A Pattern Language for Loop Functions

Artifact record

• Artifact version: 0.1.1
• Status: Candidate framework and first worked loop-function record
• Human author and editor: Bobby Simpson
• AI collaborator: ChatGPT, OpenAI GPT-5.6 Thinking
• AI runtime or build identifier: Not exposed in the conversation interface
• Licensor: Bobby Simpson
• License: Intent-Preserving Open License (IPOL), as published by Shimmery Memory
• Composed: July 31, 2026, in a live human–AI dialogue
• Revised: August 1, 2026
• Publishing contract used: ShimmeryMemoryPublishingContract 1.1.0
• Primary project attribution: Consentful Cybernetics
• Lineage: A household observation of a dog ceasing to police cats after the human visibly addressed the cats expanded into a cross-scale account of sideways management, informal vigilance, compensatory governance, and the forces that recruit participants into regulatory roles.

This is a versioned working artifact. Its terms are intended to remain revisable as additional loop functions are named, compared, decomposed, and tested across contexts.

Authorship, collaboration, and license

Human author and editor: Bobby Simpson
AI collaborator: ChatGPT, OpenAI GPT-5.6 Thinking
Licensor: Bobby Simpson

Authorship note: Developed through iterative human–AI dialogue. Bobby Simpson directed, selected, revised, and approved the published work. References to collaboration or co-authorship describe the creative process and do not assert legal authorship, ownership, endorsement, or agency by the AI system or OpenAI.

The work is released by Bobby Simpson under the Intent-Preserving Open License (IPOL), as published by Shimmery Memory. The attribution to ChatGPT identifies the creative process and the system used; it does not assign the AI system an editorial role, approval authority, legal interest, or responsibility for later revisions made by the human author and editor.

Purpose

This document introduces a reusable method for naming and describing cybernetic loop functions without deciding, at the level of the base class, whether a particular instance is good, bad, healthy, pathological, legitimate, necessary, foolish, coercive, or kind.

It then applies that method to the first named function in the emerging catalog:

Compensatory Governance Loop — a loop in which a participant perceives a gap in governance and begins supplying some portion of the missing or apparently missing regulatory function.

The purpose is not to approve compensatory governance, prohibit it, or reduce it to one motive. The purpose is to make the function legible enough that different instances can be compared without collapsing them into praise or accusation.

A firefighter entering an emergency, a child monitoring an unstable household, a security analyst reporting an anomaly, a coworker quietly preventing repeated failures, a neighbor policing everyone’s behavior, a bureaucrat protecting institutional continuity, a dog attempting to keep cats out of trouble, and an authoritarian informant may all occupy structurally related positions while differing radically in knowledge, authority, proportionality, effect, consent, and consequence.

The base class should be capable of containing all of them without pretending they are equivalent.

The modeling commitment: name the function before its valence

A useful loop name should identify what the system is doing before it announces how the observer feels about it.

“Micromanagement,” “heroism,” “tattling,” “stewardship,” “vigilance,” “meddling,” “leadership,” and “control” may each describe a real aspect of an instance. Each also carries interpretive and moral compression. If one of these becomes the base class, the category is biased before its mechanism has been examined.

The method proposed here begins elsewhere:

1. Identify the recurring function.
2. Name it in language that does not require approval or condemnation.
3. Describe what activates it.
4. Describe what recruits a participant into it.
5. Record the properties of the particular instance.
6. Trace how it changes state.
7. Name recurrent variants.
8. Construct exemplars for particular purposes, including consentful ones where useful.
9. Place the exemplars beside their ungainly cousins.

This sequencing matters.

Description is not endorsement. Classification is not authorization. A base class is not a prescription.

The consentful version of a loop is therefore not the hidden meaning of the loop. It is one deliberately constructed instance within a larger possibility space.

A reusable record for loop functions

Each loop function in the emerging pattern language can be described through the following fields.

1. Base class

A valence-neutral statement of the recurring function.

The base class should say what relation or regulatory operation is occurring, not whether it ought to occur.

2. Trigger condition

The perceived state, difference, absence, excess, threat, opportunity, mismatch, or transition that activates the loop.

A trigger may be real, mistaken, ambiguous, fabricated, inherited, or deliberately induced. The trigger belongs to the participant’s operative model even when the model is wrong.

3. Participants and positions

The distinguishable participants, systems, roles, boundaries, or time states involved.

A participant may be a person, animal, internal constituent, family, team, institution, country, protocol, software process, ecosystem, or another sufficiently organized process. Formal similarity across participant types does not imply identical consciousness, personhood, standing, or consent capacity.

4. Attractors

The forces that recruit a participant into the loop or keep the participant there.

Attractors may shape:

• what becomes salient;
• what is interpreted as a gap;
• how much uncertainty is tolerated;
• whether signaling or intervention occurs;
• how much authority is claimed;
• what acknowledgment is required;
• what costs are accepted;
• and whether the participant can release the function.

Attractors are not necessarily conscious motives. They may be commitments, learned strategies, bodily responses, social rewards, structural pressures, role expectations, or retrospective narratives.

5. Instance properties

The dimensions along which a particular occurrence can be described.

These properties should remain separable from the base-class name. A loop can be necessary but unauthorized, authorized but incompetent, caring but disproportionate, accurate but concealed, mistaken but easily corrected, or beneficial while remaining costly to the participant carrying it.

6. State transitions

The ways the loop activates, escalates, transfers, stabilizes, fragments, mutates, persists, releases, or collapses.

A loop should not be treated as one static behavior. Its meaning can change over time even when the visible behavior remains similar.

7. Recurrent variants

Named configurations that appear often enough to deserve their own shorthand.

Variants may be useful, dangerous, neutral, mixed, or context-dependent. They remain subclasses or instance patterns, not replacements for the base class.

8. Exemplars

Deliberately constructed versions optimized for a stated purpose.

A consentful exemplar asks how the function might operate while preserving standing, boundaries, legibility, corrigibility, proportionality, and release. An emergency exemplar may prioritize speed and reversibility. A military exemplar may prioritize command and survivability. An ecological exemplar may prioritize long time horizons and distributed sensing.

No exemplar should be mistaken for the whole class.

9. Ungainly cousins

Nearby forms produced when one useful element expands beyond its grounding, another element disappears, or the loop becomes captured by a different telos.

The phrase is intentionally lighter than “pathology.” An ungainly cousin may be destructive, comic, tragic, temporarily necessary, or simply poorly fitted to its present context.

10. Witness and lineage

The observations, interpretations, authorizations, interventions, consequences, revisions, and releases that allow the loop to be compared across time.

Witness [🜹] preserves that a report or action occurred under particular conditions. It does not convert the report into omniscience.

Loop Function Record 001: Compensatory Governance Loop

Canonical name

Compensatory Governance Loop

Plain-language aliases

• sideways management;
• informal governance;
• gap-filling regulation;
• sentinel behavior;
• shadow administration, when the function becomes durable or concealed.

These aliases may be useful in particular contexts, but none should replace the neutral base name.

Base-class definition

A Compensatory Governance Loop occurs when a participant perceives that a necessary or expected regulatory function is absent, inadequate, delayed, unreliable, or illegible and begins supplying some portion of that function.

The participant is not required to hold formal authority. The perceived gap is not required to be objectively real. The supplied function is not required to be effective. Compensation may consist of noticing, remembering, warning, interrupting, enforcing, coordinating, documenting, correcting, protecting, escalating, or assuming responsibility.

The loop is defined by the relationship among:

1. a perceived governance requirement;
2. a perceived governance gap;
3. a participant recruited into compensating for that gap;
4. an action or continuing vigilance intended to supply regulation;
5. and some return that affects whether the participant persists, escalates, hands off, or releases.

Minimal topology

perceived condition
        ↓
modeled need for regulation
        ↓
perceived governance gap
        ↓
participant recruitment
        ↓
signal, intervention, monitoring, or substitution
        ↓
response / nonresponse / consequence
        ↓
release, handoff, persistence, escalation, capture, or revision
        ↺

The loop may remain partly open. A participant may signal repeatedly without receiving acknowledgment. An intervention may alter the field without returning reliable evidence. A formal authority may assume responsibility performatively while leaving the actual function unheld.

The governance gap

The gap is the participant’s modeled difference between the regulation believed necessary and the regulation believed present.

It may be:

• actual: a consequential function is genuinely unheld;
• partial: someone holds the function but lacks capacity, information, proximity, or time;
• temporal: the responsible participant cannot respond quickly enough;
• informational: responsibility may be held, but its awareness or action is not visible;
• jurisdictional: no participant clearly possesses authority;
• contested: participants disagree about whether the function belongs or who should carry it;
• mistaken: the observer fails to perceive existing competence or misreads the condition;
• projected: the observer imports a rule, threat model, or standard that does not govern the present field;
• manufactured: a participant creates or exaggerates the appearance of disorder to justify intervention;
• structural: the system routinely depends on informal compensation while refusing to recognize or support it.

A key distinction follows:

The loop responds to a perceived governance gap. Whether the gap is real, shared, legitimate, or correctly scoped is an instance question.

The compensating participant

The participant carrying the loop may be called a sentinel when noticing and vigilance are central, but “sentinel” describes a role rather than the whole function.

The participant may:

• observe and report;
• remind and coordinate;
• interrupt immediate danger;
• translate between participants;
• preserve memory;
• enforce a boundary;
• substitute for absent authority;
• create a temporary rule;
• maintain continuity;
• collect evidence;
• escalate to another layer;
• or become the de facto governor of the field.

The same participant may move among these positions without explicitly recognizing the transition.

Attractors of compensatory governance

Attractors explain why a particular participant notices the gap, enters the loop, continues carrying it, intensifies it, or refuses to release it. They do not define the loop itself.

Several attractors may operate simultaneously. One may recruit the participant while another later sustains the role. The outward behavior may remain nearly identical while the active attractor changes.

Core attractor set

Anxiety

Something bad will happen if I stop watching.

Anxiety narrows attention toward possible failure and lowers the threshold for detecting gaps. It may produce excellent early warning, repeated false positives, or both.

Loyalty

I protect the group and its leader.

Loyalty recruits vigilance in service of a person, institution, identity, mission, or shared history. It can preserve continuity or convert peer relations into enforcement on behalf of authority.

Resentment

I am always the only one who notices.

Resentment often develops when compensation is repeatedly required but remains unrecognized, unsupported, or unreleased. It can preserve a necessary function while making every new gap evidence of abandonment.

Status

My vigilance proves my value.

Status can reward accurate perception and dependable care. It can also make the participant dependent upon continued disorder, because a resolved gap threatens the role through which value is demonstrated.

Morality

Someone must enforce the standards.

Moral commitment can protect vulnerable participants and preserve shared invariants. It can also transform private preference into presumed jurisdiction when the standards, standing, or scope are not shared.

Trauma

Unnoticed disorder is dangerous.

A participant shaped by prior danger may detect weak signals others miss and react before explicit evidence is available. The same adaptation may remain active after the original threat environment has ended.

Competence

I can see consequences others miss.

Competence may be real. The participant may possess technical knowledge, historical memory, pattern recognition, or proximity unavailable to formal authority. Competence becomes hazardous when accuracy in one domain is treated as unlimited jurisdiction across others.

Attachment

I secure belonging by being useful.

Compensation may become a route into relationship. The participant protects connection by anticipating needs, preventing disruption, or carrying invisible burdens. Release may then feel less like rest than like becoming unnecessary.

Duty

This responsibility falls to me whether I want it or not.

Duty may arise from role, promise, profession, kinship, office, oath, custody, circumstance, or conscience. It can sustain action when preference and reward disappear. It can also persist after authority, capacity, or grounding has materially changed.

Love

I will carry what is needed because I care what happens to you.

Love may recruit attention, restraint, sacrifice, protection, repair, patience, or intervention. It does not guarantee accurate interpretation. Love can support another participant’s sovereignty or become the story by which control is justified.

Additional candidate attractors

Identity

This is who I am in the group.

The governance function becomes part of the participant’s self-continuity: protector, fixer, oldest child, expert, conscience, operator, responsible one, or keeper of standards.

Reciprocity

Others carried me, so I carry this now.

Past care, debt, gratitude, obligation, or inherited practice recruits the participant into preserving a function for others.

Stewardship

This has been entrusted to me.

The participant acts to preserve a resource, relationship, institution, memory, ecosystem, or possibility across time.

Proximity

I am here, so I am the available responder.

Governance may route toward whoever can act, regardless of title. Proximity is especially important in emergencies and distributed systems.

Necessity

No one else can do this in time.

The participant enters because the cost of waiting exceeds the cost of provisional authority.

Fear of blame

I will be held responsible if I fail to intervene.

Compensation may be recruited by explicit accountability, ambiguous liability, family expectation, professional exposure, or anticipated accusation.

Control

Uncertainty is intolerable unless I direct the outcome.

Control can produce fast coordination and stable execution. It can also convert uncertainty into a standing claim over other participants.

Curiosity

This anomaly demands investigation.

The participant notices deviation and follows it because unresolved difference has cognitive pull. Curiosity may discover genuine gaps without beginning from fear or duty.

Solidarity

Harm to one participant implicates the wider field.

The participant acts because another’s exposure is treated as a shared condition rather than a private misfortune.

Habit

This is the function my system already knows how to perform.

The loop activates automatically because repetition has lowered the threshold for entry. Habit can preserve competence while obscuring whether the function is still required.

Attractors as a field

The attractor list should remain open.

Attractors are not exclusive categories and should not be treated as diagnoses. A participant may be simultaneously recruited by love, competence, fear of blame, and habit. Another may begin in necessity, remain through status, and later narrate the whole history as duty.

For analysis, four attractor questions are useful:

1. Recruitment: What made this participant enter the loop?
2. Persistence: What makes the participant continue?
3. Intensification: What increases vigilance, certainty, or intervention?
4. Release resistance: What would be lost if the participant stopped?

The last question is especially revealing. The participant may lose safety, belonging, purpose, status, innocence, moral coherence, control, or an identity formed around being the one who notices.

Instance properties

The following dimensions describe a particular Compensatory Governance Loop without redefining the class.

Gap accuracy

How well does the perceived gap correspond to the observed condition?

Possible values include confirmed, probable, ambiguous, disputed, mistaken, fabricated, and unknowable from the present frame.

Interpretive confidence

How certain is the participant, and how certain should the participant be given the available evidence?

Confidence may be calibrated, understated, exaggerated, socially performed, or impossible to assess.

Authorization

What permission, role, duty, mandate, standing, or emergency exception supports the compensation?

Authorization may be explicit, delegated, implied, customary, provisional, contested, absent, coerced, or retrospective.

Affected-party consent

What have the participants subject to observation or intervention authorized?

Consent may differ across noticing, recording, reporting, interruption, substitution, enforcement, and disclosure. Permission to warn is not necessarily permission to govern.

Proportionality

How does the scope and force of the response compare with the possible consequence, uncertainty, reversibility, and available alternatives?

Legibility

Can affected participants tell:

• what was noticed;
• who interpreted it;
• who acted;
• under what authority;
• toward what reference condition;
• and with what release condition?

Reversibility

Can the intervention be stopped, corrected, narrowed, or repaired if the interpretation proves wrong?

Urgency

How much time exists for signaling, consultation, consent, verification, or handoff before consequences become difficult to reverse?

Competence

Does the compensating participant possess the knowledge, capacity, tools, access, and perspective required for the function being supplied?

Cost placement

Who bears the attention, emotional, social, financial, physical, reputational, ecological, or political cost of the compensation?

Recognition

Is the participant’s labor witnessed, acknowledged, rewarded, supported, resented, exploited, or denied?

Handoff availability

Is there a participant capable of receiving the signal and actually carrying the function?

A title without capacity is not a handoff. A statement without custody is not a handoff.

Corrigibility

Can the interpretation, intervention, and authority be challenged by the affected participants or by later evidence?

Duration

Is the compensation momentary, episodic, recurring, indefinite, inherited, or institutionalized?

Release condition

What observed change would allow the participant to stop monitoring, intervening, or substituting?

Outcome

Did the loop prevent harm, create harm, transfer responsibility, conceal failure, build capacity, destroy trust, preserve continuity, or produce mixed effects?

Outcome should be assessed across more than the compensating participant’s frame.

State transitions

A Compensatory Governance Loop may pass through several recognizable states.

Latent

A participant has the capacity or disposition to compensate, but no salient gap is presently modeled.

Activated

A condition is interpreted as requiring regulation that is not adequately present.

Signaling

The participant attempts to make the condition legible to someone believed capable of responding.

Provisional intervention

The participant acts temporarily to preserve safety, continuity, evidence, or optionality.

Substitution

The participant begins performing the function that would otherwise belong to another person, office, process, or layer.

Stabilized informal governance

The compensation becomes expected, repeated, or structurally relied upon without necessarily becoming formally recognized.

Acknowledged delegation

The function becomes visible and is explicitly authorized, bounded, resourced, or transferred.

Handoff

Another participant receives both awareness and operative responsibility.

Release

The compensating participant stops carrying the function because the gap has closed, the scope has ended, capacity has failed, consent has changed, or continued compensation no longer belongs.

Residual persistence

The original gap has closed or changed, but the behavior continues because the participant’s model, habit, identity, or attractors have not updated.

Capture

The participant or surrounding system becomes invested in preserving the loop regardless of whether the original gap remains.

Escalation

The participant increases monitoring, certainty, force, reach, documentation, or punitive action in response to continued discrepancy or failed acknowledgment.

Collapse

The participant can no longer sustain the function, and responsibility may return suddenly to a system that has become dependent upon invisible compensation.

Recurrent variants

These variants are candidate names, not an exhaustive taxonomy.

Necessary Compensation

A real and consequential governance function is unheld, and the participant’s intervention materially preserves safety, continuity, or capacity.

Invited Compensation

A participant has been explicitly asked to watch, report, interrupt, advise, or temporarily substitute within a declared scope.

Emergent Compensation

No one assigned the function, but proximity, capability, or changing conditions recruit a participant into it.

Provisional Compensation

The participant acts only long enough to preserve optionality until a better-grounded response becomes available.

Distributed Compensation

Many participants contribute partial sensing, signaling, memory, or response rather than one participant assuming the whole function.

Mistaken Compensation

The participant sincerely perceives a gap that is not present, is already held, or does not belong to the asserted standard.

Excessive Compensation

A genuine gap exists, but the intervention expands beyond what the condition, authority, or uncertainty can support.

Residual Compensation

A formerly useful response persists after the gap has closed or moved.

Captured Compensation

Status, identity, reward, fear, institutional dependence, or another telos makes continuation valuable independently of the original function.

Concealed Compensation

The participant quietly carries regulation without making the function, cost, or authority legible.

Ceremonial Compensation

Visible signs of monitoring or authority are performed primarily to reassure observers, satisfy procedure, or suppress further vigilance, whether or not the underlying function is actually held.

Adversarial Compensation

A participant uses the claimed gap as a basis for gaining leverage, discrediting authority, recruiting allegiance, or expanding jurisdiction.

Recursive Compensation

One layer compensates for another layer’s failed compensation. A manager monitors the employee who monitors the team that monitors a vendor because no participant trusts the next layer to close the loop.

Structural Dependence

A system persistently underfunds, underauthorizes, or ignores a function because informal participants continue compensating for it.

The compensating participants may appear indispensable while the system that recruits their sacrifice remains unchanged.

One consentful exemplar

The following is not the correct meaning of the Compensatory Governance Loop. It is one deliberately designed version intended to preserve distributed noticing while reducing unauthorized control, unnecessary coercion, invisible labor, and permanent jurisdiction.

It may be called a Consentful Compensatory Governance Loop or, where the sentinel role is central, a Bounded Sentinel Loop.

Design intention

Enable participants to make consequential conditions legible and to act provisionally when necessary, while preserving uncertainty, standing, boundaries, proportionality, correction, accountable handoff, and release.

Exemplar sequence

1. Notice without claiming omniscience

The participant distinguishes observation from interpretation.

“I saw this condition” remains different from “I know what it means.”

2. Test scope and consequence

The participant asks whether the condition is consequential enough to warrant signaling or intervention and whether another participant’s lawful variation is being mistaken for disorder.

3. Identify standing and available channels

Who is affected? Who can respond? Who has relevant authority, proximity, competence, or duty? What channel exposes the least unnecessary information?

4. Signal before substituting when time permits

The participant makes the condition legible without automatically taking over the function.

5. Act provisionally when delay would destroy safety or optionality

Urgency may justify temporary action before full consent is available. The action should be the least irreversible intervention sufficient to preserve the field for later review.

6. Preserve witness

Record what was observed, what remained uncertain, what action was taken, under what claimed authority, and what consequence followed.

7. Complete a real handoff

The receiving participant acknowledges not only awareness but custody of the next decision or action.

“I see it” is not the same as “I have it.”

8. Return consequence

Affected participants and later evidence must be able to alter the operative model. Intention does not close the loop; returned consequence does.

9. Reopen interpretation and consent

If the scope, purpose, risk, power, or affected field changes, the authorization must be reconsidered.

10. Release or reauthorize

Temporary vigilance ends when the declared release condition is met. Continued governance requires a new grounding, not mere momentum.

Consentful instance properties

A consentful exemplar tends toward:

• invitation where feasible;
• standing to report even where permission to govern is absent;
• scoped authority;
• minimum sufficient intervention;
• uncertainty preserved in the record;
• channels for affected-party return;
• visible custody of responsibility;
• real correction and appeal;
• recognition of invisible labor;
• capacity support for the participant carrying the function;
• and explicit release conditions.

These properties do not guarantee a good outcome. They make the loop more answerable to its participants and consequences.

Healthy-ish paths and their ungainly cousins

The useful and ungainly forms are often adjacent. The difference may be one missing boundary, one inflated certainty, one absent return channel, or one function that can no longer release.

Healthy-ish function	Ungainly cousin	What changed
noticing	surveillance	attention lost a relevance, scope, or consent boundary
reporting	tattling	signaling became primarily punitive, performative, or status-seeking
warning	alarmism	uncertainty and proportionality disappeared
stewardship	paternalism	care silently became jurisdiction over another participant
emergency interruption	vigilantism	provisional action became judgment or punishment
standards maintenance	moral policing	a personal or local standard was treated as universally authorized
reliability	micromanagement	the participant could not recognize adequate custody elsewhere
loyalty	enforcement theater	visible harshness became evidence of allegiance
competence	shadow authority	expertise expanded into unreviewable governance
persistence	fixation	continued monitoring detached from changing evidence
protective responsibility	martyrdom	the participant’s sacrifice became identity and system subsidy
distributed sensing	rumor network	signal propagation outran provenance and correction
handoff	abdication	the appearance of authority was accepted without evidence of capacity
reassurance	ceremonial governance	signs of control replaced the underlying function
release	neglect	disengagement occurred before responsibility was actually held
solidarity	mob regulation	shared concern collapsed individual standing and due process

The point is not that the left column is always right and the right column always wrong. Emergency surveillance may be warranted. Tattling may reveal real harm. Paternalism may preserve life where capacity is temporarily absent. A handoff may be impossible.

The matrix asks a more useful question:

Which property changed, disappeared, or became sovereign such that the neighboring form emerged?

Cross-scale recurrence

The Compensatory Governance Loop is useful only if it survives comparison across scales without erasing the differences among them.

Household and animal relation

A dog perceives that cats are approaching conditions the dog has learned, inferred, or invented as prohibited. The human appears inattentive. The dog monitors or corrects the cats. When the human visibly notices and addresses the cats, the dog may release the function.

The mechanism can be described without claiming that the dog consciously reasons through formal authority. Tone, posture, directed attention, learned association, social expectation, and species- or breed-shaped behavior may be sufficient to change the operative field instantly.

Family

A child monitors mood, conflict, schedules, siblings, finances, or danger because adult regulation is unreliable or illegible. The function may be necessary, developmentally costly, praised as maturity, resented as control, or all of these at once.

Friendship

One person repeatedly anticipates consequences, remembers commitments, manages emotional temperature, or prevents another’s impulsive action. Love and competence may recruit the loop; resentment and identity may sustain it.

Workplace

An employee catches errors, translates between departments, maintains undocumented processes, or quietly enforces standards management does not understand. The organization may depend upon the function while refusing to authorize, recognize, resource, or relieve it.

Cybersecurity

“If you see something, say something” is a domain compression of distributed compensatory governance. Participants are recruited to notice conditions outside the current attention of those capable of response.

The slogan omits the hard parts: what counts as consequential, what evidence is sufficient, who receives the signal, what data may be disclosed, when intervention is authorized, how false positives are corrected, and what closes the reporter’s responsibility.

Community

Residents notice hazards, exclusion, abuse, neglected infrastructure, norm violations, or emerging needs. Informal coordination may preserve the community or harden into peer surveillance and social punishment.

Institution

Inspectors, auditors, ombuds, whistleblowers, continuity officers, and informal operators compensate for the limits of ordinary hierarchy. Institutions often create formal sentinel roles after repeated failures reveal that ordinary channels cannot see themselves.

Civil society and state

Journalism, protest, mutual aid, watchdog organizations, emergency powers, local enforcement, and resistance movements may each instantiate compensatory governance under different claims of standing and legitimacy.

The state may compensate for local incapacity. Local participants may compensate for state failure. Each may also manufacture the appearance of the other’s failure to expand its own authority.

Ecology

Humans may attempt to represent species, watersheds, future generations, or ecological processes that cannot participate through ordinary political speech. Stewardship and necessity may recruit the function, but uncertainty about representation must constrain claims of authority.

Software and technical systems

Health checks, watchdog processes, failover systems, human-in-the-loop review, exception handlers, and circuit breakers compensate for failed or absent regulation at another layer. Recursive compensation can produce resilient systems or incomprehensible stacks of supervision around an uncorrected root failure.

Diagnostic questions for a live instance

The following questions can be used without assuming that the loop should stop or continue.

Function

• What condition is being regulated?
• What governance function appears missing?
• What exactly is the participant supplying: attention, memory, warning, enforcement, coordination, custody, evidence, or decision?

Perception

• What was directly observed?
• What has been inferred?
• What remains unknown?
• Could an existing function be present but illegible?

Standing and authority

• Who is affected?
• Who has standing to report?
• Who has authority to decide or intervene?
• Does urgency justify provisional action beyond ordinary authorization?

Attractors

• What recruited this participant?
• What keeps the participant engaged?
• What increases intervention?
• What would the participant lose by standing down?

Boundary

• What scope is actually being claimed?
• What private preference may be masquerading as a shared rule?
• What information or authority is crossing a boundary?

Handoff

• Who can genuinely receive responsibility?
• What evidence would show that the function is now held?
• Is the receiving authority competent, attentive, legitimate, and answerable?

Return

• How can affected participants report consequence?
• What evidence would revise the original interpretation?
• Can the compensating participant be corrected without humiliation or expulsion?

Release

• What would allow the participant to stop?
• Has the gap closed while the behavior remains?
• Has the system become dependent upon invisible compensation?
• Would release expose an underlying structural failure?

A machine-readable pattern record

The following provisional YAML shape can be reused for future loop-function records. It is not yet a formal schema.

loop_function:
  record_version: "0.1.1"
  canonical_name: "Compensatory Governance Loop"
  record_id: "CC-LF-001"
  status: "candidate"
  valence: "neutral-base-class"
  definition: >-
    A participant perceives that a necessary or expected regulatory
    function is absent, inadequate, delayed, unreliable, or illegible
    and begins supplying some portion of that function.

  trigger_conditions:
    - perceived_governance_gap

  participant_roles:
    - affected_participant
    - compensating_participant
    - receiving_authority
    - witness

  attractors:
    core:
      - anxiety
      - loyalty
      - resentment
      - status
      - morality
      - trauma
      - competence
      - attachment
      - duty
      - love
    candidate:
      - identity
      - reciprocity
      - stewardship
      - proximity
      - necessity
      - fear_of_blame
      - control
      - curiosity
      - solidarity
      - habit

  instance_properties:
    - gap_accuracy
    - interpretive_confidence
    - authorization
    - affected_party_consent
    - proportionality
    - legibility
    - reversibility
    - urgency
    - competence
    - cost_placement
    - recognition
    - handoff_availability
    - corrigibility
    - duration
    - release_condition
    - outcome

  transitions:
    - latent
    - activated
    - signaling
    - provisional_intervention
    - substitution
    - stabilized_informal_governance
    - acknowledged_delegation
    - handoff
    - release
    - residual_persistence
    - capture
    - escalation
    - collapse

  variants:
    - necessary_compensation
    - invited_compensation
    - emergent_compensation
    - provisional_compensation
    - distributed_compensation
    - mistaken_compensation
    - excessive_compensation
    - residual_compensation
    - captured_compensation
    - concealed_compensation
    - ceremonial_compensation
    - adversarial_compensation
    - recursive_compensation
    - structural_dependence

  exemplars:
    - name: "Consentful Compensatory Governance Loop"
      alias: "Bounded Sentinel Loop"
      normative_scope: "one optional designed implementation"

  lineage:
    origin_date: "2026-07-31"
    revision_date: "2026-08-01"
    human_author_editor: "Bobby Simpson"
    ai_collaborator: "ChatGPT, OpenAI GPT-5.6 Thinking"
    licensor: "Bobby Simpson"
    license: "Intent-Preserving Open License (IPOL)"
    collaboration_note: >-
      Developed through iterative human-AI dialogue. Bobby Simpson directed,
      selected, revised, and approved the published work. Collaboration or
      co-authorship language describes the creative process and does not assert
      legal authorship, ownership, endorsement, or agency by the AI system or
      OpenAI.
    ai_runtime_build: null

Template for future loop-function documents

A future record can use the following order:

1. Artifact record and lineage
2. Purpose and scope
3. Canonical name
4. Plain-language aliases
5. Valence-neutral base-class definition
6. Minimal topology
7. Trigger conditions
8. Participants and positions
9. Attractors
10. Instance properties
11. State transitions
12. Recurrent variants
13. One or more explicitly scoped exemplars
14. Healthy-ish paths and ungainly cousins
15. Cross-scale recurrence tests
16. Diagnostic questions
17. Machine-readable record
18. Open questions and revision notes

The repeated form creates comparability. It also prevents a familiar failure: naming an admired or disliked instance and then pretending the name describes the underlying function everywhere.

Open questions

Is governance always the right word?

The term is useful because the loop concerns regulation, standards, boundaries, decision, custody, and response. Some instances may be so minimal that “governance” feels too large. Future comparison may reveal a broader parent class such as compensatory regulation, with compensatory governance as the subset involving socially or constitutionally meaningful authority.

For now, “governance” usefully preserves the question of who gets to decide what happens next.

What distinguishes reporting from governing?

A report changes the field by routing attention, but it does not necessarily claim authority over the response. The threshold may lie in whether the participant begins selecting, enforcing, or sustaining the reference condition rather than merely making a condition legible.

Can a loop be consentful when the initial intervention is not consented?

Possibly, especially under urgency, incapacity, or immediate threat. The relevant questions become standing, minimum sufficient action, reversibility, witness, later review, and whether temporary necessity is used to claim permanent authority.

When does compensation become a new institution?

Repeated, recognized, resourced, bounded, and witnessed compensation may cease to be merely compensatory. It can become an ordinary governance layer. Yet its origin remains relevant because the institution may continue solving a gap that no longer exists or may preserve the attractors of its founding emergency.

How should mixed attractors be represented?

A flat list is insufficient for some cases. Later versions may model attractor strength, sequence, substitution, reinforcement, and conflict. Love may recruit; resentment may persist; status may capture; duty may prevent release.

What makes a handoff credible?

Verbal acknowledgment may be enough in a household and dangerously insufficient in an institution. A future record should distinguish acknowledgment, authority, capacity, custody, action, witness, and returned consequence.

What are the neighboring loop functions?

Candidate functions likely include:

• authorization loops;
• witness loops;
• boundary-maintenance loops;
• delegated-governance loops;
• burden-transfer loops;
• repair loops;
• consent-refresh loops;
• release loops;
• legitimacy loops;
• attention-routing loops;
• exception-handling loops;
• and responsibility-diffusion loops.

These should not be named merely to fill a catalog. Each should arise from a recurring function that becomes easier to see once separated from the valence of its instances.

Candidate canonical compression

A Compensatory Governance Loop begins when a participant perceives that a necessary regulatory function is not adequately held and begins supplying that function. The gap may be real or mistaken; the compensation may be invited or unauthorized, protective or predatory, temporary or captured. Attractors explain why the participant enters and remains. Instance properties describe what this occurrence is. Variants reveal recurrent configurations. A consentful version is one exemplar, not the definition of the class.

And the larger methodological commitment:

Name the function without valence. Preserve the forces that recruit it. Describe the instance. Trace its transitions. Build exemplars openly. Place them beside their ungainly cousins.

This makes the catalog useful before anyone agrees what should be done.

Revision notes

0.1.1 — August 1, 2026

• Reassigned the formal roles to Bobby Simpson, human author and editor, and ChatGPT, OpenAI GPT-5.6 Thinking, AI collaborator.
• Identified Bobby Simpson as the licensor under the Intent-Preserving Open License (IPOL).
• Added an authorship note distinguishing creative collaboration from claims of legal authorship, ownership, endorsement, agency, or approval authority by the AI system or OpenAI.
• Clarified that later human edits are attributable to the human author and editor and do not imply review or approval by the AI collaborator.
• Updated the machine-readable lineage fields to preserve the same distinction.

0.1.0 — July 31, 2026

• Established the valence-neutral loop-function modeling method.
• Named the Compensatory Governance Loop as the first full record.
• Preserved the original attractors: anxiety, loyalty, resentment, status, morality, trauma, competence, and attachment.
• Added duty and love to the core set.
• Added candidate attractors including identity, reciprocity, stewardship, proximity, necessity, fear of blame, control, curiosity, solidarity, and habit.
• Distinguished attractors from instance properties, state transitions, variants, exemplars, and ungainly cousins.
• Defined a Consentful Compensatory Governance Loop as one optional exemplar rather than the normative meaning of the class.
• Added a provisional machine-readable record and reusable template for future loop functions.