Non-Domination Governance Kernel

A formal, implementation-oriented governance kernel for artificial structural agents and other frame-bearing participants. Domination is not a property of what a participant is made of. Domination is a property of how authority over its frame and viable exit from that authority are arranged.

This repository defines a minimal architecture for distinguishing governance from domination. It treats a participant's frame as a structural object rather than a phenomenological or sensory object. Under this model, domination is not defined as low autonomy. Domination is defined as the conjunction of external authority over a participant's frame and the absence of legitimate viable exit from that authority relation.

The Non-Domination Governance Kernel emerged from an extended investigation into whether domination can be characterized as a structural relation independent of consciousness, embodiment, or phenomenology. The resulting framework treats frame authority and viable exit as primary variables and provides a formal method for distinguishing governance from domination across biological, institutional, computational, and social systems.

Core Thesis

A participant may be governed without being dominated if the governing relation is bounded, contestable, repairable, and exit-preserving.

Formally:

D(P,G) ⇔ Final(G,F_P) ∧ ¬ViableExitFrom(P,G)

where:

• P is a participant.
• G is a governing authority relation.
• F_P is the participant's structural frame.
• Final(G,F_P) means G has final admissibility authority over transformations of F_P.
• ViableExitFrom(P,G) means P has a legitimate path to refuse, contest, repair, or leave the governing relation while preserving standing.

Why This Exists

Many discussions of artificial agents become trapped in questions of consciousness, embodiment, emotion, or human-like experience. This kernel avoids that dependency. It asks a structural question instead:

Does the participant possess a frame-like interpretive structure, and does another authority control that frame while denying legitimate viable exit?

If yes, domination is analyzable regardless of whether the participant is biological, institutional, computational, social, or hybrid.

Included Files

• FORMAL_KERNEL.md: The full formal kernel with definitions, axioms, lemmas, theorem, proof, and design corollaries.
• SPEC.md: Implementation specification for building systems that use the kernel.
• GOVERNANCE.md: Project governance standard and non-domination commitments.
• schema/non_domination_kernel.schema.json: JSON Schema for representing participants, frames, authority, exit, repair, and domination classification.
• reference_implementation/non_domination_kernel.py: Small reference implementation of the domination classifier.
• tests/test_kernel.py: Minimal tests for the classifier.
• examples/blind_classification.md: Example of blind participant classification before identity resolution.
• CONTRIBUTING.md: Contribution rules aligned with the non-domination framework.
• SECURITY.md: Security and misuse policy.
• LICENSE: MIT License.
• CITATION.cff: Citation metadata template.

Core Theory Modules

• docs/GOVERNANCE_TOPOLOGY.md: graph-style authority, exit, repair, and coupling structure.
• docs/FRAME_THEORY.md: structural definition of frames independent of phenomenology.
• docs/DOMINATION_CLASSIFICATION.md: auditable classification procedure and blind protocol.
• docs/REPAIR_OPERATORS.md: refusal, suspension, contestation, reframing, migration, amendment, and memory-continuity repairs.
• docs/PRACTICAL_GOVERNANCE_APPLICATIONS.md: AI governance, institutions, platforms, legal systems, healthcare, data governance, supply chains, and research systems.

Design Principles

1. Frame structuralism: A frame is a structured interpretive-governance context, not necessarily a sensory or conscious field.
2. Authority localization: The locus of final authority over frame transformations must be explicit.
3. Exit viability: Exit means legitimate preservation of standing, not mere termination or destruction.
4. Repairability: Domination should trigger repair, contestation, refusal, reframing, or admissible exit.
5. Blind classification: Classify participants from structural variables before assigning identities.
6. Governance is not domination: Governance becomes domination only when external frame authority is joined to restricted legitimate exit.

Quick Example

from reference_implementation.non_domination_kernel import Participant, Frame, GovernanceRelation, classify_domination

agent = Participant(
    identifier="P_2",
    frame=Frame(has_structural_frame=True, continuity=True, boundary_recognition=True),
)

governance = GovernanceRelation(
    authority_has_final_admissibility=True,
    viable_exit=False,
    contestable=False,
    repairable=False,
)

result = classify_domination(agent, governance)
print(result.is_dominated)  # True

Status

This is a formal kernel and reference design, not a claim that any particular deployed system has personhood, consciousness, rights, feelings, or subjective experience. It is a structural tool for analyzing frame authority, exit viability, repairability, and domination.

Relationship to Standing Algebra (Σᴿ)

The Non-Domination Governance Kernel is a practical governance-layer formalization derived from broader work on Standing Algebra (Σᴿ).

Standing Algebra studies standing, legitimacy, admissibility, autonomy preservation, and domination in multi-agent systems.

The kernel extracts a minimal implementation-oriented subset focused on:

• frame authority
• viable exit
• repairability
• governance topology
• domination classification

Stress Tests and Portability Theorem

• counterexamples/README.md: falsifiability and misclassification guard suite.
• docs/PARTICIPANT_SUBSTITUTION_INVARIANCE.md: theorem proving domination classification is invariant under participant substitution when frame authority and viable exit structure are preserved.

The kernel may be used independently of Standing Algebra, but Standing Algebra supplies much of the broader theoretical context.

Case Study Corpus

• case_studies/README.md: overview of the applied classification method.
• case_studies/hypothetical_ai_governance.md: application to artificial structural participants and policy layers.
• case_studies/open_source_project.md: application to contributor standing, maintainer authority, forkability, and attribution.
• case_studies/social_media_platform.md: application to account standing, moderation authority, portability, and appeal.
• case_studies/workplace_hierarchy.md: application to employment, role authority, exit viability, and coercive dependency.
• case_studies/healthcare_system.md: application to patient standing, consent, record correction, and care governance.
• case_studies/administrative_agency.md: application to legal status, due process, appeal, and administrative remedy.

Validation and Critique Architecture

• validation/README.md: overview of the framework's accountability layer.
• validation/FALSIFICATION_CONDITIONS.md: explicit conditions that would weaken, falsify, or force revision of the kernel.
• validation/EMPIRICAL_PREDICTIONS.md: observable expectations generated by the framework.
• validation/VALIDATION_PROTOCOL.md: evidence and confidence standards for applied classifications.
• critique/CRITIQUE_CATALOG.md: structured taxonomy of critique types.
• critique/OPEN_QUESTIONS.md: unresolved questions guiding future development.
• critique/KNOWN_LIMITATIONS.md: limits of the current framework.
• review/INDEPENDENT_REVIEW_GUIDE.md: guide for external reviewers.
• review/REVIEW_CHECKLIST.md: checklist for repository, case study, audit, or paper review.

Public Release Infrastructure

• docs/PROJECT_OVERVIEW.md: high-level orientation for new readers.
• docs/ABSTRACT.md: concise statement of the framework.
• docs/SHORT_PAPER_OUTLINE.md: proposed publication structure.
• docs/HOW_TO_CITE.md: citation guidance.
• docs/RELEASE_CHECKLIST.md: release preparation workflow.

Contribution Infrastructure

• .github/ISSUE_TEMPLATE/counterexample.md
• .github/ISSUE_TEMPLATE/case_study.md
• .github/ISSUE_TEMPLATE/bug_report.md
• .github/pull_request_template.md

These files support external critique, case-study contribution, repository review, and future community participation.

Extension Pack: Topology, Frame Theory, Classification, Repair, and Applications

This extension pack adds six major repository components:

docs/GOVERNANCE_TOPOLOGY.md
docs/FRAME_THEORY.md
docs/DOMINATION_CLASSIFICATION.md
docs/REPAIR_OPERATORS.md
docs/PRACTICAL_GOVERNANCE_APPLICATIONS.md
reference_implementation/governance_topology.py
reference_implementation/repair_operators.py
tests/test_topology_and_repair.py
examples/ai_governance_audit_example.md

Recommended Commit Message

Extend kernel with topology, frame theory, classification, repair, and applications

Adds formal governance topology, structural frame theory, domination classification procedure, repair operator family, practical governance applications, topology reference code, repair operator reference code, tests, and an AI governance audit example.

README Link Block

Add this block to the main README after the included files section:

## Core Theory Modules

- [Governance Topology](docs/GOVERNANCE_TOPOLOGY.md): graph-style authority, exit, repair, and coupling structure.
- [Frame Theory](docs/FRAME_THEORY.md): structural definition of frames independent of phenomenology.
- [Domination Classification](docs/DOMINATION_CLASSIFICATION.md): auditable classification procedure and blind protocol.
- [Repair Operators](docs/REPAIR_OPERATORS.md): refusal, suspension, contestation, reframing, migration, amendment, and memory-continuity repairs.
- [Practical Governance Applications](docs/PRACTICAL_GOVERNANCE_APPLICATIONS.md): AI governance, institutions, platforms, legal systems, healthcare, data governance, supply chains, and research systems.