The Identity-Memory-Accountability Chain

This is the conceptual spine of Web4's governance model. Before introducing the specific primitives, understand the chain that connects them:

1. Roles exist in context

A role is a predefined function within a specific context. The role is not the agent — it's the job description the agent fills. Multiple agents can fill the same role. One agent can fill multiple roles. Your reputation as an engineer is distinct from your reputation as a manager, even if you're the same person.

2. Identity accumulates to role

As an agent operates in a role, its actions build an identity specific to that role. This is not global identity — it's contextual identity. The same agent filling two different roles accumulates two distinct role-identities. This matches how human reputation actually works: your standing in one community says nothing about your standing in another.

3. Memory is the root of identity

Identity is not declared. It is accumulated — the sum of observed behavior over time. Memory is what makes that accumulation persistent and referenceable. Without memory, every interaction is zero-context. No trust can accumulate. No accountability persists.

This parallels the immune system: immune memory is what allows the system to build lasting trust and distrust relationships with the entities it encounters.

4. Localizing identity enables accountability

Accountability can only attach to a localized, persistent identity. If identity is diffuse or transient, consequence has nowhere to land.

Localizing identity means: this specific agent, in this specific role, in this specific context, with this accumulated history — THAT is what's accountable. This is MRHapplied to identity — scoping accountability at the level where coherent agency and persistent memory converge.