La Profilée: The Domain Translation Protocol and Its Cross-Domain Application

This paper presents a reproducible procedure — the Domain Translation Protocol — for applying La Profilée (LP) to any domain, together with its application across eight natural and made domains. The protocol rests on a single axiom: the domain supplies material (an antecedent existence indicator and an antecedent diachronic identity relation ∼_id), and LP supplies the verdict (Q1, Q2). Two verified anchor-corollaries follow: existence is anchored on the domain’s existence indicator through Q1 = R/(F·M·K) ≤ 1, and identity is anchored on the domain’s ∼_id as the core’s entry primitive. Domains are classified by what they antecedently supply into three anchor-classes — A (threshold-suspended), B (non-unique), C (antecedently absent). The roles F, M, K, R are identified as falsifiable bridge assumptions; integration capacity is represented canonically as F·M·K, the multiplicative role-class being derived and the load-bearing content being non-substitutability and role-collapse rather than the exact form. Eight runs are reported — three double-counter-read, five natural single-counter-read, one diagnosis — forming a graded proof spectrum from crystal growth and chemical species (near-unique anchors) to ecosystem (soft at both anchors). Counter-reading is a constitutive step of the procedure and recurrently corrects one systematic error class: existence conflated with identity, or anchor-plurality missed. The proof/diagnosis distinction is grounded in whether an external persistence-goal exists, not in the natural/made dichotomy. Keywords: persistence, identity, cross-domain universality, anchor classes, Frame–Module–Coupling, circularity, proof versus diagnosis, counter-reading