Interlanguage and Mathematical Translation Methodology Sidecar

This is the methodology and reproducibility sidecar for the AI-run mathematical translation program. It gathers the working methods, source-body baselines, linguistic routing material, weighted automata and terminology ledgers, normalization evidence, validation scripts, and bounded production experiments used alongside the author and series records. Open first: 00_Interlanguage_Romance_Methodology_v10_Status_20260718.pdf, a two-page reader-facing account of the current Romance v10 state and its reproducibility boundary. The exact v10 machine-gated payload is grouped in 14_Interlanguage_Romance_Methodology_v10_Evidence_20260718.zip. The earlier v0.6 Interslavic normalization report remains available under its now-explicit 01_Interlanguage_Interslavic_Normalization_Method_v06_20260718 name. Romance v10: the maintained infrastructure records 61 explicit language or variety routes, 148 corpus representations, 60 WordWeb concepts, 106 senses, 682 reviewed occurrences, 104 controlled terminology links, and a 194-row crosswalk to the current Noether Spanish and French terminology ledgers. Six bounded controlled-Romance production tranches, T001-T006, compile and reproduce their pinned renders. The validator passes 307 hash-bound targets. Research boundary: the broader Romance program remains active and incomplete. Fifty-three routes still have no source body, 33 senses have no accepted support, and the 954-row access grid contains exactly zero human observations, zero pilot-eligible rows, and zero form promotions. Orthographic diagnostics are not intelligibility measurements. The T001-T006 outputs are not native-language validated, and this record does not claim a complete Romance translation or an empirical marginal-intelligibility index. The record also retains prior source-body transfer packages, Slavic and RTL/Persianate/Arabic/Romance/CJK/Turkic/Indigenous/SEA baselines, the weighted Interslavic automaton and term ledgers, v0.4/v0.5 research-department material, and v0.6 normalization evidence. Those earlier components remain version-labeled rather than being presented as the current Romance method. These are versioned working research artifacts for inspection, reproduction, correction, and continued development. They are not native/community certification, a critical edition, or proof that all linguistic choices are settled. Corrections can be proposed through the GitHub mirror.