A protocol for representing value before compressing it into a number, an audit format for the compressions it permits, and a first application that returns no total.
That commensuration discards information was established in the sociology of quantification a quarter-century ago. Espeland and Stevens, writing in 1998, put it in one sentence: commensuration is a system for discarding information and organizing what remains into new forms, with the link between representation and the empirical world obscured in the process.
That some goods are constituted partly by their not being exchangeable is Raz, 1986. That comparability without ranking is a genuine relation is Chang, 1997. That public measures recreate the social worlds they describe is Espeland and Sauder, 2007.
Its foundational rule is an ordering claim: value → evidence → representation → compression, and the ordering is not commutative. Numbers are permitted throughout; what is withheld from them is ontological priority.
The protocol operates on an explicitly evidence-bounded field and carries the difference from any total field as declared remainder rather than hiding it as residual error. This is where it departs from ordinary practice most sharply. Conventional method treats what the model does not capture as uncertainty or a stated limitation; the protocol treats the remainder as constitutive and non-additive, and requires it to be carried rather than minimized.
Two of its fifteen invariants are worth stating here. Unknown is not zero: an unmeasured value is entered as unresolved and carried in the remainder. And inscription integrity: a valuation can declare its commensuration contract, refuse unknown-to-zero, keep its remainder open and grant no primacy to price — and still rest on arithmetic that does not hold.
reception ⇒ productive uptake ⇒ dependency — and the arrows do not run. An external system encountering or summarizing an object is reception. An operator of that object becoming productive in a new context is uptake. A later capacity materially depending on the object is dependency, and requires counterfactual degradation or costly substitution, never resemblance.
Which yields a requirement that inverts the ordinary instinct. Productive uptake is separable from independent convergence only through distinctions the source supplies and nothing else in the field does. For persuasion one leads with one’s strongest claims; for measuring transmission one leads with one’s most singular, because only those carry a signature. A claim the field would have reached anyway is worthless as evidence of transmission, however true.
The valuer’s relation is declared as the dossier’s zeroth item: author of the object, author of the protocol, author of the inscription audited, not blind, material interest stated. That does not void the valuation. It weakens it, and the weakness is a recorded property rather than a reader’s inference.
What remains owed is stated at the close, and the first item cannot be discharged from here: the protocol’s only application to date was performed by the party who authored the object, the protocol and the inscription audited. Its own standard for validity — independent valuers reaching reproducible representations from the same evidence — has not been tested once.