Reconcile dynamic coherence with evidence and existing methods

[raelovejoy]

Why

The September 24 Root Sequence work on dynamic coherence/adaptive continuity appears to close a real conceptual gap, but it also creates a research obligation: distinguish what is a useful synthesis from what established literatures already define more precisely, and correct older UCF pages that overstate cross-domain support.

Source context: Root-Sequence/root-sequence#21 (draft).

What this PR does

Adds a targeted evidence reconciliation

docs/dynamic-coherence-evidence-audit-2026-09-24.md checks the synthesis against:

• Di Paolo on enactive adaptivity / viability / agency;
• McEwen on allostasis and allostatic load;
• Holling and Walker et al. on resilience, adaptability, and transformability;
• Tognoli & Kelso on metastability;
• Shoda et al. and Nowak et al. on dynamical/attractor models of personality;
• the capability approach on real/substantive opportunity;
• Aubin's viability theory on constrained reachability;
• Klyubin/Polani/Nehaniv and Jung/Polani/Stone on information-theoretic empowerment.

The audit explicitly records what each literature can and cannot support.

Extends model v1.1 without replacing it

The temporal layer now asks:

• what changed only in current state;
• what changed in later response dynamics;
• what causal history remains retained;
• which future possibilities open or close;
• which lock-ins/dependencies were created;
• who retains the ability to revise the system.

It treats Root Sequence's agency envelope and meta-agency as provisional comparison prompts, not validated UCF measures.

Corrects active navigation/status

• Speculation no longer says its reading order mirrors how real systems grow through Chaos → Tension → Flow → Unity.
• The folder no longer describes all 30 speculation pages as complete research-ready work.
• The Foundations index flags coherence-and-systems-theory.md for high-priority rewrite.

Preserves but clearly marks overreaching older work

Adds audit notices to the older:

• systems-theory foundation;
• attractors/basins speculation;
• resilience/adaptation speculation;
• developmental-trajectories speculation;
• cross-scale speculation.

The historical text remains intact for provenance. The notices make clear which claims the current model does not adopt.

Important finding

The new synthesis does not require rewriting everything about coherence.

The strongest current UCF pieces already contain much of the needed architecture:

• contextual functional fit;
• relational consequences;
• normative judgment kept distinct;
• internal / inter-agent / systemic / temporal layers;
• anti-scalar rule;
• boundaries and timescales;
• revision and adaptation;
• explicit failure criteria.

What changes is the temporal resolution: UCF now needs to distinguish state from transition dynamics and inspect future possibility, retained history, lock-in, and revision power.

What does require correction is older material that treated:

• the four UCF labels as generalized attractor categories;
• Chaos → Tension → Flow → Unity as a developmental law;
• Flow/Unity as inherently more resilient;
• metastability/allostasis/attractors as universal cross-domain mechanisms;
• cross-scale requirements as established universals.

Proposed validation method

Do not validate UCF by showing it can describe a case.

Instead:

1. choose a bounded public, non-sensitive case in advance;
2. select a competent existing domain method as baseline;
3. freeze the UCF/dynamic prompt;
4. give both methods the same evidence;
5. compare consequential omissions, false alarms, review burden, extra information demands, disagreements, and decision-relevant findings;
6. publish null results if UCF adds nothing.

Recommended first domains: infrastructure/service fallback, transit accessibility under disruption, repair/service architecture, or community degraded mode — not trauma or consciousness.

Relationship to existing PR #5

PR #5 is a context refresh touching START-HERE / claims / evidence / PROJECT-CONTEXT. This branch is based on current main and deliberately avoids those files, so the evidence reconciliation can be reviewed independently.

What this does not claim

• no systematic review;
• no novelty proof;
• no universal dynamic-coherence variable;
• no dynamic-coherence score;
• no universal agency metric;
• no AI-consciousness or AI-trauma inference;
• no clinical trauma model;
• no empirical validation of UCF;
• no silent rewriting of historical files.

Review questions

1. Is the state-vs-transition distinction genuinely useful in UCF, or redundant with ordinary temporal analysis?
2. Should agency envelope remain Root Sequence vocabulary while UCF cites capabilities/viability/empowerment as comparators?
3. Which four-state documents should be rewritten versus archived after the audit?
4. Which competent baseline method should be used for the first added-value comparison?
5. If the comparison finds no unique value, which parts of UCF should be retired rather than expanded?