WHERE DOES THE FORGERY CORE WIN? Goodman, Goodhart, Holmström, Dawkins and Sixteen Other Rivals: A Criterion-Fixed Adjudication Across Twenty Domains

“Five wins, eleven places, four scratched. It’s not bad at all. The horse that runs in twenty races and claims twenty has told you nothing about any of them.” Highlights The forgery core is strongest where it generalises: one operation — form conserved, origin substituted, substitution undetectable in the mode of presentation — covering a painting, a metric, a coerced virtue and a mimic’s signal, with consequences that differ by case and are derived rather than described. It weakens wherever a mature field already has the special case with better instruments, or a formal literature has modelled the same cell under another name. That pattern is the result, and the eleven domains where the priority is shared are not a defeat but the shape of a real advantage. The Verdict at a Glance Twenty domains, each assessed against a named rival on criteria fixed before any comparison was run. The full map is §2. Tier Verdict Domains A Advantage strongly defensible What a forgery is · Why fidelity aggravates · How forgeries end · The measurement of character · Channel selection under capture B Genuine contribution, close antecedent, priority shared Where counterfeits are built · Signalling and the price of heredity · Mimicry, divergence and convergence · Internalisation and the counterfeit self · Self-deception · Character and situationism · Credentials and testing · Crypto-religion and absorption · Molecular domestication · The verification spiral · Legitimacy by duration C Claim not sustainable as an advantage Stemmatics and provenance reconstruction · Cryptographic authentication · Fraud, reliance and the law · Connoisseurship and technical authentication Abstract The programme’s forgery cluster — nine articles from Authorship and Its Forgery (#108) to Virtue after Forgery (#118) — claims to have found one structure behind a painted Vermeer, a gamed metric, a coerced apology, a cuckoo’s egg and a domesticated transposon. This paper asks the only question that makes such a claim assessable: in which domain, against which named rival, for which specific reason is it better — and where is it not? Ten criteria are fixed in advance, three of them chosen because the cluster fails them. Twenty domains are assessed. The verdict is a distribution. In five domains the advantage is strongly defensible. The field of art forgery has a special case where the cluster has a general structure with a tense rule that says which endings are open (Goodman, Lessing, Dutton). The aesthetics of forgery predicts that skill mitigates; the Competence Theorem predicts that damage rises with fidelity, and says why. No literature has a general account of how forgeries end; the Exit Theorem supplies four exits and a rule fixing which are open. The psychometric and situationist traditions argue about instruments; the three-channel measurement failure derives why any instrument reading a form must fail. And where Goodhart has one channel, the Capture Law has three, selected conditionally. In eleven domains the contribution is genuine and the priority is not. This paper establishes one concession the corpus has not made: Meyer and Rowan (1977) described ceremonial conformity, and the decoupling of formal structure from actual activity, forty-nine years before the programme’s Ceremony channel. They, and not Goodhart, are the rival that channel had to beat, and the corpus never named them. What survives as the programme’s own is the conditional selection rule, not the phenomenon. Three further antecedents are conceded here. Holmström (1979) modelled the cell the Decoy Placement Rule names — a principal who contracts on an observed signal while the action behind it stays unobservable — which moves that domain out of Tier A; what remains is the three destinations, the withdrawal channel above all, and the transfer to substrates with no principal and no contract. Anderson (2001) is conceded for the verification spiral, and Spence (1973) with Collins (1979) for the credential case. In four domains the claim is withdrawn as an advantage. Textual criticism reconstructed descent from shared errors a century before the Error-Provenance Lemma; cryptography solved origin-verification with mechanisms the theory can redescribe but not improve; the law of fraud has centred reliance since long before T5; and technical authentication already works by inspecting the axes the forger did not. The pattern. The core wins where a field has a case and it has a structure, or where a field has a description and it has a conditional prediction. It loses where a field has an instrument, or a model of the same cell. Not one of the six separating observations in §2.1 has been made.