Auditability Before Ontology: Operational Gates for Subjecthood Claims and a Falsifiable Framework for Stakebearing Identity and Governance

This record contains the preprint "Auditability Before Ontology: Operational Gates for Subjecthood Claims and a Falsifiable Framework for Stakebearing Identity and Governance" (v1.0).

What this is
This paper proposes a falsifiable, governance-relevant framework for evaluating claims that large language models (LLMs) constitute stakebearing subjects with persistent identity and moral standing. The central claim is methodological. Evidence of internal structure, behavioral stability, or value-like representations does not by itself justify subject-level conclusions unless explicit bridge conditions are satisfied. The framework converts those bridge conditions into measurable gates and runnable protocol sketches, and introduces a write-path discriminator intended to separate wrapper-manufactured continuity from properties attributable to the base model.

Why this exists
The stakes are practical. If subjecthood claims are treated as plausible without operational criteria, accountability for deployment harms can be displaced into metaphysical ambiguity. The goal here is auditability. If an argument implies moral standing or liability-relevant subject status, it should also imply testable properties and concrete evaluation requirements.

What is included

• Full preprint PDF (canonical for reading and citation).

• Supplementary material (protocol detail, mappings, and appendices) as provided in this record.

Scope and non-scope
This work does not attempt to prove or disprove consciousness as a metaphysical thesis. It restricts itself to properties that can, in principle, be tested and audited in real deployments, and to the governance implications of upgrading mechanism evidence into subject-level claims.

Versioning
This record uses semantic versioning. Minor versions (v1.1, v1.2) are reserved for clarity and correction without changing the core framework. Major versions (v2.0+) indicate substantive changes to gates, protocols, or argument structure.