# 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.