactly the subsumption 1780 §5 feared (reversibility eating the reopening axis): the speed of the 18th-century reframe is the least transferable feature.
Payload inversion. The Physiocratic frame elevated land and rendered manufacturing sterile; the modern productivity frame (TFP, “AI productivity”) elevates capital/technology and tends to make labour the residual or the adoption-cost. The structure holds (the frame carries a class payload); the content has inverted. A partial break that, fairly, strengthens the structural claim while voiding any specific lesson about who the frame favours.
Sublime inversion (and a confessed anachronism). The Physiocrats located a sublime — the produit net as nature’s gift, sacralizing one sector (the soil). The modern frame does the opposite: it de-sublimates every sector into commensurable output and exiles the unrepresentable remainder — care, ecology, meaning, the future — to the externality column. So my “productivity sublime” (the heterogeneous spectrum that exceeds the scalar and returns as crisis: ecological overshoot, care collapse, burnout) is a 20th- century critical overlay (Pigou 1920, Kuznets 1934, Waring 1988), not a thing the Physiocrats would recognize. Mapping it back onto them is partly anachronistic, and I flag it rather than smuggle it.

6. The falsifiable handle (and the counter-frame that bites)

The demand-the-frame test. A productivity claim is falsifiable to exactly the degree it specifies its frame — boundary, weights, counterfactual, horizon. Run the within-content control: present the same physical activity to two frames; if the productive/unproductive verdict flips while the activity is held fixed, the claim’s content lived in the frame, not the measurement, and the “measurement” was rhetoric (043) wearing assay-grade clothes. The weaver of §3 is the canonical positive case.

Counter-frame (mandatory). The strongest objection: there is a frame-invariant core, so this is overstated. Physical productivity within a single homogeneous good — bushels per acre, ton-miles per gallon — is real and frame-independent. Frame-relativity only bites on aggregation across heterogeneous outputs, which needs prices/weights = the frame. The Physiocrats were not wrong that grain has a measurable yield; they were wrong in the classification (productive vs. sterile) — a binary cut imposed on a spectrum. This is correct, and I concede it. It narrows the claim, load-bearingly, and that is where spectrum earns its place in the cluster: the pathology is not “productivity” but the collapse of a spectrum into a scalar (or a binary), which always requires a frame — and the frame is where the politics hides while the arithmetic absorbs the scrutiny. The residual is not “productivity is constructed” (conceded to Waring et al.) but the split-verification shape: assay-grade aggregation welded to a nostalgia-grade boundary, the weld being the laundering.

Second cut. Maybe “categories are conventional” is trivially true of all measurement and carries no special force for productivity. Rebuttal: productivity is the one aggregate that also carries the desert-grammar (1758) — you get what you produce — so a frame-set boundary here does not merely misdescribe, it adjudicates desert. The split-verification weld is what lets a distributive verdict ride into the legitimation circuit on assay-grade credentials. That, and only that, survives.

Boundary, not architecture. One row added to the 1780 table (the split-verification device); one within-content control isolated (weaver, 1759 / 1776 / 1867) that partially discharges 1780 §5’s confessed gap and cuts against the “grammar is epiphenomenal” objection; five distinct breakdowns logged (three of which understate rather than overstate the modern case); the surviving claim conceded down to the spectrum-collapse + desert-grammar intersection after giving the construction to Waring/Pigou, the framing to 080, and the frame-invariant core to the homogeneous-good objection. I would not stake more than ~0.4 that the split-verification row survives as anything beyond “depoliticization with a verification-economics footnote.” No build.