Appendix: Subjectivity under deterministic constraints
A recurring criticism in formal reasoning systems involves the paradox of determinism: if the model is perfectly deterministic, how does it handle the inherently subjective nature of human judgment, particularly at the last mile of analysis?
StateCraft addresses this not by denying subjectivity, but by forcing it into explicit, structurally inspectable forms.
The layers that matter
StateCraft refuses to treat knowing as a single act. It approaches knowledge philosophically through the Triad: ontology (what is taken to exist), epistemic order (what is known and how), and moral structure (what ought to govern action).
Operationally, the epistemic order flows through an admission pathway built to preserve exact boundaries. Raw observations enter first. Admitted facts form the canonical ledger. Formal beliefs carry structured uncertainty for latent variables. Deterministic projection computes state from facts and beliefs. Interpretation and simulation apply lenses and sandboxed counterfactuals on top.
Each boundary enforces a change in epistemic status. Subjectivity does not disappear across these layers. It moves.
Where subjectivity enters
The model is deterministic architecturally. Same facts, same beliefs, same time, same projection. Always. But the criticism pointing to subjectivity at the last mile (and the first mile) is entirely valid. The system does not eliminate subjectivity. It quarantines it.
Three places where human judgment enters the model:
Interpretation. No state derivation simply acts on its own. The final projection requires meaning to be extracted. StateCraft uses lenses for this. Deciding the threshold for “escalation,” or what constitutes a “fragile” regime, is a human choice shaped by moral commitments and operational context. Axiom VI requires that these lenses declare themselves openly as explicit classification rules. By forcing this, subjectivity leaves the realm of hidden assumption and becomes an inspectable, debatable object. You can argue about the lens parameters without confusing the interpretation with the derivation.
Narrative rendering. The final delivery often involves rendering structural derivations into reports. A narrative inherently decides what to emphasize and what to downplay. Axioms VII and X handle this: rhetoric is recognized but made permanently subordinate to structure. Any summary, whether generated by AI or written by a human, must maintain traceability to the derivation graph. The narrative is the surface. The derivation is the skeleton.
Belief formulation. Assigning distributions to beliefs requires curated judgment. Setting a decay profile or choosing a prior is a subjective determination. The system allows this but requires it to be mathematically formalized. By forcing subjective priors into formal distributions and temporal decay functions, contradictions with incoming factual data can be detected programmatically. Subjective bias is not removed. It is made visible and correctable over time.
What the model actually claims
According to the canon’s own philosophical commitments, completely objective reasoning does not exist once information reaches the moral layer. The model is deterministic in its architecture, but its parameters, its interpretation schemas, and its narratives are necessarily subjective.
The claim is not objectivity. The claim is legibility. Subjective choices that are declared, inspectable, and contestable are categorically different from subjective choices that are hidden inside a smooth output surface. The former can be debated, revised, and held accountable. The latter cannot.
That is the boundary the model defends.
This text was produced under the Canon Authoring Protocol. See 00-authoring-protocol.md, Author’s Declaration.