Setup–Resolution Pair

Version

v4 · 2026-08-30 · History

Prime #

1455

Origin domain

Linguistics & Semiotics

Subdomain

ordered functional pairs → Linguistics & Semiotics

Core Idea

A setup–resolution pair is an ordered two-role relation in which an initiating element opens a demand of a specified type and a second, causally linked element discharges it, so that the two together — and neither alone — constitute one completed unit. ^[1] The initiating element does not merely come first; it fixes what would count as an adequate answer, which is why the relation supports a question that bare sequence cannot answer: at any moment in between, is the demand still outstanding? ^[2]

What gets opened may be an expectation, an obligation, a question, an instability, or a dependency, and each substrate renames it accordingly. A plant awaits a payoff. An allocation awaits a release. A declaration awaits a use. An assumption awaits a discharge. Beneath the vocabularies the invariant holds steady: something has been left owed, it stays owed across an interval, and only a particular kind of thing pays it.

Three commitments make the relation what it is, and dropping any one of them collapses it into something weaker.

Typing. The setup narrows the class of admissible successors. A gun on a mantel narrows the space of later events far less tightly than an Assume P narrows the space of later proof steps, but both narrow it, and that narrowing is what makes a later element checkable against an earlier one. A first element permitting anything at all to follow has opened nothing.

Liveness. The demand persists across the interval separating the two elements, and persistence is never free — something has to carry it. The carrier may be a scope table, a court docket, a reader's memory, a ledger row, or a bit of a machine's physical state, but the pair has no existence independent of some bearer holding the demand open. Where the bearer is lossy, the interval a pair can span is bounded by how long it keeps holding.

Linkage. The second element has to be attributable to that first one. Two candidate discharges of the same type are not interchangeable wherever the system tracks which setup each one answers: releasing another process's lock does not close your acquisition, paying an invoice does not settle a different invoice, and answering a question adjacent to the one asked leaves the asked one open.

Because the unit of evaluation is the pair, its two elements can each be individually unremarkable while the composite plainly succeeds or plainly fails. Pairs also nest and chain — a resolution can serve as the next setup, and long structures such as a proof, a protocol exchange, or the life of a bond are assembled from pairs whose closures interlock. The residue of a pair that never closes is characteristic and diagnosable in every substrate: a leak, a dangling reference, an unanswered request, an abandoned subplot, a liability that never settles.

Structural Signature

A typed demand is opened → the demand is held live across an interval by some bearer → a linked element discharges it → the pair, rather than either element, registers as complete.

Recurring features:

• An initiating element that constrains the class of adequate successors rather than merely preceding them.
• A liveness interval across which the demand is outstanding, with a definite bearer holding it open.
• An attribution relation strong enough to say which setup a given discharge answers.
• A closure test that mere succession fails: the second element must satisfy the demand, not simply arrive after it.
• Role asymmetry, such that exchanging the two positions destroys the unit rather than reversing it.
• A characteristic residue on non-closure — debt, danglement, or noise — detectable without knowing what the missing resolution would have been.
• Composability, whereby closed pairs nest and any resolution may double as the setup of a further pair.

This signature is what makes the pattern checkable in substrates that have no notion of story or satisfaction. A garbage collector, a proof assistant, a double-entry ledger, and a protocol state machine all implement the same four-step shape while sharing none of each other's vocabulary; what they share is that each maintains a set of currently open demands, each can name the type of thing that would empty a given slot in that set, and each treats a non-empty set at termination as an error condition rather than as an aesthetic disappointment. ^[3]

What It Is Not

The phrase gets used loosely in ordinary talk to mean roughly “first this, then that,” and that loose use is what most often misfires. Two events can be adjacent, correlated, even causally connected, and still fail to form a pair, because a cause constrains what will happen next without thereby specifying what would count as completing it. ^[1] Rain wets a road and the road later dries; nothing was owed and nothing was paid.

It is not a judgement about quality. A pair can close crudely and still be closed; the relation asserts that the demand was discharged, not that the discharge was elegant, surprising, or deserved. An exquisitely written scene answering no standing demand is a good scene sitting outside any pair.

It does not require an author, and it does not require intent. A programmer allocating inside a loop is not planting anything, yet each allocation opens a demand exactly as a deliberately planted detail does; the demand is opened by the operation, not by a plan. Deliberate emplacement is common and often valuable, but it is a fact about how a pair came to exist rather than part of what makes it one.

It makes no claim about distance. The interval may be a single instruction or a decade, and lengthening it changes the cost of holding the demand and the odds of the bearer losing it, but not the identity of the relation.

Nor does it belong to narrative, despite the vocabulary being borrowed from there most often. Narrative is one substrate among many, and it is an unusually permissive one, since a story can survive several unclosed pairs whereas a linker cannot survive one.

Finally, it is not a promise that any resolution exists. Setups are opened all the time that nothing will ever discharge, and one of the prime's main uses is precisely to make those visible while they are still open, rather than at the point where their absence has already done damage.

Broad Use

Programming and runtime systems. Every acquire-and-release discipline is an instance: malloc and free, open and close, lock and unlock, subscribe and unsubscribe, begin and commit. So is every binding construct — a declared variable opens a demand that some later use discharges, which is why compilers report both unused declarations and uses of undeclared names as two sides of one broken pair.

Protocols and distributed systems. A SYN awaits its ACK, a request awaits its response, a challenge awaits its signed answer, a two-phase-commit prepare awaits its global decision. Timeouts exist because these systems must decide, in finite time, what to do about a demand whose discharge has not arrived. ^[4]

Formal reasoning. Natural deduction literally calls the operation discharge: an assumption introduced for a conditional proof is retired by the introduction rule that consumes it. Lemmas stated for later use, quantifiers bound then instantiated, and proof obligations emitted by a verification condition generator all sit in the same slot.

Law, finance, and administration. An escrow deposit awaits its release condition, an issued bond awaits maturity, a warrant awaits its return, a subpoena awaits compliance or quashing, an accrued liability awaits settlement. Audit is largely the practice of enumerating pairs that have not closed.

Music and performance. An antecedent phrase projects a consequent, a dominant preparation projects a cadence, a stated motif projects a recapitulation. Deceptive resolutions are interesting because listeners hold the demand tightly enough to notice the substitution.

Narrative, comedy, and games. Plant and payoff, feed line and punchline, a stated rule of the world and the later scene that exploits it, a tutorial mechanic and the level that demands it.

Clinical and investigative work. A differential diagnosis raised is a demand a later test closes by confirming or excluding it; an open action item, a logged risk, and a filed hypothesis behave identically, which is why each of these practices grows a register of things still open. ^[5]

Clarity

The confusion this prime dissolves is between things that merely happen later and things that were owed. That distinction sounds obvious stated baldly, and it is systematically lost in practice, because both look identical in a transcript, a log, or a plot summary: in each case one element appears and then another does.

The clearest symptom is the argument in which one reviewer says an ending was unearned and another replies that it was set up. Both are usually describing the same earlier material. The disagreement is over whether that material typed the later event or merely permitted it — whether a reader holding the setup could have said in advance what class of thing would discharge it, or could only say afterwards that nothing had ruled this out. ^[6] Naming the demand and its type converts an argument about taste into a question with an answer.

The same confusion wears an engineering costume. “The file handle was closed” and “this open was discharged” are different assertions, and a program can satisfy the first while violating the second on some path; the vocabulary of pairs makes them separately askable and supplies the word for half a pair, which is danglement. In institutions it appears as the gap between an item marked done and an obligation actually retired.

Manages Complexity

Closure licenses forgetting, and that is the whole of what this prime buys. Once a pair has closed, the interval between its elements can be dropped from the working set: the details of how the demand was carried, what else happened while it was open, and which intermediate steps touched it stop mattering, because nothing downstream can depend on a demand that no longer exists. ^[7]

The quantity that must actually be tracked is therefore not the history of events but the far smaller set of currently open demands. A reader who has seen a subplot pay off can stop rehearsing it and still follow the book. A compiler that has seen a register's last use can reassign it. An accountant who has matched a payment to an invoice can close the line and reason about the remaining balance alone. In every case the burden scales with concurrent openness, not with elapsed length, which is why a very long structure with disciplined closure is easier to hold than a short one that leaves five things hanging.

This also converts a diffuse worry into a countable one. “Is this system in good order?” has no procedure attached to it; “which demands are open, how long have they been open, and who is holding each” is a query. Systems that take the prime seriously build that query into themselves — open-handle counts, unmatched-entry reports, aging registers of unresolved items — and thereby cap how much any individual participant has to remember.

Abstract Reasoning

The prime licenses a five-step diagnostic that runs identically in any substrate. Given a suspected pair, first ask what class of successors the candidate initiator admits; if the honest answer is “anything,” there is no setup and the analysis stops. Second, check liveness: was the demand still outstanding at the moment the candidate resolution arrived, or had it already lapsed, been withdrawn, or been discharged by something earlier? Third, test linkage counterfactually — remove the initiator and ask whether the second element still reads as a resolution of anything, then substitute a different element of the same type and ask whether it would have discharged the demand equally well. Fourth, examine the discharge itself for whether it closes the demand, downgrades it, or merely relocates it. Fifth, name the bearer that held the demand across the interval, since a pair with no identifiable bearer is a pair the analyst is supplying rather than observing.

Running that procedure yields inferences that would otherwise take domain expertise. A system whose count of openings exceeds its count of discharges is accruing debt at a measurable rate, whatever the openings happen to be called locally. A resolution with no locatable setup is either arbitrary or evidence that a setup went unnoticed, and distinguishing those two cases is a productive investigation rather than a matter of opinion. A resolution that could be swapped for a wholly unrelated one without anything breaking reveals that the setup was undertyped, which is a defect in the setup and not in the resolution. And when a demand is discharged far earlier than expected, the remaining interval is structurally empty and will read as slack.

Knowledge Transfer

What carries across substrates is the skeleton: ordered roles, a typed demand, a liveness interval with a named bearer, an attribution relation, a discharge test, and a register of what is still open. Whether an engineer who has internalized resource discipline could therefore audit a contract portfolio, or an editor who tracks unpaid plants read a protocol trace, is this entry's conjecture rather than a demonstrated result; what can be said is that the same six questions can be put in either vocabulary, and that the failure residue they share — a demand left open past its liveness interval — is in software a catalogued defect class. ^[8]

What does not carry is everything about enforcement and tolerance, and mistaking these for part of the pattern is the usual source of bad transfer. The affective payoff — satisfaction, relief, the click of a punchline — exists only where the bearer is an experiencing party, and a linker feels nothing when a symbol resolves. Enforcement differs by orders of magnitude: a runtime aborts, a proof checker refuses, a court awards damages, a reader shrugs and keeps reading. Tolerance for unclosed pairs differs just as sharply, and it is a design parameter rather than a defect, since serial fiction deliberately runs a permanent backlog of open demands while a filesystem may tolerate none. Uniqueness differs too: a lock release must match its own acquisition exactly, whereas several quite different scenes could each honestly discharge the same narrative promise. So does the reliability of the bearer, which is what actually decides how far apart the two elements can sit. Transfer the skeleton; re-derive the enforcement regime, the tolerance budget, and the bearer's memory span locally every time.

Examples

Formal/abstract

Conditional proof in natural deduction. A prover working toward P → Q writes Assume P and opens a subproof. From that line onward, every step inside the subproof carries a dependency on something not yet earned: each is true only relative to P. The assumption is a setup whose demand is typed about as tightly as a demand can be — exactly one rule, implication introduction, discharges it, closing the subproof and exporting P → Q into the outer context with the dependency stripped away. No other move retires it, and no amount of further valid derivation inside the subproof substitutes for it.

The failure modes are instructive in both directions. A derivation that assumes P, derives Q, then simply carries on in the outer margin contains no false step. It is nonetheless incomplete as a unit, and a checker rejects it, because what has been established is not the theorem but a claim relative to an open premise. In the other direction, an implication introduction that closes a subproof never opened is not a weaker proof but a malformed one — a closure move with no scope to close; the degenerate case in which an assumption is opened and then never used is by contrast entirely legal, and is exactly what makes P → (Q → P) provable. And the scope discipline is what makes both detectable: the bearer here is the proof's own nesting structure, which holds each assumption open and refuses to let anything derived under it escape until the pairing move is made. ^[9]

Mapped back: every element of the signature is present in mechanical form. The demand is opened by the assumption, its type is fixed to a single admissible discharge, the interval is bounded by the subproof, the bearer is the scope table rather than anyone's memory, and the closure test is precisely the checker's refusal to accept a derivation with undischarged assumptions. The narrative vocabulary of anticipation and payoff has been removed entirely, and the relation survives intact — which is the strongest available evidence that the affective layer was never load-bearing.

Applied/industry

A deprecation notice and its sunset. A platform team publishes a deprecation notice for a versioned API endpoint, with a sunset date eighteen months out. The notice is a setup, and the demand it opens is definite: on that date the endpoint is withdrawn, or the notice is explicitly retracted. It simultaneously opens a mirrored demand on every client team, each of which now owes a migration. The interval is long and the bearer is distributed across changelogs, calendar reminders, and whichever engineer read the notice — notoriously lossy, which is why the pair so often fails.

Both halves fail in characteristic ways. If the date slips twice and the endpoint stays up, the immediate cost is small, but the structural cost compounds: clients learn that a sunset date types nothing, so subsequent notices open no demand at all and the mechanism is spent. If instead the endpoint disappears with no notice, the removal is indistinguishable from an outage, because an unpaired resolution cannot be recognized as one — the same event, correctly executed, reads as a fault. Programs that work externalize the bearer rather than trusting memory: a machine-readable sunset header on every response, a warning logged on each call, a dashboard of remaining callers, and a removal announcement that quotes the original notice so the two elements are visibly linked at the moment of closure. ^[10]

Mapped back: the sunset case shows the bearer question doing the decisive work, where the proof case had it handled invisibly by scope. Typing is present (removal or retraction, nothing else counts), liveness is present and expensive, and linkage has to be manufactured deliberately, since clients will not connect a removal to a notice they no longer remember. The compounding failure when setups routinely go undischarged is the same debt behaviour a leaking allocator exhibits, expressed in an organization rather than in a heap.

Structural Tensions

T1 — Too faint to register versus loud enough to telegraph. A setup must be salient enough that the bearer retains it and faint enough that it does not deliver its own resolution in advance. Miss low and the discharge lands as coincidence, since nothing was owed in anyone's mind. Miss high and the pair completes in the audience before the resolution arrives, leaving it redundant. The window is substrate-dependent: a compiler reads a declaration once, perfectly, and needs no salience, while a reader must notice a plant and then set it down. The usual lever is giving the setup other work to do where it appears.

T2 — Open demands as debt that compounds. Every live setup consumes carrying capacity in whichever bearer holds it, and unclosed pairs accumulate rather than expire quietly: dangling allocations, unanswered tickets, dropped subplots, obligations recorded and never settled. Yet some systems run permanently in the red, since serials, standing options, and open research agendas treat unresolved demands as the product rather than as failure. The hard judgement is not counting open items but separating an obligation deliberately deferred from one quietly abandoned, since both appear in the register as merely still open.

T3 — Resolutions arriving without setups. A discharge with no locatable initiator reads as arbitrary, which is what a deus ex machina, a free of never-allocated memory, and a payment against no invoice have in common. But retroactive pairing is genuinely available: audiences and auditors reconstruct a setup from whatever prior material exists, so the same event can often be rescued after the fact. That creates a standing temptation, because the honest test — would this setup have typed this resolution in advance — is nearly impossible to run once you already know the answer.

T4 — Who is obliged to remember. A pair exists only while some bearer holds the setup across the interval, and bearers differ enormously in fidelity and cost. Machines hold demands perfectly and almost free; humans hold them lossily and at expense; institutions hold them in registries that decay, get migrated, and lose rows. Porting a pattern between substrates silently changes the interval a pair can span, which is why a setup working across two pages fails across two seasons, and a lock discipline sound within one process becomes unsound across a partition.

T5 — Undertyped versus overdetermined setups. The setup must constrain adequate resolutions tightly enough that discharge is checkable, yet a setup admitting exactly one continuation has effectively already delivered it. Undertype it and almost anything counts, so nothing can be audited and nothing is earned; overtype it and the resolution becomes a formality adding no information when it arrives. This is one knob with two failure edges, the same knob whether the setup is a mantelpiece detail, an interface signature, a preregistered hypothesis, or a loan covenant.

T6 — Genuine discharge versus relocation of the demand. Many apparent resolutions do not close a demand at all; they move it. A ticket is closed by reassignment, a debt is refinanced, a question is answered with a further question, a subplot is retired by promising a sequel. Locally each pair looks complete and the register entry clears, so auditing at pair granularity undercounts what remains outstanding. Only a global count of live demands exposes the transfer, and the gap between the two counts measures how much a system is deceiving itself.

Structural–Framed Character

Setup Resolution Pair sits at the structural end of the structural–framed spectrum, with an aggregate of 0.0 and all five diagnostics reading zero. What travels is a two-role ordered relation: an initiating element opens a typed, bounded demand; a causally linked answering element discharges it; and the pair, not either element alone, is the operative higher-order unit. The setup fixes what counts as an adequate resolution, and a closure test separates completion from mere succession.

The diagnostic doing most of the work is vocabulary travel, and it reads 0.0. The everyday associations of the phrase are narrative, but the entry treats setup and payoff as one domain realization among jokes, request–response protocols, contracts, proofs, musical cadences, and staged mechanisms. None of the load-bearing roles is stated in any of those lexicons: open demand, linked discharge, closure.

Evaluative weight reads zero — the closure test is a test, not a verdict on quality. Institutional origin reads zero: contracts are one substrate among cadences and staged mechanisms, and no court or convention is needed for a setup to project its answer class. Human-practice-bound reads zero: the demand opened may be an instability or a dependency rather than an expectation or obligation. Import-vs-recognize is recognition — one finds the two ordered roles and runs the closure test in place.

The grade means the prime carries into a new domain as a pattern check rather than an interpretation. The characteristic risk is emptiness, not misfit: transfer is warranted only where the open-demand–answer–closure roles actually survive.

Substrate Independence

Setup–Resolution Pair is about as substrate-independent as a prime can be — composite 5 / 5 on the substrate-independence scale. The travelling structure is four steps that name no medium: a typed demand is opened, some bearer holds it live across an interval, a linked element discharges it, and completion is scored on the pair rather than on either half. Expectation and payoff in a story, the two halves of a joke, request and response in a protocol, obligation and performance under a contract, an assumption discharged in a proof, harmonic tension and cadence all realise it literally. Its residue on non-closure is equally portable — a leak, a dangling reference, an unsettled liability — which is why ledgers and collectors evidence it as plainly as fiction does.

• Composite substrate independence — 5 / 5
• Domain breadth — 5 / 5
• Structural abstraction — 5 / 5
• Transfer evidence — 4 / 5

Relationships to Other Abstractions

Current abstraction Setup–Resolution Pair Prime

Foundational — no parent edges in the catalog.

Children (3) — more specific cases that build on this

• Joke Domain-specific is a kind of Setup–Resolution Pair

  The proposed strict upward parent is prime:setup_resolution_pair.

• Adjacency Pair Prime is a kind of Setup–Resolution Pair

  Adjacency Pair is Setup-Resolution Pair specialized by noticeable absence, preference asymmetry, and an explicit repair affordance.

• Period Structure Domain-specific is a decomposition of Setup–Resolution Pair

  Period Structure is the tonal-cadential form of a weak opening demand discharged by a paired stronger answer.

Neighborhood in Abstraction Space

Setup–Resolution Pair sits in a sparse region of abstraction space (64^th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely rather than landing on a neighbor.

Family — Formal Systems & Structural Conditions (20 primes)

Nearest neighbors

• Connectedness — 0.71
• Adjacency Pair — 0.71
• Compellence — 0.71
• Bijectivity — 0.70
• Compatibility — 0.70

Computed from structural-signature embeddings · 2026-09-10

Not to Be Confused With

Setup–Resolution Pair must first be distinguished from Tension And Release, its nearest and most seductive neighbour, since both describe a deliberately created earlier condition that a later element settles. The difference is what the earlier element does. Tension and release requires instability: a departure from a baseline, sustained long enough for anticipation to accrue, whose release derives its value from the build. Its central claim is causal and affective — deliver the payoff cheaply and it falls flat. A setup–resolution pair requires no instability, no anticipation, and no felt value whatsoever. A variable declaration creates nothing uncomfortable and its later use satisfies no one, yet the pair is complete or incomplete on entirely objective grounds. Every tension-and-release cycle contains a pair, but most pairs carry no tension, which is why this prime survives into substrates with no experiencing party while tension and release does not.

Nor is it Adjacency Pair, though the two-slot shape is shared and conversation analysis supplies much of the intuition. An adjacency pair loads three further commitments that this prime does not carry: noticeable absence, where a missing second part is itself a social event; preference asymmetry, where two legitimate second parts differ in the interactional work they cost; and repair affordance, where a mismatch invites a request to try again. Those commitments presuppose two parties oriented to each other in real time. This prime needs no second party — a program opens and discharges demands against itself — and its second element may sit decades from the first, where noticeability and repair have no purchase. Adjacency pair is the interactional specialization; treating every pair as one imports an obligation to notice that most substrates cannot support.

Expectation Violation is the mirror image rather than a variant. It requires an active prediction and an incompatible realization that crosses a mismatch threshold, producing surprise and reorientation; the violated expectation becomes diagnostically visible precisely because it was not met. A setup–resolution pair is defined by the demand being met. The two interlock: a setup that types its successors is what makes a violation detectable, and a deceptive cadence is one prime's failure used as the other's mechanism. But the analyses run in opposite directions — one asks whether the demand closed, the other what the failure to close reveals — and conflating them reads every unresolved setup as deliberate subversion.

The distinction from Future Or Promise is one of category. A future is a first-class object, a placeholder standing in for a value committed to arrive later, which can be passed around, composed, and inspected for its state. A setup–resolution pair is a relation between two elements, and it is not itself a thing any system holds. A future is one particularly explicit implementation of the pattern: it reifies the open demand so the bearer's job becomes trivial, which is exactly why futures are useful. Most pairs have no such reification, and the analytic work — locating the demand, naming its type, finding its bearer — exists because nothing has been reified.

Related but distinct again is Callback, which specifies a handler registered in advance and invoked later by a foreign party when a condition arises. A callback determines its resolution at setup time and hands over control of when the discharge fires. A setup–resolution pair usually leaves the resolution undetermined at the moment of setup, constraining only its type, and frequently the same agent supplies both halves. Registration is one strategy for guaranteeing a pair closes; it is not what makes something a pair.

Finally, Dependency overlaps heavily and differs in temporal structure. A dependency is a directed relation in which one element relies on another being present, prior, compatible, or supplied, with a specifiable failure mode when it is not. Dependencies are typically standing conditions with no ordering of emplacement: a function depends on a library that existed for years beforehand, and nothing was opened by the writing. This prime requires that the first element actively open something that was not open before, and the second close it. A satisfied dependency is a background fact; a discharged setup is an event. Where the two coincide — a dependency created at runtime and later satisfied — the pair framing adds the liveness interval and the bearer question, which pure dependency analysis has no place for.

Solution Archetypes

No catalogued solution archetypes reference this prime yet.

References

[1] Schegloff, Emanuel A., and Harvey Sacks. "Opening Up Closings". Semiotica, 1973. Gives the classic formulation of the adjacency pair - two-utterance length, relative ordering of parts, and a first part whose pair type discriminates among admissible second parts - so that two elements form a unit only when the first types the second rather than merely preceding it. registry ↩a ↩b

[2] Schegloff, Emanuel A. "Sequencing in Conversational Openings". American Anthropologist, 1968. Introduces conditional relevance: given a first item, a second of a specified class is expectable, and until it arrives it is officially absent rather than merely missing - the standing question of whether the demand is still outstanding. registry ↩

[3] Strom, Robert E., and Shaula Yemini. "Typestate: A Programming Language Concept for Enhancing Software Reliability". IEEE Transactions on Software Engineering, 1986. Supports machine-checkable tracking of each object's currently open obligations, detecting at compile time execution sequences that are syntactically legal but undefined, such as reading an uninitialized variable or dereferencing a deallocated pointer. registry ↩

[4] Postel, Jon. Transmission Control Protocol. RFC 793, USC Information Sciences Institute for DARPA, 1981. Supports the SYN/ACK exchange and retransmission timeouts as the finite-time decision procedure for a protocol demand whose discharge has not arrived. registry ↩

[5] Elstein, Arthur S., Lee S. Shulman, and Sarah A. Sprafka. Medical Problem Solving: An Analysis of Clinical Reasoning. Harvard University Press, 1978. Establishes the hypothetico-deductive account in which a small set of diagnostic hypotheses is raised early and later data confirm or exclude each one, so a raised differential is a demand a subsequent test closes. registry ↩

[6] Barthes, Roland. "Introduction a l'analyse structurale des recits". Communications, 1966. Distinguishes cardinal functions, which are consequential and open an alternative to be settled later, from catalyses, which are merely consecutive - the difference between material that types a later event and material that only permits it. registry ↩

[7] Mohan, C., Don Haderle, Bruce Lindsay, Hamid Pirahesh, and Peter Schwarz. "ARIES: A Transaction Recovery Method Supporting Fine-Granularity Locking and Partial Rollbacks Using Write-Ahead Logging". ACM Transactions on Database Systems, 1992. Supports closure licensing forgetting: a transaction table tracks only currently active transactions and checkpointing lets everything before the oldest live one be discarded, so retained state scales with concurrent openness rather than elapsed history. registry ↩

[8] MITRE. CWE-772: Missing Release of Resource after Effective Lifetime. Common Weakness Enumeration. Supports only the software instance of the failure residue — "The product does not release a resource after its effective lifetime has ended" — that is, that a demand still open past its liveness interval is a catalogued defect class in code. It says nothing about contracts, narrative plants, or any transfer of competence between those domains. registry ↩

[9] Gentzen, Gerhard. "Untersuchungen ueber das logische Schliessen. I". Mathematische Zeitschrift, 1935. Establishes natural deduction with assumption discharge and scope discipline, so that a derivation carrying an undischarged assumption is incomplete as a unit and only the implication-introduction move retires it. registry ↩

[10] Wilde, Erik. The Sunset HTTP Header Field. RFC 8594, RFC Editor, 2019. Supports the machine-readable sunset announcement that externalizes the deprecation demand onto the responses themselves rather than onto anyone's memory. registry ↩