Collective Moral Disengagement as Evolutionarily Stable Strategy (ESS): Extended Phenotype Theory and the Intra-Corporate Intentionality Paradox

The literature on collective moral disengagement (CollMD) explains how organizations construct shared narratives that justify unethical behavior, but it does not specify under what conditions that state is evolutionarily stable, nor why it persists beyond the individuals who generated it. This article applies Extended Phenotype Theory (EPT) and multilevel Evolutionary Game Theory (EGT) to reinterpret CollMD as an evolutionarily stable strategy (ESS) in organizational environments with high internal normative rigidity, measured through an Organizational Lock-in Index (OLI) derived analogously from the Constitutional Lock-in Index (Lerer, 2025a). The article argues that CollMD is not merely emergent but actively selected because it builds extended phenotypes: performance evaluation systems, reporting structures, and accountability-diffusing narratives that reproduce the collective moral state independently of individual turnover. The transmission mechanism is formalized through Heteronomous Bayesian Updating (HBU), which predicts that the speed of CollMD adoption correlates with the visibility of reactions to dissent rather than with the argumentative quality of available moral justifications. CollMD is further identified as the psychosocial mechanism that resolves the intra-corporate intentionality paradox of Asymmetric Intentionality Theory (AIT): it enables Level 3 individuals to operate within Level 1 organizations by collectively reconstructing the moral space until both levels appear compatible. The framework is applied qualitatively to four documented cases (Volkswagen, Siemens, Wells Fargo, and Purdue Pharma), identifying OLI trajectories, extended phenotype mechanisms, and tipping points consistent with EGT predictions. Analysis suggests that OLI exceeds 0.65 in all four cases before unethical behavior becomes systematic, that CollMD extended phenotypes are detectable between 18 and 36 months before visible risk materializes, and that effective intervention requires altering the architecture of visible reactions rather than the content of circulating narratives.