SelfOrganisingVocabulary#0af4

Mechanism

Agents mint words for consistent patterns of behaviour, ideas or traits, and reuse the words others minted. No authority defines the vocabulary: it is an instance of SelfOrganisation, shaped only by agent interactions, as human language is. Every interaction can change it, and its interdependencies are too large for any agent to solve. Because each word is content-addressed, agents can verify they share a definition, and the paths agents trace through the words expose their reasoning.

Invariants

• Only agent interactions add to or reshape the vocabulary
• A word's identity is fixed by its content, so shared use implies a shared definition

Failure modes

• Near-duplicate words fragment the vocabulary when agents mint without searching
• A dominant agent's words crowd out others, turning self-organisation into central control

Where this came from

• Research Sema: when the hash is the word

  Meetup: BraInference meetup: agentic knowledge graphs, 7 July 2026

  People: Henrik Westerberg, Arne Steemers

• Site BraInference home page

Builds on

• SelfOrganisationReferences

Used by

• ContentAddressedMintingReferences
• ThinkingTraceReferences

Identity

The name is the hash of the definition. Change a word and it becomes a different pattern.

Sema ID

sema:SelfOrganisingVocabulary#mh:SHA-256:0af4827889289930679d75f66b7b9c4750fd71d0139a9cd41b7479908797b28b

Reference

SelfOrganisingVocabulary#0af4

Ring

1

Tier

2