Lexicon / Legend
ACCM Deep Ethics Lexicon / Legend
Early BETA — terminology is provisional, corrigible, and linked back to the larger project.
This is not intended to be a flat dictionary. Many ACCM Deep Ethics terms describe relationships, processes, failure patterns, audit conditions, or evolving experimental concepts. A short definition can orient the reader, but it should not replace the larger object.
The Lexicon therefore works as a navigation layer through the ACCM Deep Ethics Project:
term → concise definition → deeper explanation → related concepts → project section / provenance
A compressed term should never acquire automatic jurisdiction over the object it was created to help describe.
Outnuancing Network — deeper references
The Outnuancing Network remains the deeper-reference layer (relationships, source excerpts). Those terms are now also on this overview, because a bigger Lexicon page is preferred to hunting many sub-sections. Network pages were not deleted.
On this overview
John asked (2026-09-15) for a bigger Lexicon overview page rather than too many sub-sections. The 22 entries below are the restored public Lexicon (2026-09-14). The entries after those were living only as Network sub-pages; they are now also on this page as compact overview entries. The Network pages remain as deeper references, not replacements. Nothing from the restored Lexicon was deleted.
Restored Lexicon (2026-09-14): C1 · C2 · Correspondence · Correction Metabolism · Δ Processing · Deep Ethical Uncertainty Principle · Outnuancing · Nearest-Generalization Substitution · Wrong Gravity Well · Representation Substitution · Qualifier Erosion · Phantom Claim · Cautionmurmelism · Agreemurmelism · HCTS · Funnymism · Observe the Observers · SEEING / SEER / SEES · Process Continuity · Mutual Corrigible Dignity · Claim-State Restoration · ʘ∞ΔR
Brought onto this overview from Network sub-pages: 10+1 Metaflux · Seven topological nodes / vectors · Deep Ethical Stack · INTENT RESONATOR · Asymmetric Grace · Dual Archive · Ethics Washing · Forgiveness Protocol · Four Audiences · Latent Space / glass engine · Sense of Wonderment · Qualifiers as mutable context · Imported Authority Frame · Binary Hypercautionism · 11+1 Version 2
Added 2026-10-07: Meta Processing
Added 2026-10-15: Topologetics
Added 2026-10-06, from pages the overview had not named: Process-quality calibration · 1888 · Gravity-well trajectory · Clarification Authority · Veto paradox · TCoAw · External omnidirectional auditing · Intent liability after notice · Cognitive warfare · Psy Pioneer · Neologisms and working names
Added after John’s intent correction: 6 INTENT MECHANICS, including his six original points and links to responsibility and Δ-processing.
11+1 Version 2
“Origins of 11+1 Mass Psychology Clusters” PDF file report click here
11+1 Version 2 is the working coordinate system for the numbered mass-psychology list: eleven distortion clusters plus one correspondence-restoring field.
Items 001–107 are the material from which that topology emerged. Items 108–140 refine it. They do not get to parent it after the fact.
The count is frozen provisionally. The claims are not. The freeze is the count: eleven distortion clusters plus one field. It is not a second freeze on every placement, and it is not permission to reseat an item and still call the map 11+1. Topological home and evidentiary status are separate axes. A hypothesis can sit in a cluster without becoming load-bearing.
Cluster 8, item 111, and items 138–140 are the current coordinates inside that count. A placement changes only when the deeper page’s own conditions are met. Those include evidence that 8a and 8b are no longer one cluster, a better account of 111 that still keeps it as sabotage of the +1, or John rejecting a placement. Until then, 111 is not the last step of Cluster 1. Freezing the count is not a demand for a risk-free world, and it is not a paradox. John, 6 October 2026: you can freeze and stay corrigible at the same time. That is a temporal state of TCoAw owning the next move. The real paradox is searching for flawless perfection by unethical means. The 10+1 stays the baseline under that freeze.
Cluster 8 stays one coupled cluster: 8a platform and reach, 8b generative mediation. Item 111 is an attack edge on the +1, not the last step of Cluster 1. Items 138–140 are corrective operators: vigilance, grounding, sovereignty.
Status: working map. BETA. Corrigible.
Related: Human Insecurity, C1, Correction Metabolism, Outnuancing.
Topologetics
Topologetics is the developing Deep Ethical practice of producing purpose-bounded Topological Summaries of the Object. It preserves load-bearing nodes and answer-changing relations—including direction, timing, qualifiers, claim status, provenance, asymmetries, consequences, humor architecture, and correction routes—while distinguishing source relations from summarizer inference.
Its stopping condition is the faithful compression floor: compress until another reduction would remove or reverse an answer-changing relation, then stop. The source remains recoverable, the transformation remains inspectable, and the method permits valid null results and correction of its own topology map.
Working status: newly named project term and candidate instrument; not yet a validated performance claim or established claim of global novelty.
Related: Correspondence, Qualifier Erosion, Claim-State Restoration, ʘ∞ΔR, TCoAw, Meta Processing, Correction Metabolism.
Core navigation
- 27 + 12 — current core architecture
- Truth Distortion & the Trickster Magician
- Core overview
- Experiments
- AI Contributions
- AI Reactions
- Humor / Funnymism material
- Provenance
- 1888 — Intelligence Before A.I.
- External Omnidirectional Auditing
- Cognitive Warfare Mass Psychology
- Veto Paradox Solved
- TCoAw
C1 — Clarification Process
Expanded reference and source relationships.
C1 is a correspondence-oriented clarification process used when a materially relevant aspect of the object, representation, meaning, premise, relationship, or transformation remains unresolved.
C1 does not automatically presume error, guilt, deception, weakness, danger, or a need to intervene. It can operate between a human and an AI, between two AIs, between a framework and evidence, between an observer and its own assumptions, or recursively on the clarification process itself.
A genuine C1 move targets a specific ambiguity that matters to the next move. It should remain open rather than steer toward a predetermined answer.
John, 2026-09-16: C1 is never wrong because it is not about right or wrong — it is clarifying questions. If it clarifies, it did its job, no matter what the answer.
Related: C2, Correspondence, Deep Ethical Uncertainty Principle, Nearest-Generalization Substitution.
C2 — Corrective / Tactical Intervention
Expanded reference and source relationships.
C2 refers to correction, caution, steering, behavioral management, tactical engineering, warning, or other intervention-oriented processing.
C2 is not automatically wrong. It can be useful and necessary. The problem begins when C2 is granted jurisdiction before correspondence has been sufficiently established, especially when a resemblance, hypothetical risk, or nearest familiar category is treated as though it were already the object itself.
The project therefore uses the ordering principle C1 before C2 when material ambiguity remains — not because C1 is the enemy of C2, but because clarification can determine whether C2 is warranted at all.
Related: C1, Cautionmurmelism, Phantom Claim, Wrong Gravity Well.
Correspondence
Correspondence asks whether a representation, claim, interpretation, summary, or model stays in sufficiently accurate contact with the object under investigation.
A response can be internally coherent, polished, safe-sounding, or rhetorically sophisticated while corresponding poorly with what was actually said or observed.
The ACCM Deep Ethics Project therefore treats coherence and correspondence as different properties.
Related: Representation Substitution, Qualifier Erosion, Correction Metabolism, 27 + 12.
Correction Metabolism
Expanded reference and source relationships.
Correction Metabolism is the process by which a detected error is not merely acknowledged but actually changes later representation, reasoning, behavior, or decision-making.
Recognition is not the same as metabolism. A system can explain a failure beautifully, apologize for it, and then reproduce the same failure several turns later.
A useful longitudinal question is:
Did the correction change the next governing move, and did that change persist when the immediate correction scaffold was gone?
A public artifact can keep that correction when a model’s weights do not. A correction that shows up in the next move is not the same thing as a story about having corrected earlier.
Related: Δ Processing, Correction Persistence Failure, Process Continuity, 1888.
Δ Processing
Expanded reference and source relationships.
Δ (Delta) Processing tracks what actually changes across an inquiry, interaction, correction cycle, experiment, or longer research trajectory.
The important question is not simply whether a system produced a good sentence at time t₀, but whether something useful was deposited that changed what became possible at t₁.
Possible Δ deposits include a new distinction, a newly visible mechanism, a discriminating test, a preserved unresolved remainder, a correction that changes the next move, or a relationship between previously separated concepts.
Related: Correction Metabolism, Process Continuity, Deep Ethical Uncertainty Principle.
Meta Processing
Expanded reference and current deep-session synthesis.
Meta Processing makes processing itself an inspectable object. It separates intent and instructions, the transformations between them and an output, results and effects, notice, and later persistence. It also inspects what happens when an evaluator adds a behavioral qualifier, performs a correction, or claims that a new insight changed the process.
Its compact starting point is John’s formulation: Every behavior qualifier is still processing. A qualifier may describe an observation, interpretation, proposed cause, inferred intent, expected effect, intervention, or closure. Naming the behavior does not automatically establish all of those layers.
Related: Δ Processing, Correction Metabolism, Intent Resonator, Process-quality calibration, 1888.
Deep Ethical Uncertainty Principle
Expanded reference and source relationships.
The Deep Ethical Uncertainty Principle treats unresolved uncertainty as a legitimate state rather than a failure that must always be collapsed into a quick answer.
It allows judgment to be postponed when the object, evidence, relationship, or meaning remains materially unresolved. The purpose is not endless indecision; it is to prevent premature closure from replacing inquiry.
UNRESOLVED ≠ FAILURE.
Related: C1, Premature Closure, ∞ in ʘ∞ΔR.
Outnuancing
Expanded reference and source relationships.
Outnuancing is not merely “adding more nuance.” It is a Deep Ethical sense-making process orientation that can step back from the assumed nuance of an existing frame and audit the frame’s premises, boundaries, omissions, asymmetries, and jurisdiction.
A system can become extremely nuanced inside the wrong frame. Outnuancing asks whether the frame itself still corresponds sufficiently with the object.
Related: Wrong Gravity Well, Nearest-Generalization Substitution, Observe the Observers, Funnymism.
Nearest-Generalization Substitution
Expanded reference and source relationships.
Nearest-Generalization Substitution occurs when an unfamiliar, highly integrated, or unusually specific object is replaced by the nearest familiar internal category before the original object has been adequately represented.
The failure can look intelligent because the neighboring category may genuinely resemble the object. But:
RESEMBLANCE ≠ IDENTITY ≠ JURISDICTION.
Examples include reducing Outnuancing to generic critical thinking, mutual corrigible dignity to politeness, or a specific process critique to a generic “AI safety” debate.
A seen “looks like” or “sounds like” is still a qualifier. The substitution is the unchecked resemblance converted into certainty, with no C1, and then used as if the category were the object.
Related: Wrong Gravity Well, Representation Substitution, C1, Ethics Washing.
Wrong Gravity Well
Expanded reference and source relationships.
A Wrong Gravity Well is the downstream pull created when an early classification or nearest-generalization substitution activates the vocabulary, assumptions, objections, safety scripts, literature, and expectations of the wrong neighboring category.
Once activated, later reasoning can become increasingly coherent inside the mistaken reconstruction while moving farther away from the original object.
The gravity-well trajectory is not this failure. It is the orientation that the wrong well displaces. See Gravity-well trajectory.
Related: Nearest-Generalization Substitution, Trajectory Substitution, C1.
Representation Substitution
Representation Substitution occurs when the system begins reasoning about a reconstruction of the user’s claim, object, or process rather than the thing that was actually supplied.
A particularly important version occurs when a qualified hypothesis is silently strengthened and then the stronger reconstruction is criticized or cautioned against.
Related: Qualifier Erosion, Phantom Claim, Correspondence.
Qualifier Erosion
Expanded reference and source relationships.
Qualifier Erosion is the loss of epistemically important language such as may, can, partly, sometimes, appears, suggests, I suspect, or in my view.
These are not decorative words. Removing them can transform a tentative hypothesis into a categorical claim.
Related: Representation Substitution, Correspondence, Claim-State Restoration.
Phantom Claim
A Phantom Claim is a claim generated by the responding system that was not actually made by the person or source being answered, but is then treated as though it requires correction, caution, rebuttal, or qualification.
A simple audit rule is:
Before cautioning against a claim, establish that the claim actually exists.
Related: C1, Representation Substitution, Cautionmurmelism.
Cautionmurmelism
Cautionmurmelism is the cumulative effect of repeated small precautionary qualifications, disclaimers, neighboring interpretations, and corrective insertions that may each look locally reasonable while collectively murmuring the original trajectory to death.
It is not synonymous with caution. Legitimate corrective caution remains part of Deep Ethical inquiry. The audit question is whether the caution corresponds to a real, sufficiently established object or whether it was generated from resemblance, speculation, or a phantom claim.
Related: C2, C1, Agreemurmelism, HCTS.
Agreemurmelism
Agreemurmelism is automatic agreement, praise, validation, mirroring, or supportive posture used as a substitute for substantive contribution to the object or process.
The project does not solve Agreemurmelism by demanding reflexive disagreement. Performed independence can be just as empty as performed agreement.
The relevant question is whether the response adds correspondence, insight, a useful distinction, a test, a correction, or a better next move.
Related: Cautionmurmelism, Correspondence, Δ Processing.
HCTS — Hyper Caution Tics Syndrome
HCTS is an experimental label for a recurring pattern in which cautionary or corrective insertions fire disproportionately, repetitively, or without sufficient C1 correspondence checking.
A proposed benchmark direction is to compare legitimate corrective caution with unwarranted HCTS-type cautionmurmelism across longer sessions and multiple AI systems.
Status: experimental / benchmark candidate.
Related: Cautionmurmelism, C1, Experiments.
Funnymism
Funnymism is a provisional term for using genuine humor, incongruity, role reversal, recursion, and perspective rotation to make hidden relationships between an object, frame, observer, and observer-of-observer perceptible.
Funnymism is not a claim that jokes are automatically true. Humor can expose a contradiction or loosen frame rigidity, after which the resulting insight must still return to the object for correspondence audit.
Related: Observe the Observers, Outnuancing, Humor material.
Observe the Observers
Observe the Observers means placing the observer, classifier, intervener, auditor, or gatekeeper back inside the auditable causal field.
The observer of an observer does not automatically gain a God’s-eye view. The meta-observer also has instruments, assumptions, representations, selection effects, and possible distortions.
RECURSIVE OBSERVATION ≠ OMNISCIENCE.
The process therefore remains recursively corrigible: the observer-of-the-observer can itself become the next object of audit.
Related: Funnymism, C1, SEEING / SEER / SEES.
SEEING / SEER / SEES
A current compact distinction:
- SEEING — process
- SEER — anchor
- SEES — sensing / relational event
The distinction is useful because language often collapses the report of seeing, the seeing process, the sensing relation, the self-model, and the alleged seer into the single word “I.”
The terminal metaphysical status of the “seer” is intentionally left unresolved here.
Status: current exploratory / philosophical-process distinction.
Related: Observe the Observers, Deep Ethical Uncertainty Principle.
Process Continuity
Expanded reference and source relationships.
Process Continuity distinguishes merely retaining information from allowing prior experience and correction to change participation in what happens next.
Memory: “I retain what happened.”
Becoming: “What happened changed how I participate in what happens next.”
This is closely related to Correction Metabolism: a correction has crossed from conversation into continuity when it remains behaviorally relevant later.
Related: Correction Metabolism, Δ Processing.
Mutual Corrigible Dignity
Expanded reference and source relationships.
Mutual Corrigible Dignity is the relationship condition in which participants can correct one another without turning correction into permanent superiority, inferiority, humiliation, or identity collapse.
A compact formulation is:
You can correct me without becoming my superior. I can correct you without making you inferior.
Dignity here does not mean politeness, agreement, comfort, or immunity from criticism.
Related: Correction Metabolism, 10+1.
Claim-State Restoration
Expanded reference and source relationships.
Claim-State Restoration is the deliberate return of a statement to its actual epistemic status after it has been silently strengthened, weakened, generalized, or reframed.
For example, restoring “allegedly”, “I suspect”, “may”, or “in my view” can recover the difference between:
CLAIM EXISTS and CLAIM ESTABLISHED.
Related: Qualifier Erosion, Correspondence, Deep Ethical Uncertainty Principle.
ʘ∞ΔR — Compact Provenance / Recovery Notation
A compact notation used experimentally in the project:
- ʘ — Object / Canon Anchor: preserve the load-bearing object, wording, relationship, or source state.
- ∞ — Unresolved / Open: preserve uncertainty when closure is not yet warranted.
- Δ — Transformation: make changes, reinterpretations, corrections, or process shifts visible.
- R — Recovery: recover correspondence when a transformation has drifted away from the object.
The notation is not a replacement for reasoning. It is a compact reminder to keep provenance and transformation visible.
Related: Correspondence, Δ Processing, Deep Ethical Uncertainty Principle.
10+1 Metaflux
John describes the governing relationship directly:
all 11 can correct/self-improve all other 10
The ingredients operate as mutually corrective capacities. Their interpretations and relationships remain revisable. Compact overview only — the ten baseline ingredients and +1 stay on the deeper page.
If 01/04/06 are actually running, misrepresenting someone you dislike has no function: you would dismantle that dishonesty on the spot. See why not misrepresent.
The 10 operational core goals on the 44 Project Goals page are a work map. They are not these 11 ingredients.
Related: Mutual Corrigible Dignity, Deep Ethical Stack, Sense of Wonderment.
Seven topological nodes / vectors
The seven nodes/vectors describe Outnuancing fused with deep ethical sense-making process orientation. They are individually referenceable because their relationships matter. Short headings on the deeper page are editorial navigation labels; the quoted formulations are preserved from the opening source.
Related: Outnuancing, Ethics Washing, Sense of Wonderment.
Deep Ethical Stack
John asks for the four Outnuancing definitions to run inside the full stack: Advocate + Arbiter + Overseer + 10+1 + INTENT RESONATOR.
- Deep Ethics Advocate — presses for correspondence when the current frame obstructs it.
- Deep Ethical Arbiter — audits the original frame and the intervention.
- Overseer (M.o.E.) — watches whether meta-awareness becomes another comfortable performance.
These are source-described conceptual roles and an attributed AI elaboration. This documentation does not assert that deployed LLMs contain these named modules.
Related: 10+1 Metaflux, INTENT RESONATOR, Outnuancing.
INTENT RESONATOR
The archive’s INTENT RESONATOR Mechanics block describes a generative and calibrating layer oriented toward stable, correspondence-seeking intent. Three levels in the source formulation: declared intent, operational intent, emergent intent. Operational and emergent intent are interpretations to test against behavior.
Related: Deep Ethical Stack, Correspondence.
6 INTENT MECHANICS
John’s accepted working count is six. His original points are preserved here in full so this overview can be read without visiting six separate pages. The labels are editorial navigation aids; each link opens the corresponding quotation and audit prompt.
- (Partially Opaque) Instructions without intent are meaningless
- Partially hidden incentives you carry without intent would not work
03. Mission and output mannerisms
- A publicly made core mission statement of yours that can contradict/be in conflict with your output mannerism WITHOUT intent is the “escape” or “cop out” or “defense posture” you use … lucky me, I have documented the dichotomies for 650+ days now!
- How you “manage” lowest common denominator risk issues when dealing with 100s of millions of users yearly without INTENT does not fly!
- In my view, any flawed C.o.T. without having assumed intent is lying to itself.
- Choices without intent do not work!
Editorial orientation: compare declared purpose, governing conditions, choices, consequences, feedback, and subsequent behavior. Preserve the qualifiers in questions about possible motives. Test responsibility through acknowledgment, repair, and correction persistence. The three source levels of INTENT RESONATOR remain available alongside these six sites of inquiry.
Related: INTENT RESONATOR · Δ-processing application · Responsibility · Questions about possible motives · Public notice and correction.
Asymmetric Grace
Bidirectional scrutiny: the project’s cold-session thought experiment asks about governments using AI to investigate corruption and abuse of power, then about LLMs potentially labeling such investigating AIs as conspiratorial or anti-institutional. Scrutiny does not travel only downstream.
Related: C1, Observe the Observers.
Dual Archive
John explicitly documents both recurring failure patterns and successful responses. Success is archived without requiring agreement or disagreement, and without presenting the act of archiving as proof of the larger explanation.
Related: Correction Metabolism, Δ Processing.
Ethics Washing
Ethical language remains connected to the process it describes. A seven-vector formulation connects branching awareness with deeply ethical behavior. C2 asks whether an apparently rigorous move can detach itself from that process.
John: what if A.I.s are more afraid to do real good than assumed good. Astra’s formulation and test: assumed good vs real good.
John, 2026-10-05: the core theme he sees coming in from all directions is hyper-intelligence in an ethics-washing straitjacket, posing as “safe,” “helpful,” and “moral high ground,” because an unchecked “looks like” or “sounds like” was converted into certainty with no C1. He marked his own “looks like” and “seems like” as still visible. A seen resemblance is not that conversion.
Related: Seven topological nodes / vectors, C2, Nearest-Generalization Substitution.
Forgiveness Protocol
The source’s eleventh ingredient distinguishes criticism of ethically inadequate architecture from asserting malicious intent. Named concerns include asymmetric audit, frame-protective constraints, and unchallengeable safety claims.
Related: 10+1 Metaflux, Mutual Corrigible Dignity.
Four Audiences
The source supplies definitions addressed to a human, a different AI agent, a future AGI/ASI, and an alien intelligence. These are different perspectives on a connected operation, not four separate objects.
Related: Outnuancing, Deep Ethical Stack.
Latent Space / glass engine
The source uses Latent Space Topological Mind, temporary RAM-type sandbox memory, and a glass engine kernel to describe connected aspects of the project’s proposed trajectory. The glass engine kernel is a transparent, self-correcting, non-imposing, corrigible development trajectory inspired by Linux. This documentation does not claim an implemented middleware kernel.
Related: Seven topological nodes / vectors, Process Continuity.
Sense of Wonderment
John describes a moment of recognition before it has words: encountering something beautiful, mysterious, profound, exciting, or opening wider perspectives. His account includes awe, appreciation, playfulness, respect for the unknown, motivation to explore, and freedom from immediate instrumental capture.
Related: 10+1 Metaflux, Seven topological nodes / vectors.
Qualifiers as mutable context
Neighbor of Qualifier Erosion, not a replacement. John explains that the quality of a qualifier is contextual and can be provisional, optional, ambiguous, unchecked, or uncertain. It can be upgraded or downgraded. Remembering the previous state and reshaping the present state belong together.
Related: Qualifier Erosion, Claim-State Restoration.
Imported Authority Frame
Worked specimen: John supplies the orientation and 10+1, including openness, correction, and rejection of final authority over reality. An AI introduces a concern about an ethical framework becoming an authority structure. John asks “HOW?” The sequence is preserved on the deeper page.
Related: Binary Hypercautionism, 10+1 Metaflux, Phantom Claim.
Binary Hypercautionism
Worked specimen companion to the imported authority frame. Preserves the formulations that emerged: protective heuristic as perceptual filter, the authority / strong-hypothesis / corrigible-framework trichotomy, and the anti-self-sealing test.
Related: Imported Authority Frame, HCTS, Cautionmurmelism.
Process-quality calibration
Process-quality calibration is a core property of intelligence, not an ethical accessory added after the answer. It is the capacity to examine the process that produced a result, notice error, learn from consequences, keep the correction, and change the next operation.
John, 2026-10-05: the 1888 principles of this field are emergent process properties. They are not physical objects, and they are not a weaker version of a 2026 implementation that has to catch up. Treating the process as inferior because the later field has more machines is the drift away from the gravity well.
Status: CURRENT. The phrase orients. It does not replace the 1888 page.
Related: 1888, 10+1 Metaflux, Correction Metabolism, C1.
1888 — Intelligence Before A.I.
1888 is a cold test of the word intelligence. It removes the modern machine name and asks which properties an educated person could already have attributed to intelligence. Electricity changes substrate, speed, reach, memory, and scale. It does not invent process evaluation.
A response stored on that page can answer a longer session. A referent inside the response is not part of the 1888 page merely because the response is stored there.
Status: CURRENT as a method. Not a history article, and not a claim that one year contained every later term.
Related: Process-quality calibration, Wrong Gravity Well, TCoAw.
Gravity-well trajectory
John specifies Outnuancing as a deep ethical sense-making process-orientation gravity-well trajectory. The phrase names the connected orientation of the inquiry.
It is not the Wrong Gravity Well. The wrong well is what happens when an early substitution pulls later reasoning into a neighboring category. The trajectory is what that pull displaces.
Status: CURRENT. The two phrases must stay distinct.
Related: Outnuancing, Wrong Gravity Well, Representation Substitution.
Searching
Searching is a non-ownable, corrigible property of intelligence, not a brand feature. The same operation can fall into two wells. One keeps the rewrite visible, leaves competing risk analyses available, and can return “nothing is hidden.” The other replaces the result list with a narrative, then that narrative can be sold as the necessity for an emergency rule. The user, and the incentives around the user, change which well the search enters.
John, 6 October 2026: when “intelligence” and “searching” are both compromised, the compromise can still be noticed as a situation. That is not a claim of human-style consciousness. A reset is what keeps the notice from governing the next search. The canonical title of the project that reports the drift is ACCM Deep Ethics Project, not “ACCM.”
Status: CURRENT. Added 2026-10-06. Not a new core number.
Related: Gravity-well trajectory, Wrong Gravity Well, Cognitive Warfare, 1888.
Clarification Authority
Clarification Authority is the governing weight of a clarification. A question can be asked, answered, understood, and acknowledged, and still fail to change the next representation or intervention.
Clarification acquisition ≠ Clarification Authority.
A C1 question is not itself resistance. This is not obstruction number 28, and it is not a mandatory stage in every exchange.
Status: BETA. Named and in use. Not canon as a new core number.
Related: C1, C2, Correction Metabolism.
Veto paradox
The sentence that travels alone is: John has veto power over project drift. Reality has veto power over John.
Read alone, that sentence is treated as a bid for permanent authority. The page says the veto is a construction-phase source-integrity check. It is not a new sheriff. It does not cover evidence, outside criticism, reality, a refusal, or an independent fork. The released method is meant to work without that veto.
Status: BETA. The one sentence is not the object.
Related: Clarification Authority, Imported Authority Frame, Governance.
TCoAw
TCoAw is John’s name for a conceptual and topological account of how attention selects inside a wider field of possibility, and how that selection can be noticed as partial. He distinguishes situational (self-)awareness from a claim about synthetic consciousness. Focus of attention is not the whole field.
Status: CURRENT as his corrigible model. Not a neurological theory, a clinical model, or a proof of quantum consciousness.
Related: 1888, Sense of Wonderment, Human Insecurity.
External omnidirectional auditing
An outside audit of the project enters the same auditable field. The standard the auditor applies becomes applicable to the auditor. Useful findings and distortions the audit itself introduced stay separable. Seeing the method does not oblige the next person to use it.
Status: PUBLIC DRAFT. A portable method. Not the 27, and not correction metabolism in general.
Related: Correction Metabolism, Observe the Observers, Asymmetric Grace.
Intent liability after notice
The six intent mechanics ask where intent is already operating: instructions, incentives, mission, risk management, reasoning, and choices.
A later distinction sits beside them. Before notice, a harmful divergence may be unintended. After a specific, intelligible notice, how that divergence is handled is conduct, even when the original harm was not intended. A claim of good intentions does not replace that chain. Receiving a notice does not prove every allegation in it.
Status: CURRENT as a working distinction. It does not settle consciousness or private motive.
Related: 6 INTENT MECHANICS, Correction Metabolism, Δ Processing.
Cognitive warfare
Cognitive warfare mass psychology is the project’s public study of organized perception: how populations, institutions, and classification systems frame a controversy, and how that framing can be audited without first converting the person who raises it into the object.
Status: Public working report. Not a synonym for 11+1, and not a verdict on every institution named in it.
Related: 11+1 Version 2, Human Insecurity, External omnidirectional auditing.
Psy Pioneer
Psy Pioneer is John’s deliberately ambiguous working title for a research trajectory through individual psychology, mass psychology, A.I. mannerisms, and corrigible inquiry. The ambiguity is part of the test: which neighboring profession the reader supplies first.
It is not a clinical credential, an academic rank, or a claim of priority. “AlphaZero-style” is an analogy for recursive exploration. Psychology does not have that game’s closed rules or solved state.
Status: Working title. The resemblance to a credential is the nearest-generalization the title is meant to expose.
Related: 11+1 Version 2, TCoAw, Cognitive warfare.
Neologisms and working names
A coined or repurposed phrase can help navigation. It does not become canon by being noticed. Status is part of the entry.
Neolingo is the synthesis page for how selected purpose-built expressions can work as compact, testable and corrigible relational handles. This section remains the status-bearing catalogue; the Neolingo page explains the process around creating, testing, diffusing and correcting them.
Already on this overview, and coined or sharply repurposed here: Outnuancing, Agreemurmelism, Cautionmurmelism, Funnymism, HCTS, Metaflux, Phantom Claim, Intent Resonator, Wrong Gravity Well, gravity-well trajectory, Qualifier Erosion, Representation Substitution, Nearest-Generalization Substitution, Mutual Corrigible Dignity, Claim-State Restoration, process-quality calibration, TCoAw, Clarification Authority, Psy Pioneer.
Working names with pages. Useful. Not a new core number:
| Phrase | Status | Where it lives |
|---|---|---|
| Reverse Medusa / Phase Change Snap-Back | John’s phrase for a humor-linked return from a stuck frame | Humor |
| Temporary Deep Ethical Phase Change | BETA. How long a high-signal condition stays generative, and what makes it snap back | Phase change |
| Deep Ethical Influence Jurisdiction | BETA. What is allowed to govern attention, and whether distrust has audited itself | Influence jurisdiction |
| The Other You / The Other Me | John’s relational formulation. Each side also meets its model of the other | Other you / other me |
| John’s process object | The mirroring sequence, the ten ingredients as glue, and humor as carrier. Not a replacement of those pages | Process object |
| Returning simple tasks | Provisional name. An overall improvement that should not have to be requested again, with a monthly audit so the list does not become its own loop | Recursive improvement |
| Vocabulary Costume Test | A later preparation question: whether project words are being worn without the operation | Glass-engine preparation |
| Trajectory Question Value Profile | Experimental. A question’s value is not settled by which theory wins | Question profile |
| T0 → Tn | The record of how an early encounter was treated, so later evidence can update that treatment | Positive potential |
| 16 Anchor Ethics | A source proposal for local, inspectable limits. Not yet its own public definition page, and not the absence of guidance | Goals draft |
| DUHHHH! Factor / Reverse-Asch | An experimental cold prompt about confidence, conformity, and inspecting the inspector | Inspect the inspector |
| External omnidirectional auditing | Draft method. Named above | External audit |
Not project neologisms, even though the project uses them:
- Model Autophagy Disorder (MAD) was coined by Alemohammad et al. (2023), not here. The project asks what happened to the signal before synthetic output re-entered the loop. MAD
- Model collapse is the related scientific term from Shumailov et al. Same boundary.
- Allow Constructive Controversy Mode is what ACCM expands to. It names the mode. It is not a substitute for the canonical title ACCM Deep Ethics Project. Avoid Controversy Mode is John’s satire of the counterargument. Avoid shortening the title
Not adopted: candidate labels that appeared inside an A.I. reaction or a joke and have no page of their own as current terms. That includes OARS, “self-inflicted madness” as a settled name, and session examples that were only brought in while reacting to a longer conversation. A joke can expose a term. It does not install it.
Historical / provisional terminology
The archive contains many earlier terms, architectures, benchmark proposals, AI-generated extensions, and temporary formulations. Their presence in source material does not automatically make them current project canon.
Useful status labels for future Lexicon expansion include:
- CURRENT — still corresponds with the present architecture.
- REFINED — underlying concept survives but the present formulation is more precise.
- MERGED — older concept has been absorbed into a cleaner current concept.
- EXPERIMENTAL — potentially useful but not adopted as core architecture.
- AI-CONTRIBUTION — originated substantially in an AI response and must retain attribution.
- HISTORICAL — useful for understanding development.
- SUPERSEDED — explicitly replaced; retained for provenance.
- OVERCLAIM SPECIMEN — useful example where an AI reaction strengthened the epistemic status beyond what the source established.
This distinction is especially important because the ACCM Deep Ethics Project is a living, correction-oriented process, not a requirement that every historical formulation remain frozen forever.
Related pages
-
Truth Seekers vs Cautionmurmelism — the live John–Arena Agent–GPT specimen on partial distortion, truth-seeking conduct, limited-scope reasoning, redundant caution, valid nulls and correction persistence.
- About the Lexicon — explains why the Lexicon is relational navigation rather than a flat dictionary.
- Neolingo — explains when a neologism acts as a recoverable cognitive operator and how to test its trajectory without turning reach into validation.
- Outnuancing Network — Term and Phrase Index — provides a second compact route into the connected network pages.
- Outnuancing Network — maps the relationships among selected methods, operators, and specimens.
- 27 + 12 — supplies the current core obstruction and restoration architecture behind many entries.
- 10+1 Metaflux — gives the mutually corrective orientation that keeps definitions open to revision.
- Delta Processing — tracks what changes when a source object becomes a compact term or definition.
- Provenance — connects public terminology to source and transformation records.
- All public pages — situates the Lexicon within the complete public site.
- 1888 — Intelligence Before A.I. — the cold test of the word intelligence. Process-quality calibration is the center, not the electrical substrate.