Society & Economicspreprint2026-08-17

Layered Causal Realism: An Overview

Open access0 citations

Abstract

Layered Causal Realism (LCR) is a methodological framework for diagnosing and constraining explanatory migration across domain boundaries. It imposes two related constraints: any explanatory inference that crosses an explanatory-domain boundary requires independent authorization, and cross-layer correlation—including stable covariation, causal continuity, and predictive correspondence—does not by itself constitute such authorization. This Overview presents the methodological foundations of LCR and examines their development across a six-paper series, in which the framework is brought to bear on agency, phenomenal attribution, identity, normativity, and civilizational relations. LCR I reconstructs agency under determinism by separating action generation, deliberative experience, and the attribution of responsibility into distinct explanatory dimensions. LCR I.S extends this analysis to the diachronic structure of agency, showing how a subject inherits its own deliberative history through causal continuity across successive rounds of decision. LCR II develops a structural account of phenomenal attribution, introducing Rualia as a structurally warranted attributional position under conditions of ontological underdetermination. LCR III reconceives identity as task-conditioned adjudication organized around constitutive attributes rather than universally applicable criteria. LCR IV separates authorship from normative persistence, establishing an Invariant under which unresolved wrongful distortion remains normatively reachable even after the author-position becomes vacant. LCR V extends the analysis to civilizational scale, rejecting the assumption that motivational topology is fixed by intelligence alone and examining how heterogeneous motivational structures can generate complementarity, persistent friction, and an enduring normative layer. Taken together, these applications reveal a broader implication of LCR: properties long organized around the same subject need not share a single explanatory condition, allowing the subject to be treated not as the endpoint at which every explanation must converge, but as a point at which multiple causal, attributional, identity, and normative structures intersect.

// Source

View paper (DOI)Open access versionOpenAlexZenodo (CERN European Organization for Nuclear Research)Published 2026-08-17

Authors: Chenghao Qian