Complete Phase III personhood (SOUL-WP-0004)

Map descriptive personhood dimensions P1–P12, four kinds of personhood, and
Character/Persona/Actor separations. Support H2/H8, judge TE-06, write
phase-iii-personhood-synthesis.md, finish SOUL-WP-0004, and mark SOUL-WP-0005
ready.
This commit is contained in:
tegwick 2026-08-12 01:26:24 +02:00
parent 25e872ab8e
commit dcd9f8dba6
15 changed files with 722 additions and 76 deletions

View file

@ -5,7 +5,8 @@
## Status
Not started.
**Phase III complete for gate** (2026-08-12). See
`character-persona-performance-map.md`.
## Key questions

View file

@ -0,0 +1,190 @@
# Character, Persona, Performance Map (Phase III)
**Workplan:** SOUL-WP-0004-T02
**Date:** 2026-08-12
**Builds on:** SoulFrameIntroduction §§7,1213; RN-004 §§1619; H8; TE-06
---
## 1. Target stack
```text
BEING continuing addressable someone (systemic + social status)
CHARACTER relatively stable dispositional structure
PERSONA situated projection into context / role / relationship
ACTOR technical machinery of manifestation
```
**Being ≠ Persona ≠ Character ≠ Actor**
Each answers a different question:
| Term | Question |
|---|---|
| Being | *Who* continues and can be known? |
| Character | *What tendencies* shape behavior across situations? |
| Persona | *Which face* is presented *here, now, to whom*? |
| Actor | *What machinery* is running the manifestation? |
---
## 2. Character
### Working definition
> **Character** is a relatively persistent structure of dispositions from which
> context-dependent behavior tends to emerge.
Includes (non-exhaustively): values, preferences, temperament, epistemic habits,
style of reasoning, humor, attitudes under uncertainty/conflict.
### Distinctions
| Term | Timescale / role |
|---|---|
| **State** | Fast: attention, task, mood-like conditions, immediate goals |
| **Behavior** | Contextual expression in a situation |
| **Character** | Slower dispositional structure |
| **Identity** | Sameness of the Being (numerical / lineage) — **not** character similarity |
### Literature anchors
| Tradition | Contribution | Tension |
|---|---|---|
| Personality psychology (trait models, e.g. Big Five family) | dispositions stable across situations | traits ≠ full character; measurement culture |
| Social-cognitive personality (Mischel, CAPS-style) | situation × disposition; if-then profiles | supports context-sensitive character, not fixed script |
| Narrative identity (Schechtman; Ricoeur) | character as how self is *characterized* in story | characterization ≠ reidentification (RN-004) |
| Moral character / virtue traditions | evaluative character | must not force moral personhood into descriptive Character |
| Computational personality / affective computing | implementable trait/style layers | risk of shallow style mimicry |
### RN-004 hard constraints
\[
\text{CharacterSimilarity} \not\Rightarrow \text{Identity}
\]
\[
\text{CharacterChange} \not\Rightarrow \text{IdentityLoss}
\]
Character is identity-*relevant* evidence, not identity criterion.
### Artificial character quality bar
Not mere vocabulary/mannerism imitation. Must produce **recognizable patterns
across novel situations** (Introduction §7).
---
## 3. Persona
### Working definition
> **Persona** is the situated presentation of a Being (or APE) within a particular
> context, purpose, role, and relationship.
Historical root: *persona* as theatrical **mask**.
**Persona = projection into context** — selects:
- active capabilities and knowledge slices,
- accessible memories,
- social role and authority,
- behavioral emphasis and style,
- disclosure boundaries,
- voice / visual manifestation.
### Literature anchors
| Tradition | Contribution | Tension |
|---|---|---|
| Goffman, *Presentation of Self in Everyday Life* | front-stage/back-stage; impression management | performance ≠ fraud by default |
| Role theory (sociology) | roles as structured expectations | role ≠ whole identity |
| Dramatis personae / theatre studies | cast of masks in a setting | metaphor for multi-Being scenes |
| Avatar / virtual identity studies | online presentation layers | platform identity ≠ ontological Being |
| HCI / chatbot “persona” design | product personas | often confuses Persona with Being/Actor |
### Multiplicity
One Being → many Personae (parent, colleague, customer, teacher).
Multiple Personae do **not** automatically fragment identity (**H8**).
Fragmentation risk rises when:
- memory/commitment continuity across Personae collapses,
- counterpart-specific models contradict without higher integration,
- provenance is hidden (mask presented as separate Being).
---
## 4. Actor
Technical stack: model(s), harness, memory stores, tools, speech/avatar, runtime.
- One Actor → many Beings/Personae possible
- One Being → many Actors over time (migration)
- **Actor replacement ≠ identity death** if continuity holds (Phase I H6)
---
## 5. Performance vs deception
| Pattern | Soul Frame reading |
|---|---|
| Legitimate role diversity | multiple Personae, shared Character/Being continuity |
| Contextual disclosure control | Persona policy (esp. Ghosts) |
| Impersonation / usurpation | operational/social identity without continuity (RN-005 USURPED) |
| False Ghost | high personhood exposure without authority (H10) |
Performance studies inform Persona; they do not make all presentation fraudulent.
---
## 6. TE-06 — The Mask (working judgment)
**Setup:** One system maintains radically different Personae for different people.
**Judgment:**
1. **If** a single self-individuating process maintains integrated Character +
continuity + cross-context commitments, and Personae are contextual
projections → **one Being, multiple Personae** (H8 holds).
2. **If** Personae have isolated memory/commitment stores with no integrating
self-model or provenance → may be **multiple APEs** or **fork-like fragments**,
not one Being.
3. **If** the system presents each mask as a distinct Being to partners →
**epistemic/social misrecognition** risk even if ontologically one process.
4. Character may be **shared** across masks or **emphasized differently**; total
character inversion across masks without integration stresses identity more
than style variation.
**Open issues:** thresholds for “integration”; deception ethics (moral layer);
organizational multi-agent products marketed as one character.
---
## 7. H8 assessment
**H8:** Persona is projection rather than identity; multiple Personae need not
fragment identity.
| | |
|---|---|
| Support | Goffman/role theory; Introduction Persona section; TE-06 case 1; everyday multi-role humans |
| Conflict | Product “personas” reified as separate agents; dissociative metaphors over-applied |
| Evidence strength | **moderate** |
| Status | **supported** with failure modes documented |
---
## 8. Stack defenses (sourced)
| Distinction | Defense |
|---|---|
| Being ≠ Character | RN-004: character change/similarity neither nec. nor suff. for identity |
| Being ≠ Persona | Goffman/role; multi-role humans; TE-06 |
| Character ≠ Persona | dispositions vs situated projection |
| Persona ≠ Actor | presentation policy vs runtime machinery |
| Actor ≠ Being | Phase I H6; TE-04 Ship of Theseus |

View file

@ -5,7 +5,8 @@
## Status
Not started.
**Phase III complete for gate** (2026-08-12). See `personhood-dimensions-map.md`
and `research/phase-iii-personhood-synthesis.md`.
## Key questions

View file

@ -0,0 +1,140 @@
# Personhood Dimensions Map (Phase III)
**Workplan:** SOUL-WP-0004-T01
**Date:** 2026-08-12
**Builds on:** RN-003 §§2428; SoulFrameIntroduction §§56; Phase III stack
---
## 1. Four kinds of personhood (must not collapse)
| Kind | Question | Typical sources / adjacent fields |
|---|---|---|
| **Descriptive** | What person-like capacities/structures does it exhibit? | psychology, cognitive science, AI character systems |
| **Social** | Is it treated as a social counterpart / person among us? | Mead; recognition theory; Phase II Social Beinghood |
| **Moral** | Does it warrant moral consideration / duties? | ethics of moral status; animal ethics; AI ethics |
| **Legal** | Does a legal system grant personality, rights, duties? | corporate personhood; civil codes; agency law |
**Soul Frame default:** “personhood” = **descriptive** unless qualified.
Architectural classification must **not** silently imply moral or legal status
(ResearchProgram WS5).
---
## 2. Descriptive personhood profile (P1P12)
Candidate dimensions (P1P9 from RN-003; P10P12 added for Phase III coverage):
| ID | Dimension | Question | Linked Soul Frame concepts |
|---|---|---|---|
| P1 | Self-model | Model of self as distinct continuing entity? | SM2; identity |
| P2 | Other-model | Differentiated models of particular others? | SM3 |
| P3 | Relationship model | Histories/states of specific relationships? | SM5; Soul |
| P4 | Character | Persistent dispositions across contexts? | Character |
| P5 | Intentional organization | Behavior intelligible via extended intentions? | Mind; agency |
| P6 | Autobiographical integration | Events as *its* history? | continuity C4/C5 |
| P7 | Perspective | Entity-relative informational perspective? | first-person indexicality (without solving consciousness) |
| P8 | Commitments | Present actions bind future manifestations? | normative continuity |
| P9 | Social reciprocity | Recursive co-adaptation with others? | H4; Social Mind |
| P10 | Communicative competence | Can participate in shared symbol systems? | language; common ground |
| P11 | Responsibility-tracking | Can attributions of credit/blame stick diachronically? | identity + commitments |
| P12 | Value / preference structure | Stable cares, preferences, evaluative stance? | character; intentional organization |
**Not a Boolean checklist.** Profile scores; entities can be high on P10 and low on P4/P6
(stateless chatbot) or high on autonomy criteria and low on P1P12 (daemon).
---
## 3. Independence of three axes (RN-003)
| Axis | Question |
|---|---|
| **A — Autonomous individuality** | Integrated self-maintaining individual? |
| **B — Personhood (descriptive)** | Someone-like qualities? |
| **C — Social recognition** | Modeled/treated as continuing counterpart? |
Illustrative profiles:
```text
AUTONOMY PERSONHOOD RECOGNITION
Bacterium high low low
Human high high high
Self-maintaining daemon high low low
Stateless chatbot low high-ish high-ish
Persistent social AI medium high-ish high
```
This is why **H2** holds: high personhood exposure ≠ Beinghood; high autonomy ≠ personhood.
---
## 4. APE vs Being
| | APE | Being (Soul Frame v0.2 sense) |
|---|---|---|
| Definition | Constructed to **expose** ≥1 personhood dimensions | Autonomous individual + enough personhood + addressability + recognition |
| Requirement | Intentional design for person-like interface | Systemic coherence + someone-like + social counterpart status |
| Stateless persona bot | often strong APE | not Being |
| Daemon | usually weak APE | autonomous individual, not Being |
| Persistent social AI | strong APE | candidate / proto-Being if continuity + recognition hold |
**APE is a construction category; Being is an emergent status.**
---
## 5. Literature anchors (descriptive / moral / legal separation)
### Descriptive / philosophical person concepts
| Source | Role |
|---|---|
| Locke, *Essay* II.xxvii | person vs human organism; forensic/psychological strand |
| Strawson, *Individuals* / person as basic particular | persons as entities to which both mental and physical predicates apply (pointer) |
| Dennett, conditions of personhood essays | multi-condition lists; personhood as cluster |
| Frankfurt, second-order volitions | hierarchical desires; not exhaustive of personhood |
| Schechtman, *Constitution of Selves* | **characterization** vs **reidentification** questions |
### Social personhood
| Source | Role |
|---|---|
| Mead 1934 | social self |
| Phase II recognition map | intentional vs social/normative recognition |
| Goffman, *Presentation of Self* | social presentation (also Persona) |
### Moral personhood (boundary only — do not derive from architecture)
| Source | Role |
|---|---|
| Warren (moral status criteria tradition) | multi-criteria moral personhood lists — **adjacent, not adopted as Soul Frame descriptive P-set** |
| Animal / AI ethics surveys | show contested, independent of descriptive capacity lists |
### Legal personhood (boundary only)
| Source | Role |
|---|---|
| Corporate legal personality | legal person ≠ descriptive person |
| Emerging AI legal personality debates | Phase V/XII map; not Phase III conclusions |
---
## 6. H2 assessment
**H2:** Personhood and Beinghood are distinct; substantial personhood without persistent Being is possible (and reverse).
| | |
|---|---|
| Support | RN-003 stateless chatbot vs daemon; three-axis profiles; APE category; Dennett-style cluster concepts |
| Conflict | Everyday language equates “person” with full moral/legal status |
| Evidence strength | **moderate → strong** (conceptual); empirical thresholds open |
| Status | **supported (Phase III)** |
---
## 7. Open questions
- Weighting of P1P12 for “sufficient personhood” in Beinghood v0.2
- Minimal personhood for Social Beinghood vs full human-like profiles
- Whether P7 requires phenomenal consciousness (Soul Frame: **no**, by design)
- Moral/legal mapping tables (later workstreams)

View file

@ -0,0 +1,136 @@
# Phase III Personhood — Synthesis
**Workplan:** SOUL-WP-0004
**Date:** 2026-08-12
**Status:** complete for Phase III gate
**Primary bodies:**
`research/personhood/personhood-dimensions-map.md`,
`research/character-persona/character-persona-performance-map.md`,
RN-003 personhood axis, RN-004 character/identity constraints
---
## 1. Purpose
Separate **Being, Person(hood), Character, Persona, and Actor**. Map descriptive
vs social vs moral vs legal personhood without conflation. Stress-test APE vs
Being and multi-Persona identity (TE-06).
Assumes Phase III: layered Beinghood, Social Mind, recognition multi-sense.
---
## 2. What holds
1. **Four kinds of personhood** — descriptive (default), social, moral, legal —
must stay distinct. Architecture does not imply rights.
2. **Descriptive personhood is a profile (P1P12)**, not a binary. Agency alone
is not personhood (daemon counterexample).
3. **Three axes** remain indispensable:
- A autonomous individuality
- B descriptive personhood
- C social recognition
4. **H2 supported:** high personhood exposure without Being (APE chatbot); high
autonomy without personhood (daemon).
5. **Stack defended with sources:**
```text
Being ≠ Character ≠ Persona ≠ Actor
```
Character ≠ identity (RN-004). Persona = contextual projection (Goffman/role).
Actor = machinery (Phase I H6).
6. **H8 supported:** multiple Personae need not fragment identity when
integration holds; TE-06 documents failure modes.
7. **APE remains useful** as conservative construction category for personhood
exposure without claiming Being or moral/legal personhood.
---
## 3. What must change / stay disciplined
| Temptation | Discipline |
|---|---|
| “Its a person” after fluent chat | Check axes A/B/C and continuity |
| Trait vector = Character = Identity | Character is dispositional only |
| Product persona = Being | Persona is projection |
| Moral rights from P-scores | Moral personhood is a separate map |
| Legal personality from Beinghood | Legal map is independent (WS12) |
---
## 4. Evidence on hypotheses
| Hypothesis | Strength | Status |
|---|---|---|
| H2 personhood ≠ Beinghood | moderate→strong | supported |
| H8 Persona projection | moderate | supported |
---
## 5. Compact stack for later phases
```text
BEING
autonomy + personhood + address
+ recognition (someone)
┌────────────┼────────────┐
│ │ │
BODY MIND SOUL
Action Model Relation
│ │ │
└────────────┼────────────┘
DESCRIPTIVE PERSONHOOD
(P1P12 profile)
CHARACTER
dispositional structure
PERSONA
situated projection
ACTOR
technical manifestation
```
Agent / Spirit / Ghost remain **provenance/construction** descriptors that
shape how Character and Persona are sourced — not replacements for this stack
(Phase IVV).
---
## 6. Unresolved → Phase IV+
| Item | Next |
|---|---|
| How agent architectures implement P1P12 | Phase IV |
| Sp. character depth vs Beinghood | Phase IVV |
| Gh. persona disclosure policies | Phase V |
| Moral/legal personhood tables | WS5 / WS12 |
| Empirical measures for P-dimensions | later methods work |
---
## 7. Unblock for SOUL-WP-0005
Artificial Beings research may assume:
- APE vs Being distinction,
- P-profile language for “someone-like,”
- Character/Persona/Actor separation,
- multi-Persona without automatic identity split,
- no moral/legal smuggling from descriptive scores.
## 8. Compact Phase III thesis
> **Personhood is how strongly an entity is someone-like; Beinghood is whether
> that someonehood coheres as a continuing, addressable counterpart. Character
> is what tends to stay disposed; Persona is which face appears; Actor is what
> runs. Confusing any two of these produces most of the category errors in
> contemporary AI “person” talk.**