Exploration: The Methodology Landscape, Pulled Up Under the Convergence Grounding

Status: The methodology arrangement brought to the same maturity as entity/biology: full lens stack run on N=21 methodologies across the 4 layers, read through the now-registered Convergence Domain. This is the "methodology output" — what the methodology, analyzed as a methodology, tells us about itself, and the feedback that lets the other analyses be done more accurately. Builds on: methodology-landscape.md (the prior N=13 qualitative pass), data/domains/convergence.v1.json, exploration-convergence-grounds-the-higher-layer-practice.md.

1. Framing

The methodology arrangement positions 21 analytical methods in the methodology's own L1–L4 lattice (Domain-Analysis → Graph-Construction → Graph-Semantics → Applied-Analysis). It is not a substrate→ecosystem chain — it is the landscape of methods that produce landscapes. Grounded by the Convergence Domain: a cluster here is a Determination (Dt); a method's membership shift between layers is a Collapse (Cl) (it crosses a coherence threshold at one layer but not another); the corpus space is a finite Sp2. Claims are bounded at Ds2 (classical — descriptive cartography, no interference/over-reach), per the licensed-claims bound.

2. The landscape (lens stack, N=21, hierarchical + stability, all 4 layers)

Silhouettes rise through the stack (L1 0.56 → L4 0.68): the structure sharpens at the applied layer. One attractor recurs at every layer, and the cross-method stability run confirms it as the robust core:

• The formal-inquiry attractor (the dominant Dt). Robust core across all layers: scientific-method, mathematical-proof, statistical-inference, analytical-philosophy, phenomenology, engineering-design (design-thinking / ethnography join at most layers) — and structural-primitive-analysis sits inside it at every single layer.
• The strategic-analysis frameworks fragment into the prior note's three thinner sub-attractors: snapshot-decomposition (SWOT, PESTEL, Cynefin, Lean Canvas, Porter), iterative-cycle (OODA, OKRs, DMAIC, TOC), trajectory-mapping (Wardley, Scenario-Planning, Balanced-Scorecard).
• PESTEL is a recurring singleton (its own cluster at L1, L3, L4) — a stable Dt1 boundary attractor at the minimal, Cx-only corner.

3. Why the landscape has this shape — the convergence grounding does real explanatory work

The cluster structure is explained, not just described, by which Convergence-Domain primitives each method instantiates:

This is the grounding revealing structure, not imposing it: the methodology landscape's attractors are sorted by how much of the convergence process the method runs. Snapshot methods stop at the Landscape Triad; only formal-inquiry methods reach Collapse + Determination.

4. The self-referential payoff — what this tells us about our methodology

structural-primitive-analysis clusters with scientific-method, mathematical-proof, statistical-inference at all four layers — never with the strategic frameworks. Therefore:

1. Our method's structural peer group is formal science, not strategy consulting. It is a full-convergence inquiry method. This validates the analyst-cartography stance: what we do is structured inquiry that converges to determinations (Dt), in the Ds2-classical regime — exactly the discipline we converged on conversationally, now confirmed by the method's own position in its own landscape.
2. Paper 11's comparison set is settled: the honest peers are scientific-method / mathematical-proof / statistical-inference; the closest strategic neighbor (for the "vs other frameworks" figure) remains Wardley. The four under-served roles from the N=13 note (pair-load classification, filter stringency, core triads, cross-arrangement coupling) still hold and are now positioned as "what a full-convergence method adds over Landscape-Triad-only methods."
3. The recursion has its sharpest form here: the method, analyzed by itself, lands among the methods that converge to determinations — i.e. the methodology is itself an instance of the Convergence Domain it discovered, and its own landscape confirms it.

5. The feedback loop (why this makes the other analyses more accurate)

Because our method is a full-convergence formal-inquiry method, the discipline for the entity/biology analyses is now derived from the methodology landscape itself, not asserted:

• We are licensed to claim Dt-determinations (anchors) and Cl-events (phase transitions) — that is what full-convergence methods do.
• We are not licensed to leave Ds2 (classical): no interference / emergent-magic over-reach.
• A clustering scalar alone is Landscape-Triad output (no Cl) — by construction not a finding. A finding requires the Information-Gain triad {Ds, Dy, Cl}. This is the exact "confirm/deny on a 0.5 is wrong" rule, now re-derived from where our method sits among methods that don't converge.

So the loop closes: analyzing the methodology as a methodology tells us it is a full-convergence inquiry method, which tells us precisely what the entity/biology analyses may and may not claim — and that is the corrected discipline, independently re-grounded.

6. Bounded follow-ups (flagged, not chased)

• Meta-stability not run. cluster-meta-stability ARR=methodology failed: it requires stability for pair + full signatures (only raw was run) and it referenced entity's application-architecture level — i.e. it appears not arrangement-aware for non-entity arrangements. Precise tooling item; the per-layer cluster+stability (raw) is sufficient for the result above. Do not chase into the weeds.
• The N=13 note's open follow-ups (Sc=2 weighted by usage, Sc=3 trajectory 1960s→2026) remain valid and are now convergence-framed (a Sc=3 trajectory would be a Dy/Cl history of the methodology landscape).