soul-frame/history/260928-scope-intent-gap-assessment.md

Scope vs. Intent — Gap Assessment (2026-09-28)

Context: SCOPE.md updated the same day to reflect actual repo state (Phases I–IX complete, SOUL-WP-0001…0010 finished). This note assesses that state against INTENT.md's standing purpose, six determinations, and success criterion, and ranks the gaps.

---

1. Assessment against INTENT's six determinations

#	INTENT: "determine…"	State	Verdict
1	what is already well established	13 workstream literature maps, SF-R### ledger, five phase syntheses	Met for the traversed fields; fiction set and some WS depth (WS11, WS13) remain thin
2	what Soul Frame recombines usefully	research/concept-crosswalk.md, related-work §, phase syntheses	Met
3	what concepts require refinement	H1–H10 all phase-reviewed; RQ1–RQ8 carry working answers with explicit residuals	Met as process; residuals live in OpenQuestions
4	where terminology conflicts with usage	Terminology standard; RQ4 Soul retention decided with disclaimer policy; Deppert false-friend flagged	Met
5	what genuinely new propositions emerge	B1–B8 / P1–P12 profiles, three-axis independence, Sp./Gh. as provenance descriptors, recognition test, continuity channel set	Met, but propositions are profile-shaped with no aggregation/threshold principle (→ G1)
6a	can these support a rigorous theory	Paper draft exists, adversarially reviewed	Partially met — rigor currently rests on per-dimension analysis; sufficiency questions ("how much personhood is enough for Beinghood v0.2") are open everywhere they matter (→ G1)
6b	…and eventually a practical architecture	Deliberately downstream (INTENT itself: "downstream of conceptual clarity")	Out of repo scope — not a gap to close here

2. Assessment against the success criterion

INTENT's success criterion: one coherent vocabulary describing nine reference entity types (human, animal, short-lived AI agent, persistent autonomous Agent-being, Einstein Sp., authorized Gh., original artificial companion, distributed AI, radically non-human system) while still making meaningful distinctions among them.

Finding: this test has never been run as such. The closest artifacts cover fragments — personhood-dimensions-map.md §3 profiles five entities on three coarse axes; thought experiments each stress one entity class. No document applies the full vocabulary (B1–B8, P1–P12, continuity channels, addressability/recognition, Agent/Sp./Gh. descriptors) across all nine and checks that the distinctions survive. The program's own success criterion is currently asserted, not exercised (→ G2).

3. Gap register

ID	Gap	Source	Addressable now?
G1	No principled weighting/sufficiency scheme for the B-, P-, and continuity profiles; "sufficient personhood," "enough closure," and continuity thresholds are undefined wherever they are load-bearing	RQ1, RQ2, RQ5 (OQ-043); paper §17 item 2	Yes — conceptual work, no dependency
G2	Success-criterion coverage matrix (nine entity types × full vocabulary) never produced	INTENT success criterion	Yes — synthesis of existing material
G3	Soul packaging unresolved: relations vs consequences vs both; Cluster D formal depth	RQ4 remainder, flagged paper-critical	Yes — bounded concept work
G4	Metrics instruments scoped but not built (B3, psychological continuity)	WP-0010 residual	Gated — open decision 0401cf28; do not start without resolution
G5	Companion line items 2–4 (moral status survey, non-derivation statement, jurisdictional tracking)	Charter	Gated — charter's own next-step rule; slice-by-slice
G6	Paper not yet circulated; internal-review stage depends on coulomb-research being operational	WP-0009-T06 decision	Blocked externally — facility still being set up
G7	Thin coverage: WS11 (death/persistence depth), WS13 fiction set breadth, hybrid display conventions, eval harnesses	§17 item 6; RQ7/RQ8 residuals	Yes, but low urgency — none is load-bearing for the paper

4. Ranking — most relevant gaps

G1, G2, G3, in that order, are the gaps worth addressing now:

• G1 is the deepest structural debt: every profile the framework introduces ends in "weighting and thresholds open." Determination 6a cannot be claimed while the framework has no stated position on aggregation — even a principled refusal to weight (context-of-claim relativity) would be a position; silence is not.
• G2 is the cheapest high-value item: it directly exercises the success criterion, likely surfaces vocabulary strain exactly where INTENT predicts it (distributed AI, radically non-human), and produces the paper's strongest possible exhibit.
• G3 is the only remaining item OpenQuestions marks paper-critical.
• G4–G6 are gated or blocked by design; G7 is real but not urgent.

Disposition: G1–G3 are picked up by SOUL-WP-0011 (created with this assessment). G4–G6 stay with their existing gates. G7 remains in OpenQuestions/§17 as agenda, not debt.