Skip to content

Off the Record

Treat a source's condition as a post-transmission use-license: the information is transmitted and the journalist's orientation affected, but a defined class of downstream speech acts — publication, attribution, citation as evidence — is barred until the source upgrades or an independent source confirms.

Core Idea

"Off the record" is a use-license agreement attached to a piece of information at the moment of its transfer from source to journalist: the source conveys the information on the condition that the journalist not publish it, not attribute it to any source, and — in the strictest reading — not cite it as evidence in any public record. The information is transmitted; what the license constrains is the class of downstream speech acts the recipient may perform with it. The license is therefore a post-transmission restriction: the journalist's beliefs and orientation are affected by the information, and the license leaves this untouched; what it governs is attribution and publication. A journalist who receives off-the-record information may use it to orient their reporting — framing questions to other sources, identifying which documents to request, knowing which leads to pursue — but may not publish the content with or without attribution unless a second, on-the-record source independently confirms the same facts. This lead-versus-evidence distinction is the operational core of the journalism practice. The license is asymmetric: the source may later go on the record, upgrading the license; the journalist has no unilateral path to do so. The journalism attribution ladder — on the record, on background, on deep background, off the record — is the professional vocabulary for negotiating the license terms at the start of an interaction, with each tier permitting a progressively narrower class of downstream uses. Breach modes are three: direct breach (publishing anyway), drift (misclassifying off-the-record as background and attributing to a vague category), and laundering (passing the information to a colleague or second contact who then reports it as though from a separate source).

Structural Signature

Sig role-phrases:

  • the source — the party originating the information and attaching the license at the moment of transfer
  • the journalist-recipient — the party who receives the information, whose beliefs and orientation are permanently affected but whose downstream public acts are bound
  • the post-transmission license — the restriction that survives the transfer, governing not what the recipient knows but the class of speech acts they may perform with it
  • the attribution ladder — the four discrete tiers (on the record, on background, on deep background, off the record), each licensing a progressively narrower class of downstream uses, settled up front
  • the lead-versus-evidence boundary — the operative line: the tip may steer questions, document requests, and leads, but may not be the cause of a published claim absent independent confirmation
  • the upgrade asymmetry — only the source holds the path to go on the record; the recipient has no unilateral route to lift the restriction
  • the independent-confirmation path — a second on-the-record source supplying a separate, publishable basis so the content can appear without breaching the original license
  • the three breach modes — direct breach (publishing anyway), drift (reclassifying off-the-record as background), laundering (routing through a colleague to report as if from a separate source)

What It Is Not

  • Not buying silence or controlling the reporter's mind. The information is transmitted and the journalist's beliefs and orientation are permanently affected; the license leaves all of that untouched. Knowing is not doing — what is bound is a specific class of downstream public moves (publication, attribution, citation as evidence), not what the recipient thinks or how it shapes their reporting. The source buys constraint on speech acts, not secrecy in the reporter's head.
  • Not "you cannot use it at all." The operative line is lead versus evidence: an off-the-record tip may legitimately steer which questions get asked, which documents get requested, which leads get run — it just cannot be the cause of a published claim absent independent on-the-record confirmation. So "can I use this?" is the wrong question (always yes, as a lead); the right one is "can this appear, and can it be attributed?"
  • Not a vague promise of confidentiality. It is a specific use-license with parties, terms, exceptions, an upgrade path, and breach remedies — a contract-like object, not a general gesture toward discretion. Reading it as a fuzzy assurance hides the lead/evidence boundary and the three distinct breach modes that the license frame makes crisp.
  • Not the other ladder tiers. On background, deep background, and on the record each license a different class of downstream uses; off the record is the strictest, permitting publication or attribution of the content only after a separate basis arises. Drift — quietly reclassifying off-the-record material as background and attributing it to a vague category — is precisely a breach, not a tier the reporter may slide to.
  • Not unilaterally liftable by the reporter. The upgrade path is asymmetric: only the source may go on the record, and an independent on-the-record source may supply a separate publishable basis. Deciding the embargo has lapsed, or that the fact is "out there anyway," is breach, because the right to lift the restriction belongs to the source alone.
  • Not the broad confidentiality duty, consent, or access control. It is a specific license variety — information already transferred to a particular recipient may not be republished or attributed — not the general normative duty of confidentiality. Consent governs whether information may be collected; access control gatekeeps who receives it; off the record assumes receipt and governs downstream use.

Scope of Application

"Off the record" lives across the reporting contexts of journalism where a source negotiates the terms on which a journalist may use what they are told — wherever the attribution ladder governs a source-reporter exchange; its reach is within that domain. The structurally identical post-transmission use-licenses elsewhere — legal settlement privilege (FRE 408), "without prejudice" mediation, therapy confidentiality, intelligence handling caveats and the Chatham House Rule, peer-review non-disclosure, clinical-trial data-monitoring embargoes — are fellow instances of the general "use-license on information" template (carried by confidentiality, access_control, contract, and consent), not uses of this journalism construct, so they are analogy and stay out of the map.

  • Investigative and accountability reporting — leak-driven and document-based reporting where the most consequential tips arrive off the record and must be steered into independently confirmable claims.
  • Political and government reporting — official and staff sources negotiating background versus off-the-record terms before speaking, the up-front tier choice governing every downstream attribution.
  • National-security and intelligence reporting — sensitive disclosures conveyed under the strictest tier, where the lead-versus-evidence boundary and independent confirmation are the only publishable path.
  • Business and financial reporting — executives and insiders conveying market-sensitive information off the record, the upgrade asymmetry deciding when it can surface.
  • The attribution ladder as surrounding apparatus — on the record, on background, on deep background, and off the record, the four-tier vocabulary negotiated at the start of any source interaction across all beats.

Clarity

Reading "off the record" as a use-license rather than a vague promise of secrecy separates what the journalist now knows from what the journalist may do with what they know — and that separation dissolves the most common confusion about the practice, the belief that the source is buying silence or controlling the reporter's mind. The source controls neither: the information is transmitted, the journalist's orientation is permanently affected, and the license leaves all of that untouched. What it binds is a specific class of downstream speech acts — publication, attribution, citation as evidence — while leaving another class open. That is the operational heart of the practice, and the term makes it crisp as the lead-versus-evidence boundary: an off-the-record tip may legitimately steer which questions get asked, which documents get requested, which leads get run, yet it may not become the cause of a published claim unless a second, on-the-record source independently confirms the same facts. The sharp question for a working reporter is therefore not "can I use this?" — they always can, as a lead — but can this appear, and can it be attributed?, which the license answers and a bare promise of confidentiality does not.

The license frame also makes the practice's asymmetry and its failure surface legible. Because only the source holds the upgrade path — the source may go on the record; the journalist has no unilateral route to do so — the term clarifies why the negotiation happens up front and why the reporter cannot simply decide the embargo has lapsed. And it turns three otherwise blurry lapses into distinct, namable violations: direct breach (publishing anyway), drift (quietly reclassifying off-the-record as background and attributing to a vague category), and laundering (routing the information through a colleague or second contact so it can be reported as if from a separate source). Naming these as three different breaches of one license, rather than as degrees of "burning a source," lets the practitioner ask the diagnostic question the looser vocabulary hides: which term of the license is being violated, and through which channel?

Manages Complexity

Source-reporter exchanges could, in principle, carry an open-ended continuum of use conditions — who may be told, under what name, with how much hedging, in what venue, for how long — and a reporter renegotiating those terms afresh with every source, on every tip, would face an unmanageable space of bespoke arrangements. The journalism attribution ladder compresses that continuum into four discrete, named tiers — on the record, on background, on deep background, off the record — each licensing a progressively narrower class of downstream uses, so the parties settle a single tier up front and the whole bundle of permitted and forbidden speech acts follows from that one label for the information's lifetime in the reporter's hands. The deeper compression is the lead-versus-evidence boundary, which collapses the reporter's recurring question from a per-item judgment to a single binary. Because the license is a post-transmission restriction — it governs publication and attribution, not the reporter's beliefs or orientation — the reporter never has to ask "can I use this?" (the answer is always yes, as a lead: to frame questions, request documents, run leads) but only the one question the tier answers: can this appear, and can it be attributed? That single boundary sorts every downstream act cleanly into the licit (orientation) and the constrained (publication without independent on-the-record confirmation), so the operational rule reads off the tier rather than off the content. The frame also fixes a small, decidable structure over the failure surface. The license's asymmetry reduces to one fact the reporter tracks — only the source holds the upgrade path (the source may go on the record; the reporter has no unilateral route) — which settles why the terms are negotiated up front and why the reporter cannot declare the embargo lapsed. And the otherwise diffuse notion of "burning a source" resolves into three named, distinguishable breaches of one license: direct breach (publishing anyway), drift (reclassifying off-the-record as background and attributing to a vague category), and laundering (routing the information through a colleague so it can be reported as if from a separate source). Instead of an unbounded space of confidentiality arrangements and an amorphous sense of wrongdoing, the practitioner tracks one tier, one lead/evidence boundary, one upgrade asymmetry, and three breach channels — and reads off what may be published, what may be attributed, who can lift the restriction, and exactly which term is violated when something goes wrong.

Abstract Reasoning

Reading the practice as a use-license — with one tier, one lead/evidence boundary, one upgrade asymmetry, and three breach channels — turns a murky promise into a contract-like object whose inferences run off those terms.

Diagnostic — classify a lapse by which license term it violates and through which channel, and audit a published claim's provenance. The signature inference takes a suspected wrong and asks not "was a source burned?" but which term of the license is being violated, and through which channel? — sorting the lapse into three distinguishable breaches: direct breach (the content published anyway), drift (off-the-record quietly reclassified as background and attributed to a vague category), or laundering (the information routed through a colleague or second contact so it can be reported as if from a separate source). Each has a distinct signature in the published artifact — an attributed quote, a "sources familiar with" hedge, a suspiciously independent second source — so the breach type is read off how the information surfaced. A second diagnostic audits provenance to decide whether the license was in fact respected: a published claim is licit only if a second, on-the-record source was the cause of the publication and the off-the-record tip merely the lead, so the analyst inspects whether independent confirmation exists rather than whether the fact happens to be true.

Interventionist — negotiate the tier up front, and re-license publication only through the source's upgrade path or independent confirmation. The lever the license exposes is the tier choice settled at the start of the interaction: selecting on-record, background, deep background, or off-the-record fixes the entire bundle of permitted and forbidden downstream speech acts for the information's lifetime in the reporter's hands, so the intervention that controls everything downstream is the up-front negotiation, not a later judgment. Two paths re-open publication, and the construct predicts the effect of each: the source going on the record upgrades the license (the only route the asymmetry permits), and an independent on-the-record source confirming the same facts supplies a separate, publishable basis — in both cases the off-the-record content can now appear, by a path that did not breach the original license. The sharp negative prediction is that the reporter has no unilateral path: deciding the embargo has lapsed, or that the information is "out there anyway," is forecast to constitute breach because the upgrade right belongs to the source alone.

Boundary-drawing — knowing is not doing, the lead/evidence line bounds licit use, and the upgrade right is asymmetric. The license draws its central boundary between what the reporter now knows and what the reporter may do with it: the information is transmitted and the reporter's orientation permanently affected, and the license leaves all of that untouched, binding only a specific class of downstream speech acts (publication, attribution, citation as evidence). This bounds the source's control — they buy neither silence nor influence over the reporter's mind, only the downstream public moves. The operative inner boundary is lead-versus-evidence: an off-the-record tip may legitimately steer which questions are asked, which documents requested, which leads run, but may not become the cause of a published claim absent independent on-record confirmation — so "can I use this?" is ruled the wrong question (always yes, as a lead), replaced by "can this appear, and can it be attributed?" The asymmetry bounds who may move the line: only the source holds the upgrade path, which is why the terms must be set before transmission and why the reporter cannot declare the restriction expired.

Predictive — the contract terms forecast what is permitted, and the asymmetry forecasts where the embargo holds. Because the license has parties, terms, duration, exceptions, and a breach remedy, the analyst predicts the permitted and forbidden acts by reading those terms off the agreed tier rather than re-deriving them per item: the tier predicts what may be published and what may be attributed, the exception classes predict the carve-outs, and the breach modes predict the remedy. The upgrade asymmetry predicts the dynamics — the restriction holds until the source acts or independent confirmation arrives, and any publication preceding one of those is predicted to read, on provenance audit, as one of the three breaches.

Knowledge Transfer

Within journalism the construct transfers as mechanism, intact, across the reporting situations that share the attribution ladder. The contract-like reading (parties, terms, duration, exceptions, breach remedy), the four-tier ladder (on the record, on background, on deep background, off the record — each licensing a progressively narrower class of downstream speech acts, settled up front), the lead-versus-evidence boundary (an off-the-record tip may steer questions, document requests, and leads, but may not be the cause of a published claim absent independent on-the-record confirmation), the upgrade asymmetry (only the source holds the path to go on the record; the reporter has no unilateral route), and the three named breaches (direct breach, drift, laundering) all carry without translation. A reporter who has internalized one off-the-record exchange applies the same license to every source negotiation, reading off what may be published, what may be attributed, who can lift the restriction, and exactly which term is violated when something goes wrong. This is genuine within-domain mechanism transfer, and it is what places the practice in this layer.

Beyond journalism the honest account is a shared-abstract-mechanism one, and here the shared structure is genuinely strong: the same post-transmission use-license — information transmitted under a restriction on downstream use that survives the transmission — recurs across distinct substrates as real co-instances. Legal settlement discussions (FRE 408 and analogues making settlement communications inadmissible to prove liability; "without prejudice" communications and caucus confidentiality in mediation), therapy and clinical confidentiality (what is said in session, bound except for statutory carve-outs like duty to warn and mandated reporting), intelligence and diplomacy (classification and handling caveats — NOFORN, ORCON, the Chatham House Rule), academic peer review (manuscripts received under a pre-publication non-disclosure license), and clinical-trial data monitoring committees (interim data held under strict non-disclosure until unblinding) all instantiate the same license shape: prior agreement on use restrictions, defined exception classes, enforcement through professional norms and sometimes law. But the crucial honesty is that what recurs is the general license-attached-to-information template (a possible "use-license on information" family), not the journalism construct — and that template's substrate-independent residue factors cleanly into catalog primes that each travel on their own terms: the broader normative class of confidentiality, access_control (the structural cousin governing access where this governs use), contract and commitment (the license-as-agreement layer), consent (the participation-permission layer), and accountability (the breach-enforcement layer). So the cross-domain reasoner should carry that general template and its constituent primes, while recognizing that each domain's apparatus stays home-bound: the journalism on/background/deep-background/off ladder, the lead-versus-evidence rule, and the three breach modes are journalism-particular, just as FRE 408 and attorney-client/mediation privilege are legal-particular and the duty-to-warn carve-outs are clinical-particular. Calling a settlement privilege or a Chatham House Rule "off the record" borrows the journalism label for a fellow instance of the same template and is analogy to be marked — the shared object is the use-license, not the journalism construct.

The construct's defining boundary is part of what makes the shared structure recognizable across all these substrates: knowing is not doing. The information is transmitted and the recipient's orientation permanently affected; the license leaves all of that untouched and binds only a specific class of downstream public moves (publication, attribution, citation as evidence). The source buys neither silence nor influence over the recipient's mind — only the downstream acts — and that separation holds identically for a reporter, a therapist's client, a settlement counterparty, or a peer reviewer. The construct stays distinct from its catalog neighbors accordingly: it is not the broad normative duty of confidentiality but a specific license variety in which information transmitted to a particular recipient may not be republished or attributed; it is not consent (which governs whether information may be collected, where this governs what may be done with information already transferred); and it is not access_control (gatekeeping who receives information, where this assumes the information has already been received). Strip the journalism idiom and what remains is the general use-license-over-information template carried by confidentiality, access_control, contract, commitment, consent, and accountability — the boundary between this domain-specific abstraction and the substrate-independent mechanisms it composes (see Structural Core vs. Domain Accent).

Examples

Canonical

During the Watergate investigation (1972–74), Washington Post reporter Bob Woodward met a senior FBI source — revealed in 2005 to be Associate Director Mark Felt — whom the reporters nicknamed "Deep Throat." Felt spoke on deep background: Woodward could use what he learned to orient the reporting but could not attribute it to Felt or to the FBI, nor publish it as a sourced claim. Felt's guidance steered which leads to pursue and confirmed the direction of the investigation, while the Post's published stories rested on independent, separately sourced confirmation. Felt's identity was protected for more than three decades and lifted only when he himself chose to go public. The arrangement is the textbook enactment of the license: the information used as a lead, never as attributable evidence, its restriction liftable only by the source.

Mapped back: Felt is the source and Woodward the journalist-recipient; "deep background" is a rung of the attribution ladder, settled up front. Using the tips to know which leads to run while publishing only separately confirmed claims is exactly the lead-versus-evidence boundary. That the embargo held until Felt himself went public in 2005 is the upgrade asymmetry — the recipient never held a unilateral route.

Applied / In Practice

In day-to-day political journalism the ladder is negotiated explicitly before an interview begins. A White House or agency official routinely offers to brief "on background," and the resulting stories attribute the information to "a senior administration official" — a background tier, not off the record, but fixed by the same up-front tier choice. When a source insists on off the record, the reporter may use the guidance only to steer questions toward on-the-record officials and to know which documents to request. Newsroom standards manuals — at the Associated Press and The New York Times, among others — codify the ladder and police its failure modes: they discourage vague, drifting attribution and require editors to know the identity of anonymous sources, precisely to prevent laundering the material through a second contact. The abstract license becomes enforceable institutional practice.

Mapped back: The "on background / off the record" choice made before the interview is the attribution ladder negotiated up front, and using an off-the-record tip only to steer questions is the lead-versus-evidence boundary. The standards manuals' rules against vague attribution and undisclosed second sources are institutional guards against two of the three breach modes — drift and laundering.

Structural Tensions

T1: Knowing versus doing (the license binds acts but the mind is already changed). The license constrains a defined class of downstream speech acts — publication, attribution, citation — while leaving the reporter's beliefs and orientation permanently affected: knowing is not doing. That separation is the construct's power and its limit at once. For the reporter it is a real, enforceable restriction on public moves. But for the source it delivers less control than the phrase "off the record" seems to promise: the information is already transmitted, the reporting already steered, and there is no un-telling. A source who goes off the record to prevent a story may instead have caused it, having handed over a lead that reorients the entire investigation toward independent confirmation. The source buys constraint on speech acts, never secrecy in the reporter's head. Diagnostic: Is the protection sought a bar on publication and attribution (the license delivers it) or control over what the reporter now knows and pursues (it cannot)?

T2: Lead versus evidence (the same tip licitly steers and, but-for, grounds the claim). The operative boundary is that an off-the-record tip may steer which questions are asked and which documents requested, but may not be the cause of a published claim absent independent on-the-record confirmation. Clean in principle, the causal line blurs in practice: once a tip points the reporter at exactly the fact that a second source then confirms, the published claim is formally "caused" by the confirmation yet is but-for the original tip. How independent is a confirmation the off-the-record source directed you to find? The license depends on a provenance distinction — lead versus evidence — that the reporting process continuously threatens to collapse, since the whole value of the tip is to make the confirming evidence findable. Diagnostic: Would the published claim stand on its independent source alone, or is the "independent" confirmation merely the off-the-record tip re-sourced through a path the tip itself opened?

T3: Four discrete tiers versus the continuum of use conditions (what the ladder compresses). Source-reporter exchanges could carry an open-ended range of conditions — who may be told, under what description, in what venue, for how long — and the attribution ladder compresses all of it into four named tiers, so the parties settle one label up front and the whole bundle of permitted acts follows. That compression is what makes the practice tractable and the terms enforceable. But it also flattens distinctions a bespoke arrangement could draw: a source's actual intent may sit between "background" and "off the record," or attach conditions the ladder has no rung for, and forcing the exchange onto four tiers can misrepresent what was really agreed. The economy that makes the license readable is bought by discarding the finer terms real confidentiality negotiations sometimes need. Diagnostic: Does one of the four tiers capture what this source actually intends, or is the real agreement being distorted to fit a rung it does not match?

T4: Source trust versus reporter strand (the upgrade asymmetry cuts both ways). Only the source holds the path to lift the restriction; the reporter has no unilateral route, and deciding the fact is "out there anyway" is itself a breach. This asymmetry is load-bearing for the whole institution: sources speak only because they, and they alone, control whether their words can surface, so the reporter's powerlessness to lift is precisely what makes candor possible. The same asymmetry can strand the reporter holding a true, confirmed, public-interest fact that cannot be published because the source will not relent and no independent confirmation has arrived. The feature that secures the source's trust is the feature that can bury the story — and the reporter cannot resolve the bind without either the source's upgrade or a separate publishable basis. Diagnostic: Is the restriction still serving the trust that made the source speak, or has it become a lock on information the public interest needs and only the source can open?

T5: Crisp definition versus covert breach (the subtle violations are the invisible ones). The license frame turns "burning a source" into three named, distinguishable breaches — direct breach, drift, laundering. Definitionally crisp; but the three differ sharply in detectability. Direct breach leaves an attributed quote anyone can see. Drift (reclassifying off-the-record as background, attributing to a vague category) and laundering (routing the material through a colleague to report as if independently sourced) are engineered to be invisible in the published artifact — a "sources familiar with" hedge or a suspiciously independent second source is a signature only to those who already know the provenance. And enforcement rests largely on professional norms and newsroom standards manuals, not law, so the breaches hardest to detect are policed by self-regulation precisely where external verification is weakest. The clarity of the categories does not translate into the enforceability of the boundary. Diagnostic: Would this breach leave a trace in the published artifact, or is it a drift/laundering violation detectable only by someone who knows how the information actually reached print?

T6: Autonomy versus reduction (the journalism construct or the general use-license template). "Off the record" is a journalism-particular construct — the on/background/deep-background/off ladder, the lead-versus-evidence rule, the three breach modes — and within reporting it transfers as mechanism intact. But its cross-domain cousins (settlement privilege under FRE 408, "without prejudice" mediation, therapy confidentiality, intelligence handling caveats and the Chatham House Rule, peer-review non-disclosure) are not uses of the journalism construct; they are fellow instances of a general post-transmission use-license template, whose substrate-independent residue factors into catalog primes that travel on their own terms — confidentiality, access_control (governing receipt where this governs use), contract/commitment (the agreement layer), consent (the collection layer), and accountability (the enforcement layer). Calling a Chatham House Rule "off the record" borrows the journalism label for a sibling instance. Diagnostic: Resolve toward the general use-license template and its parents (confidentiality, access_control, contract, consent, accountability) when reasoning across domains; toward "off the record" when the attribution ladder and lead/evidence rule are the actual apparatus in play.

Structural–Framed Character

"Off the record" sits at the framed end of the structural–framed spectrum — framed-leaning: a journalism-particular use-license constituted by a source-reporter practice, contract-like in structure but a technique rather than a verdict. On evaluative_weight it is largely neutral as an agreement (parties, terms, an upgrade path), though its three breach modes carry a normative charge (breaching a license is a wrong), so it sits above a pure mechanism while below a censure-label. On human_practice_bound it is high: the license is constituted by the practice of source-reporter exchange under an attribution ladder and dissolves without it — with no journalist, no publication, and no attribution to bar, there is only a person telling another person something, nothing being licensed. Institutional_origin is pronounced: the on-record / on-background / deep-background / off-the-record ladder, the lead-versus-evidence rule, and the three named breaches (direct breach, drift, laundering) are journalism-ethics furniture codified in newsroom standards manuals, an apparatus of a professional tradition rather than a fact of nature. On vocab_travels it scores low: the attribution-ladder idiom is pinned to reporting. And on import_vs_recognize the transfer is bimodal — within journalism it ports as mechanism across investigative, political, security, and business reporting, but settlement privilege, Chatham House Rule, and therapy confidentiality are co-instances of a shared template, not imports of "off the record."

Here the portable content is a post-transmission use-license on information — the general template the construct co-instantiates with legal, clinical, intelligence, and peer-review confidentiality — and its substrate-independent residue factors cleanly into several catalog primes the entry genuinely rests on: confidentiality (the normative class), access_control (governing receipt where this governs use), contract/commitment (the agreement layer), consent (the collection layer), and accountability (the enforcement layer). Each travels on its own terms. But together they do not pull "off the record" off the framed pole, because that template is exactly what the construct instantiates from its umbrellas, not what makes "off the record" itself travel: the cross-domain reach belongs to the use-license template and its constituent primes, while the attribution ladder, the lead-versus-evidence rule, and the three breach modes stay home. Its character: a practice-constituted, contract-like journalism use-license, structural only in the post-transmission use-license template it borrows from its umbrellas and specializes to the newsroom's attribution ladder.

Structural Core vs. Domain Accent

This section decides why "off the record" is a domain-specific abstraction and not a prime — why its cross-domain reach belongs to a general template and its constituent primes while the newsroom apparatus stays home.

What is skeletal (could lift toward a cross-domain prime). Strip the journalism and a thin relational structure survives: a post-transmission use-license — information is transmitted under a restriction on downstream use that survives the transmission, so the recipient's knowledge and orientation are permanently affected while a defined class of downstream acts is barred until an agreed condition lifts the bar. The portable pieces are abstract: parties, terms, an upgrade path, exception classes, and a breach remedy, all attached to already-transferred information. This skeleton is genuinely substrate-portable, which is why its residue factors cleanly into catalog primes that each travel on their own terms — confidentiality (the normative class), access_control (governing receipt where this governs use), contract/commitment (the agreement layer), consent (the collection layer), and accountability (the enforcement layer). But this general "use-license on information" template is the core "off the record" shares with settlement privilege and Chatham House Rule, not what makes the journalism construct distinctive.

What is domain-bound. What is distinctive is journalism-ethics furniture and does not survive extraction. The attribution ladder — on the record, on background, on deep background, off the record — is a four-tier newsroom vocabulary; the lead-versus-evidence boundary (a tip may steer questions and document requests but may not be the cause of a published claim absent independent on-the-record confirmation) is a reporting-specific operative rule; the independent-confirmation path presupposes publishable sourcing; and the three breach modes (direct breach, drift, laundering) are named against newsroom sourcing practice and codified in standards manuals. The decisive test: settlement privilege (FRE 408), "without prejudice" mediation, therapy confidentiality, intelligence handling caveats, and peer-review non-disclosure are structurally identical use-licenses — but each has its own apparatus (FRE 408 is legal-particular, duty-to-warn carve-outs are clinical-particular), and calling any of them "off the record" borrows the journalism label for a sibling. Remove the reporting practice and there is no attribution ladder or lead/evidence rule, only the bare use-license template.

Why this does not clear the prime bar. A prime's vocabulary travels and its transfer is recognition of the same mechanism, not analogy. "Off the record"'s transfer is bimodal. Within journalism it moves as full mechanism — the contract-like reading, the four-tier ladder, the lead/evidence boundary, the upgrade asymmetry, and the three breach modes carry intact across investigative, political, national-security, and business reporting, because each is a source-reporter exchange under the same ladder (recognition, not analogy). Beyond journalism the cross-domain cousins are genuine co-instances of the general use-license template, not imports of the journalism construct: settlement privilege, mediation confidentiality, clinical confidentiality, intelligence caveats, and peer-review NDAs all instantiate prior-agreement-on-use-restrictions with defined exceptions and norm-or-law enforcement — sharing the object (the use-license) but not the newsroom apparatus. The genuinely portable structure is that general template and its constituent primes (confidentiality, access_control, contract, commitment, consent, accountability), each of which travels on its own terms. So the cross-domain reach belongs to the template and its parents; the disciplined move is to carry them when reasoning across domains, and reserve "off the record" for where the attribution ladder and lead/evidence rule are the actual apparatus in play. It clears the domain-specific bar comfortably for journalism, but its only substrate-spanning content is already carried, in more general form, by the template it composes.

Relationships to Other Abstractions

Local relationship map for Off the RecordParents appear above the current abstraction, mutual partners to the right, and children below. Node labels state whether each abstraction is prime or domain-specific; colors identify relation types.Off the RecordDOMAINPrime abstraction: Commitment — is a decomposition ofCommitmentPRIMEPrime abstraction: Information Use License — is a kind ofInformationUse LicensePRIME

Current abstraction Off the Record Domain-specific

Parents (2) — more general patterns this builds on

  • Off the Record is a kind of Information Use License Prime

    Off The Record is a strict specialization of Information Use License.

  • Off the Record is a decomposition of Commitment Prime

    Accepting information Off the Record is a present binding act that constrains the recipient's later publication and attribution, with release terms and breach consequences.

Hierarchy paths (2) — routes to 1 parentless root

Not to Be Confused With

  • On background / on deep background / on the record (the other ladder tiers). The adjacent rungs of the attribution ladder, each licensing a different class of downstream uses — background permits publication with vague attribution ("a senior official"), the record permits full named attribution. Off the record is the strictest tier, barring publication and attribution outright. Quietly reclassifying off-the-record material as background is drift — a breach, not a legitimate slide. Tell: may the content be published with some attribution now (background/on-record), or only after the source upgrades or an independent source confirms (off the record)?

  • Anonymous sourcing. Publishing a claim while withholding the source's name — "according to a person familiar with the matter." This is a background-tier arrangement in which the information does appear; off the record bars the content from appearing at all. Conflating them is exactly the drift breach. Tell: does the reporting publish the substance under an unnamed attribution (anonymous sourcing) or refrain from publishing the substance entirely (off the record)?

  • News embargo. A timing agreement: information may be published, but only after a set release time. Off the record governs whether and with what attribution the content may ever appear, not merely when. Tell: is the restriction a hold-until-a-time on publishable material (embargo) or a bar on publication/attribution that only the source or independent confirmation can lift (off the record)?

  • Reporter's privilege / shield laws. The legal right of a journalist to resist compelled disclosure of a confidential source's identity to courts or government. Off the record is the source-reporter agreement governing use of the information; the privilege is a legal protection against outside compulsion. Tell: is the issue an agreement between source and reporter about downstream use (off the record) or a legal defense against a subpoena demanding the source's name (reporter's privilege)?

  • Confidentiality (the broad duty). The general normative obligation not to disclose entrusted information. Off the record is a specific license variety within that broad class — information already transmitted to a particular recipient may not be republished or attributed — with defined parties, terms, an upgrade path, and breach modes, not a diffuse duty. Tell: is it a general obligation of discretion (confidentiality) or a specific contract-like use-license with an attribution ladder and breach remedies (off the record)?

  • The post-transmission use-license template (parent) and its co-instances. The substrate-general shape the construct instantiates — information transmitted under a downstream-use restriction that survives transmission — carried by confidentiality, access_control, contract, consent, and accountability. Settlement privilege (FRE 408), the Chatham House Rule, and therapy confidentiality are fellow co-instances of this template, not uses of the journalism construct. Tell: is the newsroom attribution ladder and lead/evidence rule the apparatus (off the record), or a use-license in another institution's clothing (a co-instance of the template)? (Treated more fully in Structural Core vs. Domain Accent.)

Neighborhood in Abstraction Space

Off the Record sits in a moderately populated region (46th percentile for distinctiveness): it has near-neighbors but no dense thicket of look-alikes.

Family — Journalistic Sourcing & Institutional Trust (13 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-07-12