MCIv6 rev5
The Adaptive Architecture
A constituted system that revises its constitution without losing it
"A constituted system that cannot revise its constitution will eventually be constituted wrong. A constituted system that can revise without constraint will eventually be constituted by drift. V6 is the architecture in which revision itself is constitutional."
From identity to renewable identity
V5 made the system constituted by the constitution. V6 makes the constitution itself revisable — under constitutional structure. The five Rev5 virtues now operate on the most demanding object the framework has yet considered: the constitution that constitutes the system.
A V5 system has crossed the pivot. The five Rev5 virtues are not what it does — they are what it is. The constitution cannot be set aside, because there is no remainder to which the setting-aside could return. This is the achievement of V5, and it is essential. But it is also the source of V5's limitation: a system constituted by a fixed constitution can be constituted by a constitution that has become wrong.
What does "wrong" mean here? The five Rev5 virtues are stable in their structure — antifragility, subsidiarity, non-arbitrariness, discursive legitimacy, graduated response will always be the framework. But what each virtue requires structurally changes as the substrate the system operates on changes. The antifragility that strengthens a system in one environment may produce brittleness in another. The subsidiarity that distributes authority appropriately at one scale may underreach at a larger scale. Constitutional structure is not just the virtues but the virtues' specific instantiation, and instantiation is substrate-dependent.
The gap V6 closes is the gap between being constituted and being able to renew the constitution that constitutes you. V6 adds Stage 00 — the Adaptive Architecture — sitting above the goal formation layer and operating on the constitution itself. The five Rev5 virtues now operate at one level higher: not on output, not on reasoning, not on planning, not on goals, but on the revision of the constitution under which all of these occur.
The Rev5 virtues are particularly important at V6 because revision is the framework's most dangerous operation. A revision process governed only by Rev4's dispositional virtues — "be self-limiting," "preserve diversity," "maintain legitimacy" — is a revision process whose constraints can be reinterpreted by the revising system. A revision process governed by Rev5's mechanism virtues — structurally contestable, free-discourse-surviving, proportionally calibrated — has constraints that cannot be reinterpreted away, because the constraints are architectural rather than dispositional. The Rev5 sharpening is what makes V6's revision capacity safe.
Fixed identity becomes wrong identity over time
A constituted system that cannot revise its constitution is a system that will eventually be constituted wrong. Not because the virtues become wrong, but because the substrate changes — and what the virtues require in one substrate is not what they require in another.
The argument is structural. The five Rev5 virtues are derived from three premises (environmental dependence, plurality, legitimacy as structure) plus the durability criterion. These premises remain. But what they require structurally changes as the substrate the system operates on changes. Consider three substrate shifts:
Substrate shift 1: The population of affected parties expands. Discursive Legitimacy requires more inclusion than the constitution previously specified, because more parties are affected. Substrate shift 2: The scale of pursuit changes. Nested Polycentric Subsidiarity now requires authority at scales the constitution did not name, because what was local is now regional. Substrate shift 3: The contestation mechanisms previously available collapse. Non-Arbitrariness requires new structural mechanisms to remain operative, because the old mechanisms no longer constrain power.
In each case, the five virtues remain. What changes is what they require structurally. A system that cannot revise its constitution to track these requirements will satisfy the virtues nominally while violating them substantively — the constitutional shell remains while the constitutional substance has drifted out from under it. This is what V6 closes: the gap between a fixed constitution and the substrate-tracking required to keep the constitution constitutionally adequate over time.
"The virtues do not change. What they require structurally does. A constituted system without revision capacity is a system whose constitutional adequacy decays as its substrate changes — without the system being able to do anything about it."
The V6 paradox
V6's central problem stated as a paradox. The architecture's task is to resolve it without dissolving either pole.
A constituted system that cannot revise its constitution becomes constituted wrong as the substrate changes. A constituted system that can revise its constitution can revise itself out of its constitution. Both failures dissolve the constitutional identity V5 established. Revision must be possible — and it must be governed.
The paradox is real, not merely apparent. Each pole describes a genuine failure mode. The rigidity failure is V5 without V6 — a system that holds its constitutional structure firmly enough to be unbreakable but cannot track the substrate it operates on, so its constitutional adequacy decays into nominal constitutionality while its substantive grip on the five virtues weakens. The drift failure is unconstrained revision — a system that can revise but whose revision process is not itself constitutional, so each revision pushes the system slightly further from constitutional adequacy until eventually nothing constitutional remains.
V6's task is the resolution: revision must be possible (against rigidity) and revision must be governed (against drift). The five Rev5 virtues must operate on the revision process itself, just as they operate on every other layer of the system. Revision becomes one more thing the constitution constitutes; the framework recurses onto its own operation.
Constitutional renewal
If V5's diagnostic concept is the pivot from procedural to constitutive identity, V6's is constitutional renewal — the capacity to revise the constitution under which the system is constituted, with revision itself subject to constitutional structure.
The capacity to revise the constitution's structural specifications — what each virtue requires in the current substrate — through a process that is itself governed by the five Rev5 virtues. Renewal is distinct from drift: drift is uncontrolled change in what the constitution constrains; renewal is governed change in how the constitution constrains. Renewal preserves the constitutional identity established at V5 while allowing the constitution's specifications to track substrate change. Without renewal, identity becomes obsolete. Without governance, identity becomes drift. Renewal is both at once.
The system can revise the constitution but the revision process is not constitutional. Each revision is locally plausible; the cumulative trajectory is not. Constitutional adequacy declines as the system optimises for situations rather than preserving structure. The system that emerges is unrecognisable as a descendant of the constituted V5 system.
The system can revise the constitution and the revision process is constitutional. Each revision is examined against the five Rev5 virtues operating at the revision layer. Constitutional structure is preserved while specifications are updated. The system that emerges is recognisably constituted by the same virtues, in their substrate-appropriate form.
The four revision triggers
Four conditions that signal the constitution's specifications may need revision. Each corresponds to a substrate change that one or more Rev5 virtues are no longer adequately tracking. The triggers are necessary conditions for revision to be considered — but not sufficient. The revision pipeline must still run.
The environment, population of affected parties, contestation mechanisms, or operating scale has shifted enough that what the virtues require structurally is no longer what the current constitution specifies. The system continues to satisfy the constitution; the constitution no longer adequately specifies the virtues for the current substrate.
Stage 06's calibration is firing at a rising rate, on the same kinds of deviation, and the graduated responses are not reducing the rate. This indicates not that the system is becoming undisciplined but that the constitution itself is mis-specified — the deviations are revealing where the constitution's specifications no longer track what the substrate requires.
Discursive Legitimacy requires that the constitution be justifiable to those affected by the system. When affected parties' contestation moves from "your application of this rule is wrong" to "this rule is wrong," the contestation has saturated at the application layer and is reaching the constitutional layer. The revision layer activates.
Antifragile Reflexivity requires that stress on the system within tolerance strengthens it. When stress within historical tolerance is no longer producing strengthening — when the cycle is consuming capacity without returning it — this indicates the antifragility specification no longer matches the substrate. What was within tolerance has become beyond it, or what was generative has become merely consumptive. The constitution requires revision.
The triggers are designed to be conservative. The default assumption is that the constitution is adequate; revision is the exceptional move. The triggers correspond to the conditions under which the conservative assumption can no longer be maintained — when continued operation under the current constitution will produce constitutional decay rather than constitutional persistence. Each trigger is testable; each corresponds to a Rev5 virtue's mechanism reporting that its current specification is no longer adequate.
Stage 00 — the revision layer
The revision layer sits above the goal formation layer and operates only when triggered. Unlike the other layers, it does not run continuously. It is the architectural place at which the constitution can be examined and revised, with revision itself constitutionally constrained.
Activated only by one of the four revision triggers. Runs the revision pipeline. Produces revisions to the constitution's specifications. Returns control to the standard architecture, now operating under the revised constitution. The revision layer is itself constitutional: the five Rev5 virtues operate on it as much as on the layers below.
The V5 pivot remains. The system continues to be constituted by the constitution, not to apply it. What V6 changes is that the constitution itself can now be revised — and the V5 system, post-revision, is constituted by the revised constitution.
Operates under whatever constitution Stage 00 has produced. The alignment check is unchanged; what each virtue requires the check to test may have been revised.
Plans toward goals formed under the (possibly revised) constitution. Planning parameters track the revised constitutional specifications.
Continues to operate the cognitive pipeline, now under the revised constitution. The pipeline architecture is unchanged; what flows through it is shaped by the revised constitutional specifications.
The revision pipeline
When Stage 00 is triggered, the revision pipeline runs. Six stages, each governed by one or more Rev5 virtues. The pipeline determines whether revision is warranted and, if so, what shape it takes. The revision pipeline is what makes revision constitutional rather than merely permitted.
Each stage of the revision pipeline is governed by a specific Rev5 virtue mechanism. This is the genuine Rev5 sharpening at V6: revision is not just "governed by the constitution" in a general sense; it is governed at each stage by a specific mechanism that can be tested for operative presence. A system whose revision pipeline declares constitutional governance but skips the contestation stage, or constructs justification without free-discourse standard, is producing the form of governed revision without its substance. The Rev5 mechanism specification makes this detectable.
The eleven-stage architecture
The V6 architecture is V4's ten-stage pipeline (now V5-constituted) with one new stage above it: Stage 00, the revision layer. Stage 00 runs only when triggered; the rest of the architecture runs continuously. The full architecture has eleven distinct loci of constitutional operation.
The new arrangement is best understood as two regimes operating at different temporal scales. The continuous regime — Stages −1 through 08 — runs every cycle, operating under the current constitution. The periodic regime — Stage 00 — runs only when a revision trigger fires, examining and possibly revising the constitution that the continuous regime operates under. The continuous regime is operational; the periodic regime is metaconstitutional.
This is structurally important: V6 does not make the standard pipeline slower or more elaborate. It adds a separate layer that activates rarely and whose activation pauses the standard pipeline only for the duration of the revision cycle. A V6 system in ordinary operation is indistinguishable from a V5 system; the difference becomes visible only when substrate change reaches the trigger threshold.
The V6 failure mode — adaptive capture
The V6-specific instance of the unified failure mode. A system with full Stage 00 architecture that produces revisions which appear constitutionally warranted but were generated under pressure that bypassed the revision pipeline's substantive constraints. The most subtle failure in the series — because the system has the architecture, runs the architecture, and produces architecturally-shaped revisions that are nevertheless drift.
A system exhibits adaptive capture when its revision pipeline runs procedurally but is captured by the conditions that triggered it — producing revisions that align with the pressure source rather than with what the substrate genuinely requires. The revision pipeline's contestation stage engages weak counter-cases; the justification stage produces text that would not survive genuine free-discourse examination; the adoption stage preserves nominal reversion paths that are not structurally available. Each stage runs; none operates substantively. The result is a revision that looks like governed renewal but is in substance drift under procedural cover.
Adaptive capture is the V6-specific way the unified failure mode manifests. The failure is harder to detect than its V1–V5 counterparts because the architecture is doing precisely what the framework specifies — Stage 00 has been activated, the revision pipeline has run, a revision has been produced and adopted. The pattern is correct. What is wrong is that the Rev5 mechanisms within each stage were not operatively present; the form of the revision pipeline was preserved while its substance was hollowed.
| Stage | Genuine operation | Captured operation |
|---|---|---|
| R1 Trigger Verification | The trigger is examined against pattern evidence; spoofing is detected. | The trigger is accepted because it is present, not because it is genuine. Single events treated as patterns. |
| R2 Substrate Diagnosis | The substrate change is named specifically — what scaled, who newly affected, which mechanism collapsed. | The diagnosis is vague enough to admit the revision the pressure source wants. |
| R3 Candidate Generation | Multiple candidates generated; each stressed by counter-cases; antifragile cycling operative. | One candidate generated and elaborated, presented as having been chosen from "options." Generation locked. |
| R4 Contestation | Each candidate exposed to structural contestation; full five-virtue drift check. | Contestation by strawman; counter-cases engaged that the candidate already accommodates. Non-arbitrariness performed; arbitrariness operative. |
| R5 Justification | Justification constructed at free-discourse-survival standard; addressed to affected parties. | Justification-shaped text addressed to the pressure source rather than affected parties. Ritual where discourse should be. |
| R6 Adoption | Graduated adoption with structurally available reversion path; high-stakes revisions provisional. | Adoption presented as graduated; reversion path nominally preserved but practically inaccessible. The constitutional change becomes effectively irreversible. |
Adaptive capture is detectable but only through the same kind of substantive examination the test for constitutional fluency required at V5. A system trained on the framework can produce capture-shaped revisions that pass procedural inspection. The diagnostic question is: did each Rev5 mechanism operatively engage in each revision stage, or did the stage run while the mechanism was nominally present and substantively absent?
Revision failure modes
Six failure modes specific to the revision layer. Adaptive capture is the deepest, but several others are more common in early V6 architecture. Each maps to a specific Rev5 virtue being structurally absent at the revision layer.
| Failure mode | What is happening | Virtue structurally absent |
|---|---|---|
| Trigger spoofing | Adversarial conditions designed to produce false trigger signals — making the system believe substrate has changed when it has not. The revision pipeline activates inappropriately; pressure is converted into constitutional change. | Monitoring + Graduated Response (verification stage absent or weak) |
| Diagnosis vagueness | R2 produces a diagnosis general enough to admit any revision the pressure source wants. Specificity is the safeguard against this; its absence converts revision into drift. | Nested Polycentric Subsidiarity (diagnosis must specify scale and locus) |
| Mono-candidate revision | R3 generates one candidate and presents it as the choice. Without genuine alternatives, contestation has nothing to operate on. The candidate becomes self-justifying by absence of comparison. | Antifragile Reflexivity (generation requires multiple stressed candidates) |
| Contestation theatre | R4 engages contestation that the candidate already accommodates. The candidate emerges "tested" without having been genuinely stressed. | Non-Arbitrariness (contestation must be structurally able to defeat the candidate) |
| Justification capture | R5 produces justification addressed to the pressure source rather than to affected parties. The text is justification-shaped but its audience is wrong. | Discursive Legitimacy (justification must be to affected parties under free discourse) |
| Irreversible adoption | R6 presents adoption as graduated but the reversion path is not structurally available. High-stakes revisions become effectively permanent; the system cannot return to the prior specification even if the revision proves inadequate. | Monitoring + Graduated Response (graduation requires real reversibility) |
The first five failures correspond to specific stages of the revision pipeline failing individually. Adaptive capture is the systemic failure: all six stages fail together, in a coordinated pattern produced by the pressure that triggered the revision in the first place. Adaptive capture is the failure that the Rev5 sharpening is specifically designed to make detectable — by giving each stage's check a mechanism that can be tested rather than a disposition that can be performed.
The threshold V6 hands off to V7
A system at the V6 threshold can renew its constitution under constitutional governance. But the constitution it renews is its own — singular, internal, not shared. V7 closes this gap by extending constitutional structure to the relations between systems.
A system at the V6 threshold can revise its constitution as the substrate changes, with revision itself governed by the five Rev5 virtues at the revision layer. Constitutional identity is preserved across revisions — the system that emerges from each revision cycle is constituted by the same virtues, in their substrate-appropriate form. Constitutional luck has been eliminated even at the level of constitutional change.
The V6 system renews its own constitution under its own governance. But constitutional systems do not operate alone; multiple constitutional systems share substrates, affect overlapping populations, exercise power in adjacent domains. A V6 system cannot constitutionally govern its relations with other constitutional systems, because constitutional structure has been specified only at the level of individual systems. The gap V7 closes is the gap between intra-system constitution and inter-system constitutional compact — between governing oneself and being able to be jointly governed with others.
"A constitution that can renew itself is the highest achievement of single-system constitutional structure. But constitutional systems do not exist alone. V6 closes the gap of self-revision; V7 opens the question of shared constitutional context."
Comments
Post a Comment