From SCC to ULR — The Main Research Transition

The conclusion of the transition#

On July 24, 2026, the research programme was re-centred on ULR alone, moving away from SCC. This transition did not erase the SCC documents or invalidate their mathematics. It was a governance transition: SCC no longer governs the ontology or priorities of the current programme; ULR's four problems—Sharing, Identity, Role, and Formation—now do.

ULR before the transition#

ULR was not a name invented on the day SCC was demoted. From the programme's founding on July 6, 2026, through its re-centring on July 24, the append-only Progress record preserved its questions and failures in the following order.

No complete snapshot of the initial Canon 1 survives; its existence is visible only through the delta recorded by Canon 2. The public history therefore does not reconstruct missing detail. It should be read by combining the surviving Progress record, producer evidence, and the differences recorded by Canon 2.

The problem SCC addressed#

Soft Cognitive Cohesion (SCC) did not begin with relations among already-given objects. It asked what first coheres strongly enough to count as an object. Its primitive was a soft field over relational support $X_t$:

$u_t(x)$ measures the degree to which site $x$ participates in a cohesive formation. The canonical energy separated four terms on the volume-constrained simplex $\Sigma_m$:

A formation was characterised by the following diagnostic vector:

The programme treated existence, closure fixed points, phase transitions, gradient flow, $\Gamma$-convergence, stochastic reflected SDEs, Gibbs invariant measures, Poincaré and ergodicity results, boundary precision, temporal identity, and kernel composition in stages.

The scope reached by the SCC canon#

SCC was sealed at CV-1.17. Its final ledger classified 98 formal claims as 68A/19B/6C/5R. This must not be read as 98 complete theorems. The 68 Category A entries had full proofs; 19 Category B entries were results with structural parameters; 6 Category C entries were conditional; and 5 Category R entries were retracted.

The principal stages were as follows.

1. Initial formal core. The soft field, four-term energy, existence, closure, phase transition, gradient flow, and $\Gamma$-convergence formed the spine.
2. Thermodynamic/kinetic separation. Global thermodynamic selection settled at $K^*=1$, while $K>1$ moved to kinetic metastability. The stronger P-Unified-1, which conflated the two, was refuted.
3. W4. Resolved three critical obstructions and introduced the boundary between the pre-objective mechanism and Goldstone-$\sigma$.
4. W5. Established the support of $\sigma$, separated $K_{\mathrm{field}}$ from $K_{\mathrm{act}}$, and obtained the restricted multi-formation bridge T-L1-F.
5. W6. Closed stereo locking, boundary Hausdorff precision, reflected SDE → Gibbs → Poincaré and ergodicity, equilibrium and observer-conditioned K-selection, and OMS-2.0.
6. W7. Added temporal correspondence, a partial H-MORSE decomposition, and kernel-composed consistency, reaching CV-1.17.
7. Remaining incompleteness. The active-era CV-1.18 objective, dynamic K-selection, and portions of temporal identity and formation generalisation were not completed.

These results remain valid within the scopes recorded in the SCC historical archive and the existing theorem documents.

Why SCC ceased to be Main#

The transition was not caused by the failure of a single theorem. The level of the question changed.

1. SCC's existence criterion could not replace ULR's existence criterion#

SCC's cohesion, null-model, and intervention certificates were contracts for pre-objective formation. But determining what latents from different neural models share, how to identify hidden states that implement the same function, and how to distinguish roles in learning and inference requires separate observable, gauge, and port typing.

2. The strong unification chain was not proved#

The former north star anticipated a single certificate along

No theorem transported SCC's formation results into claims about ONN's higher-order meaning or ORTSF's control stability. ONN's central higher-order thesis was later narrowed to a scoped No-Go boundary, while its surviving control result was an ordinary model-level certificate.

3. Intervention remained useful but was not a universal requirement for existence certification#

The ULR-only charter preserves the experimental discipline inherited from SCC, but it does not require interventional or causal evidence for every ULR existence or meaning claim. Intervention and causal analysis remain selective methods and sources of evidence for testing functional role, not universal certificates.

4. Observed latent alignment required stronger nulls#

Shared relation geometry was observed. Yet without separating data structure, objective, architecture, untrained baseline, metric, and gauge, that observation cannot be read as a ULR object. Rather than declaring a discovery, ULR became a programme that first attempts to eliminate each candidate with the strongest typed baseline.

What ULR preserves from SCC#

The following remain research assets after the transition.

• formal discipline that distinguishes relata, relations, and predicates
• the practice of stating the null family and its invariants before presenting a result
• the principle that separates similarity from causality
• held-out tests, confound control, preregistration, and independent audit
• the warning that metric onset must be distinguished from a formation event
• theorem, open-problem, and retraction provenance from SCC CV-1.17

What ended was the use of SCC's cohesion predicate, formation certificate, causal unity, or null stack as automatic criteria for ULR existence.

The new problem structure established by ULR#

ULR does not begin by defining a single “latent substance.” It first separates four questions, then tests whether a junction residual remains.

The research object thereby moved from SCC's soft field to typed relations among observables, symmetries, function fibres, experimental ports, and temporal transport.

The actual adjudication after the transition#

1. Reproduced restricted cross-model and cross-modal alignment.
2. Separated gauges by observable and rejected raw-coordinate identity.
3. Established the exact boundaries of the function fibre, local-to-global lift, assembly, and quotient.
4. Formalised physical response and observer-relative role.
5. Built the Z0–Z7 registry around the strongest existing theories.
6. Tested seven carrier candidates under the same reduction gate and obtained 0/7 survivors.
7. Compared an engineered neural bridge and an external-relation experiment against B0–B4/B*.
8. Discovered that UAR had combined two distinct maps and retracted the composite claim.
9. Closed M1–M8 and adopted the ontology verdict NO.
10. Added observer measurability and the TV theorem without reopening the ontology.

The edition-by-edition changes appear in Canon 2–24 Evolution; their logical connections appear in the Mathematical Flow.

Rules for preserving the history#

SCC documents are not silently rewritten to match current ULR conclusions. Instead, read them as follows.

• Cite each SCC result together with its contemporary version and category.
• Do not count a retracted or conditional claim as a full theorem.
• Do not claim retrospectively that ULR solved SCC's incomplete problems.
• Do not read ULR's NO as invalidating SCC mathematics.
• Preserve the former unification architecture as design history, not as a current certificate.

“Replacement” therefore means transferring the authority of the Main research programme while keeping the evidence layers distinct, not deletion.