soul-frame/research/metrics-scoping-note.md

# Empirical/Operational Metrics Scoping Note

**Workplan:** SOUL-WP-0009-T04, resolved 2026-09-28
**Purpose:** scope — not build — measurable proxies for the A-axis (B1–B8,
autonomous individuality), the B-axis (P1–P12, descriptive personhood), and
the continuity profile. Soul Frame is currently qualitative/classificatory;
this note identifies which axes are plausibly measurable now with existing
instruments vs. which need new ones, so a future workplan can pick this up
without re-deriving the landscape.

## Method

For each dimension: does an existing measurement tradition already target
something close to it, or would Soul Frame need to invent an instrument?

## A-axis (autonomous individuality, B1–B8)

| Code | Dimension | Plausibly measurable now | Rationale |
|---|---|---|---|
| B1 | Individuability | Partial | Individuality/information-theoretic measures (Krakauer et al. 2020, SF-R019–034 cluster) give a graded formal proxy; applying them outside biological systems is untested. |
| B2 | Temporal integrity | Partial | Uptime/session-continuity logs give a crude proxy for artificial systems; no validated cross-substrate metric. |
| B3 | Constitutive integration | **No** | No existing instrument distinguishes "integrated organization" from "loosely coupled components" quantitatively; closure-of-constraints formalizations (Montévil & Mossio 2015) are the closest theoretical basis but are not operationalized as a measurement protocol. |
| B4 | Self-maintenance participation | Partial | For artificial systems, resource/error-correction telemetry (self-healing, retry, recovery behavior) is a plausible proxy already collected by most production systems — needs mapping, not invention. |
| B5 | Endogenous viability | **No** | Requires distinguishing internally-driven from externally-propped persistence; heteronomous-individual cases (Terminology.md) are exactly the hard cases this would need to separate, and no metric currently does. |
| B6 | Adaptive regulation | Partial | Control-theoretic and RL literature already measures regulation quality (error minimization, model accuracy) — directly reusable, though not originally framed as a Beinghood metric. |
| B7 | Interactional asymmetry | **No** | Novel to Soul Frame's framing; no existing instrument. |
| B8 | Identity-preserving transformation | Partial | Overlaps with the continuity profile (see below) — measurable to the extent continuity is measurable. |

## B-axis (descriptive personhood, P1–P12)

Not itemized in current concept files beyond the "profile, not checklist"
framing (`concepts/personhood.md`, `Terminology.md`). **Scoping finding:**
before metrics can be scoped per P-code, P1–P12 need to be enumerated with
the same specificity B1–B8 already have — this is a prerequisite gap, not
a metrics gap. Recommend a follow-up task (not opened here) to write out
P1–P12 explicitly in `concepts/personhood.md` before attempting measurement.

## Continuity profile

| Dimension | Plausibly measurable now | Rationale |
|---|---|---|
| Causal | Yes | Provenance chains, cryptographic attestation, audit logs — mature tooling exists for artificial systems. |
| Organizational | Partial | Architecture/component-graph diffing can detect discontinuities; no agreed threshold for "still one organization." |
| Psychological | **No** | No validated metric for "the same adaptive/modeling process," as opposed to similar outputs. |
| Autobiographical/narrative | Partial | Memory-store diffing and narrative-consistency NLP metrics exist but measure surface consistency, not the underlying continuity claim. |
| Normative | **No** | No existing instrument for "the same locus of commitments/obligations" independent of legal/contractual proxies (which measure something adjacent, not identical). |
| Relational | Partial | Social-graph and reputation-system metrics are mature for *tracking* relationships; whether they validly evidence *identity* continuity (as opposed to just relationship persistence) is unresolved. |

## Overall scoping conclusion

- **Most measurable now:** causal continuity, adaptive regulation (B6),
  self-maintenance telemetry (B4) — existing instruments, direct reuse.
- **Needs new instruments:** constitutive integration (B3), endogenous
  viability (B5), interactional asymmetry (B7), psychological continuity,
  normative continuity.
- **Blocked on prior work:** B-axis (P1–P12) metrics cannot be scoped until
  the P-codes are enumerated with B1–B8's level of specificity.

**Recommendation:** a future workplan (not opened as part of SOUL-WP-0009)
should (1) enumerate P1–P12 explicitly, then (2) prioritize instrument
development for constitutive integration and psychological continuity —
these are cited most often across the thought-experiment catalog (TE-02,
TE-04, TE-05, TE-08) and are the two dimensions where "No" above is doing
the most argumentative work in the framework.