Divine Architecture: Non-Algorithmic World-Selection and Rahmani Branch-Optimization in Strong Occasionalism

ABSTRACT The Leibnizian best-possible-world theodicy subjects divine sovereignty to an external axiological calculus, generating an intractable architectonic problem. Contemporary rivals (libertarian agent-causation, Thomistic concurrentism, Molinism) attempt escape via metaphysical "hatches" that, upon analysis, collapse into hidden occasionalism or incoherence. This paper argues that a coherent theology of divine sovereignty and love requires embracing strong occasionalism as its foundation. Building on two companion papers—which establish the incommensurability of world-types via Category Theory and resolve the occasionalist "author of sin" paradox via Formal Agency—we construct the Two-Tier Sovereign Model. In Tier 1, God makes a non-algorithmic, sovereign selection among incommensurable world-types (e.g., C_W1: Redemptive Struggle). In Tier 2, within the chosen type, God instantiates the Rahmani-Optimal Branch via a Threefold Filter that optimizes for essential consistency, maximal formal agency, and contextual divine mercy (rahmah). This model synthesizes absolute divine causality with meaningful creaturely responsibility, reframes the problem of evil as a question of sovereign expressive choice and rahmani-optimal context-setting, and yields the Rahmani-Optimality Thesis: our world is the optimally merciful instantiation of a sovereignly chosen kind of goodness. The result is a robust, non-consequentialist theological architecture that affirms God as both perfectly sovereign and perfectly good.