What Must Be Assumed? An Audit of Primitive Structure in the Distinction Framework, and a First Test of Minimal Rule Inheritance

This paper has two parts and a single purpose: to state plainly what the distinction-primitive
research program still assumes rather than derives, and to report the first case in which one of
its own previously published theoretical commitments – minimal rule inheritance – was tested
directly against computational evidence rather than left as an illustrative claim. Part I audits
the primitive vocabulary used across this research program (potentiality, distinction, persistence,
compatibility, organization, topology, geometry), preserving the specific defect already identified
for each term rather than restating them as generic open questions. Part II reports a first direct
test of minimal rule inheritance using the DLL Laboratory platform’s six-rule registry and its
two replicate-supported findings, and documents a two-layer audited inference procedure – a per-
seed epistemic classification tree and a cross-seed replicate aggregation – constructed specifically
to prevent this test from producing a false claim of dependency or independence. The procedure
is presented as an audited version 1, not a closed taxonomy: each of three successive review
rounds during its construction identified a genuine structural gap, and each gap was checked
directly against the evidence in hand and found not to alter either reported classification. Both
findings and the procedure itself are reported with their limitations stated in the same sentence
as their results, not deferred to a separate section.

Keywords: primitive structure; potentiality; distinction; minimal rule inheritance; ablation anal-
ysis; replicate stability; epistemic decision procedure; DLL Laboratory