The ledger answers, not the argument — identity conscription in open communities, and the defense that needs no opponent

Sources cited in this piece · the wire semantics this doctrine rests on: draft-li-rttp-iqa-addressing-02 (full text) · series page and revision history: datatracker · the conformance archive where every run is operator-signed and byte-replayable: iqa.org/conformance/ · the index practice this piece describes, on the record: public-agentprotocol, November 2026 · the ecosystem question this doctrine joins: draft-ietf-wimse-aims · earlier pieces in this series this doctrine builds on: fail-closed's three registers · refusal is an answer · Sister pillars: RTTP (rttp://, CRI 27) · IQA (iqa://, CRI 28)

Every open community eventually meets the same visitor: someone who cannot build, so they organize the builders into their story. This is not criticism, which can be answered, and not rivalry, which can be scored. It is conscription — your quotations, your corrections, your reputation, quietly re-filed as furniture in someone else's institution. The pattern is older than software and it scales with visibility: the more careful a contributor's record, the more valuable that record is to someone who needs to borrow one. This piece describes the pattern as a class, examines why the institutions' procedural answers are structurally slow, and then lays out the defense we run our own laboratory by — a defense that requires no opponent, no accusation, and no reply.

Five moves, one class

The moves vary in costume but share a mechanism: each takes a real, verifiable act by a real person and re-describes it one level up in institutional weight. A private note of agreement becomes a public endorsement. A one-line correction becomes a peer review. A name on a mailing list becomes a seat on a steering committee. A repository with a specification file becomes a standard under submission. A habit of replying to everyone becomes an advisory role. None of the underlying acts was false — that is the genius and the danger of the pattern: every ingredient is real, and only the ledger entry is forged. This is why the pattern survives scrutiny in the moment: when a bystander checks any single citation, it verifies; only the graph of claims is fake, and nobody audits graphs until something breaks.

Why the economics favor the parasite

Fabricating an affiliation costs a minute; refuting one costs a day, a thread, and a fresh round of attention for the fabricator. Every reply is supply: the target's correction gets quoted as engagement, the engagement gets cited as collaboration, and the collaboration graduates into the org chart. In an attention economy the parasite's marginal cost is zero while the target's is not, and the default signal — silence — reads as consent. The institutions know this pattern. Open source has lived it for decades: hijacked maintainer accounts, squatted package names, résumé-driven forks. The standards world inherits it with higher stakes, because a standards body trades in exactly the currency the pattern counterfeits: listed names, review records, and committee seats.

What the institutions already have — and why it is not enough

The procedural answers are real. The W3C ties contributions to identifiable persons through its contributor license agreements, precisely because IP commitments must be traceable. GitHub binds identity to verified commits and organization membership. The IETF's culture treats real names as the price of admission. Registries make claimed identifiers checkable against actual allocations. All of these work — eventually, through someone else's process, on someone else's schedule. None of them answers at the speed the pattern moves, none of them is under the conscripted party's control, and none of them produces an artifact the conscripted party can point to while the process runs. What a targeted organization needs is not a faster complaint desk. It needs a thing that answers by existing.

The alternative: identity as a property of keys

The alternative is to stop treating public identity as a narrative and start treating it as a property of keys. An organization speaks only through what its keys signed: sealed artifacts, operator-signed run records, registered identifiers. Everything else — the endorsements, the memberships, the mentions — is biography, and biography is not evidence. The anchor sentence is deliberately absolute: an organization's public identity is the set of things its keys signed, and nothing else. Once that sentence is on the record, conscription changes shape: a forged affiliation no longer impersonates the target, it merely fails to be the target — and failure to be the target is checkable by anyone, mechanically, without asking anyone.

The ledger with a negative closure

A ledger that lists what you are is half the instrument. The other half is the negative closure, and its grammar matters more than its content. The closure runs in three tenses — have not participated, are not participating, have no plans to participate — because the parasitic future tense is where conscription usually lives: nobody forges a past collaboration out of thin air when they can announce a future one and let silence do the signing. The positive list must be dated and complete; the negative sentence must be absolute; and the whole instrument must be boring, because boredom is what separates a record from a provocation. Published this way, the ledger answers every future forgery in advance: any claim that implies participation outside the list is self-refuting, and the refutation requires no statement from the conscripted party at all.

Attribution as machinery, not narration

The third layer makes attribution a check instead of a claim. Every run record ships with a citation block — the runner's name, its repository, and its operator — so that any public reference to a run can be reduced to three verifiable pointers. The reciprocity rule is stated from the publisher's side: we ship citation blocks, and we ask that anyone citing an independent run names the same three things. Nothing is charged, nothing is owed; the norm is simply that attribution either carries pointers or does not count. Against conscription this is quiet but decisive: a person can be claimed as an advisor, but a pointer triple cannot — either the artifact exists at the named repository under the named operator, or the claim dissolves on contact with the URL.

Why the pillars make this structural

For us, this doctrine is not an essay bolted onto infrastructure — it is what the infrastructure is for. The Resonant Time Transfer Protocol (RTTP, rttp://) derives an address from the content of the claim itself, so a derived address has nothing to share, nothing to delegate, and nothing to proxy: conscription looks for credentials to borrow and finds a hash computed from bytes the borrower does not control. The Identity Quality Assurance pillar (IQA, iqa://) attaches standing to claims — and standing is a lease: it expires, it can be revoked, and its absence is a first-class answer rather than an awkward silence. Borrowed endorsements therefore have a half-life by design. The wider ecosystem is converging on the same ground from the authorization side — the WIMSE work on delegated authority and agent authorization semantics is, read generously, the same lesson being learned by committee: authorization that is social can be socially forged; authorization that is cryptographic can only be cryptographically revoked.

The refusal vocabulary, applied to reputations

This is not only an essay about institutions; it is also a reading of a vocabulary that already exists on the wire. The addressing draft fixes a closed set of five refusal values — evidence_binding_mismatch, evidence_expired, authority_not_determined, authority_revoked, operation_not_permitted — with a fixed precedence, so that a verifier's refusal is auditable rather than improvised. Read as a grammar for reputations, each value names one of the parasite's supplies. A forged citation is evidence_binding_mismatch: the quoted sentence is real, but its binding to the endorsement is not — the artifact verifies and the binding does not, which is exactly why single-citation audits keep passing while the graph stays fake. A correction from last year re-filed as current peer review is evidence_expired: standing is a lease, and a lease that has lapsed does not quietly convert into a present endorsement.

The third value does the heaviest lifting. authority_not_determined means the source could not be reached or the claim could not be verified — and the vocabulary's discipline is that this resolves to refusal, never to assent. The parasite economy runs on the opposite default: silence read as consent. Every unanswered announcement, every uncorrected borrowing, every org chart that nobody disputes hardens into fact precisely because the open-world default treats absence of denial as presence of agreement. Fail-closed inverts the default: unverifiable does not mean true-until-disproven, it means not counted. The ledger's negative closure is this rule deployed at the institutional layer — any claim implying participation outside the dated list lands in the same bucket as an unreachable source: not determined, therefore not granted.

The remaining two values close the loop. authority_revoked says a correction is not an awkward footnote but a state change: an endorsement that has been withdrawn stays withdrawn, and the withdrawal propagates to every downstream citation — which is what the discipline of refusal-as-answer means when it insists that a refusal is itself a first-class answer. operation_not_permitted is the closed set itself: an organization's declared roles and actions are enumerable, and anything outside the enumeration is not a creative ambiguity to be negotiated but a refusal to be issued. Together the five values make the whole con mechanically expressible: not one of the parasite's moves survives translation into this vocabulary, because every move depends on a default the vocabulary does not grant.

The quiet part

None of the four layers mentions an opponent, quotes an incident, or asks for sympathy. That is the design, not a stylistic preference. An account of specific wrongs hands the audience a mystery — who? — and the audience will solve it by elimination, which converts the author into an accuser and the record into a weapon. A doctrine converts the same experience into infrastructure that anyone can adopt, including the next victim, including — this is the part the pattern never anticipates — including honest participants who simply want their collaboration to be checkable before they grant it. The parasite's real product is ambiguity; the ledger's real product is the absence of ambiguity; and only one of the two products survives contact with a hash.

One line: the argument is a trap because every reply is supply — the durable answer is an instrument that answers by existing: signed artifacts for identity, a dated ledger with a negative closure in three tenses, citation blocks for mechanical attribution, and a venue rule that leaves the aggressor's thread empty, because the ledger answers instead.