Complete Phase II interaction (SOUL-WP-0003)

Map model-based mind, co-adaptive social cognition, and recognition/social
Beinghood into instruments and workstream notes. Support H3/H4/H7, judge TE-09,
write phase-ii-interaction-synthesis.md, finish SOUL-WP-0003, and mark
SOUL-WP-0004 ready.

research/social-cognition/README.md

@@ -5,7 +5,8 @@
 
 ## Status
 
-Not started.
+**Phase II complete for gate** (2026-08-12). See `co-adaptive-model-building.md`
+and foundations RN-006.
 
 ## Key questions
 

research/social-cognition/co-adaptive-model-building.md

@@ -0,0 +1,96 @@
+# Co-Adaptive Model Building — Social Cognition Map (Phase II)
+
+**Workplan:** SOUL-WP-0003-T02  
+**Date:** 2026-08-12  
+**Primary research body:** `research/foundations/Research Note 006 — Social Mind and Co-Adaptive Model Building.md`
+
+---
+
+## 1. Working definition (RN-006)
+
+> **Co-Adaptive Model Building** is the recurrent process through which two or
+> more Beings update action-guiding models of themselves, one another, their
+> relationships, and their shared situation—potentially including bounded models
+> of one another's models—while actions generated from those models alter the
+> interaction process and thereby create the evidence for subsequent updates.
+
+**Not:** mere exchange or synchronization of identical models.  
+**Not:** private Theory of Mind alone.
+
+---
+
+## 2. Tradition map
+
+| Tradition | Key sources | Soul Frame uptake |
+|---|---|---|
+| Theory of Mind | Premack & Woodruff 1978 | Other-Model (SM3) |
+| Inverse planning | Baker, Saxe & Tenenbaum 2009; Shum et al. 2019 | Computational other-modeling |
+| Symbolic interaction | Mead 1934 | Reflected-Self; Generalized Social Model |
+| Dialogue / alignment | Pickering & Garrod 2004 | Interactive co-adaptation mechanisms |
+| Common ground | Clark & Wilkes-Gibbs 1986; Clark & Brennan 1991 | SM7 Common-Ground Model |
+| Shared intentionality | Tomasello et al. 2005; Bratman 1992 | Strong form of cooperation — **not** definition of Social Mind |
+| Joint action | Sebanz et al. 2006 | Coordination without full mental-state identity |
+| Participatory sense-making | De Jaegher & Di Paolo 2007, 2010 | Interaction process as constitutive |
+| Second-person neuroscience | Schilbach et al. 2013 | Interaction-centered social cognition |
+| Recursive multi-agent | Gmytrasiewicz & Doshi 2005 (I-POMDP) | Formal recursion bounds |
+| Relational schemas | Fiske 1992 | Relationship Model (SM5) types |
+
+---
+
+## 3. Social Mind model families (SM1–SM8)
+
+| ID | Model | Question |
+|---|---|---|
+| SM1 | World | What is the shared environment? |
+| SM2 | Self | What am I? |
+| SM3 | Other | What is B? |
+| SM4 | Reflected-Self | What does B model about me? |
+| SM5 | Relationship | What is A↔B? |
+| SM6 | Shared Situation | What is happening between us now? |
+| SM7 | Common Ground | What do we sufficiently share? |
+| SM8 | Generalized Social | Roles, norms, institutions, community |
+
+A Social Mind is a **coupled system** of these, not a single “user model.”
+
+---
+
+## 4. Distinctions that must not collapse
+
+| Concept | Relation to co-adaptation |
+|---|---|
+| Model equality | Same model contents — stronger than needed |
+| Model alignment | Compatible enough for coordination |
+| Common ground | Mutual acceptance for current purposes |
+| Shared intentionality | Joint goals/commitments — special case |
+| Recognition | Standing as counterpart — may begin before deep co-adaptation |
+
+Agreement and cooperation are **possible outcomes**, not definitional requirements
+of co-adaptive model building (conflict and deception are also co-adaptive).
+
+---
+
+## 5. H4 assessment
+
+**H4:** Persistent interaction produces recursively adapting models of self, other,
+and relationship.
+
+| | |
+|---|---|
+| Support | Mead; alignment; common ground; participatory sense-making; inverse planning; I-POMDP recursion; RN-006 loop |
+| Conflict | Purely individualistic ToM that treats interaction as mere input; strong “no representation” enactivism if “model” is read narrowly |
+| Evidence strength | **moderate → strong** for the co-adaptation thesis as synthesis |
+| Status | **supported** (Phase II) |
+
+Refinement: co-adaptation is **joint evolution through interaction**, and models
+can be self-confirming or self-defeating because they change the evidence stream.
+
+---
+
+## 6. TE implications
+
+- **TE-01 Stateless Agent:** partner may co-adapt *as if* continuous; co-adaptation
+  on one side does not create diachronic Beinghood on the other.
+- **TE-06 The Mask:** multiple Personae require differentiated Relationship and
+  Reflected-Self models per counterpart.
+- **TE-09 Unrecognized Intelligence:** co-adaptation may not start until intentional
+  recognition begins; systemic individuality can precede social modeling.