This paper develops a layered framework for auditing theoretical discourse by distinguishing among L1, L2, and L3 claims. Its central thesis is that many philosophical and scientific confusions arise not from metaphysics as such, but from unmanaged metaphysical intrusion into theoretical discourse. Drawing on STT-CR, the paper argues that the value of a theory should be assessed not by how strongly it negates prior theories, but by how clearly it specifies construction conditions, domain limits, defects, minimal completions, and empirical accountability. The discussion is developed through two case studies: block universe theory and temporal experience, and determinism and free will. The paper concludes that the aim is not anti-metaphysics, but metaphysical management: keeping scientific discourse within L1/L2 as far as possible, while preventing unmarked L3 claims from functioning as if they were internal theoretical items.
Teruhito Kojima (Fri,) studied this question.