Why Pascal's Wager Needs a Dual-System Framework: Evaluative System Asymmetry and the Limits of Expected Utility

This article examines Pascal's Wager as a boundary case in decision theory and the philosophy of probability, rather than as a religious or theological argument. The central claim is that standard expected utility analysis fails not merely due to technical issues such as infinite utilities or competing hypotheses, but because it presupposes a unified outcome space where none is structurally available. Three considerations motivate the analysis. First, Pascal's Wager was not intended as a general argument for theism, but as a rhetorical device addressed to skeptics maximally distant from faith. Interpreting the wager as a universal rational proof is a historical and methodological misapplication. Second, both Pascal and his modern critics implicitly assume that the problem belongs to a single evaluative system. Building on the dual-system framework developed in Foundations of Intellectual Substrate Mining and Gospel Formalization, this article argues that Pascal's Wager necessarily spans two asymmetric systems—the finite probabilistic framework of earthly decision-making and a distinct evaluative regime invoked by its promised outcomes. Single-system probability is therefore inapplicable by category, not merely by technical failure. Third, Pascal's claim that the earthly cost of faith is small is a deliberate missionary simplification, not an ontological assertion. The Christian tradition recognizes that, within worldly terms, the cost of faith is often substantial. Philosophical difficulty arises only when this rhetorical reduction is incorporated literally into probabilistic models. The article introduces the concept of evaluative system asymmetry to characterize this structural limitation and distinguishes it from the epistemic limitations emphasized in recent work on transformative experience (Paul, 2014; Pettigrew, 2015, 2019). The argument extends these contributions by showing that cross-system decisions lack the unified outcome space that expected utility theory requires—a structural problem that persists even with perfect knowledge of post-decision states. Taken together, these points reframe Pascal's Wager as a limit case for probabilistic rationality. The article argues that expected utility theory remains precise and powerful within its proper domain, but loses normative authority when applied to decisions involving transitions between incommensurable evaluative systems. This limitation generalizes beyond Pascal's Wager to any transformative decision where choosing alters the chooser's evaluative framework itself. Acknowledgment: This article was revised following correspondence with Professor Alan Hájek, whose remarks highlighted the relevance of transformative-experience frameworks in decision theory and prompted clarification of the decision-theoretic implications of evaluative system asymmetry.