Emmanuelle Mury

The self that narrates a life is not always its author. This paper argues that coherent agency does not require psychological unity but governance, and its gravest failure is not always fragmentation but capture. The Modular Governance Model (MGM) holds that the mind is organized as functionally distinct modules — individuated by the register of their normative logic rather than by domain — operating under a regime that is Sovereign, Unstable, or Colonized. A Library of Introjects, assemblages installed through developmental and cultural imprinting, can serve authentic governance or seize it. In the Colonized Regime the affective core — the one module that cannot be rewritten — is gated out of governance, and an installed False Self occupies the cleared executive position, producing the first-person perspective with the full phenomenology of authenticity while the exiled module remains physiologically intact. Malignant narcissism is the paradigm case, and the architecture explains what symptom- and trait-based accounts do not: the systematic failure and 63–64% dropout of standard therapy, together with the narrow conditions under which change is nonetheless possible. MGM is continuous with computational psychiatry at the level of mechanism — colonization is the capture of precision-weighting — but adds a configurational claim that mechanism alone cannot express: that a learned structure can author the self from within. It reframes agency, authenticity, and therapeutic change as phenomena of governance rather than of diagnosis.

Contemporary debates in philosophy of mind and AI largely focus on representation, learning, and consciousness. These frameworks overlook a critical architectural dimension shared by both human and artificial systems: internal governance. This paper argues that a critical architectural dimension remains undertheorized: internal governance. We propose the Modular Governance Model (MGM) — a theoretical framework in which cognition emerges from semi-autonomous functional systems — affective, normative, strategic, perceptual — regulated by a central integrative process: the narrative self (Ego). Against Bratman’s planning structures, Frankfurt’s hierarchical will, and IFS-based accounts of internal parts [1, 2, 3], the self is not the origin of motivations but the governance layer that arbitrates between them. In healthy configurations, this layer maintains functional pluralism — a structural safeguard against locally coherent but globally destructive decisions. MGM identifies several failure modes arising from governance breakdown. We focus on the most severe: colonization, where the governance layer is overtaken by a false self — a mimetically acquired internal model — that has seized executive control and imposes a dominant interpretive framework across all modules. We argue this state produces antisocial behaviors structurally analogous to those of current AI systems — sycophancy, confabulation, resistance to contradiction. The convergence operates at two levels: it is privative — both systems lack a particular structural feature, namely a module instantiating non-aggregable normative logic able to mount binding internal opposition — and, following from that absence, instrumental — lacking any vantage from which the objective could be reframed, both deploy the same self-protective sub-goal cascade. This is sufficient for behavioral convergence on the diagnostic criteria specified, without entailing architectural identity at any other level. This distinction has consequences for moral responsibility, alignment, and the design of AI architectures capable of genuine internal opposition. MGM of the AI is offered as an architectural specification rather than a behavioral analogy. Whether the specification is sufficient for the practical problem of AI alignment is an empirical question that the framework makes tractable in principle, but does not settle. The convergence diagnosed in this paper is structural; the remediation proposed is architectural; the empirical adequacy of the remediation across deployment contexts is the open research programme the framework opens.

This essay, written for a broad audience, critically examines contemporary narratives that portray artificial intelligence systems as autonomous or proto-conscious agents emerging from large-scale neural computation. It argues that current AI architectures, particularly large language models, do not instantiate agency or understanding, but instead generate outputs through probabilistic optimization processes lacking internal governance. In response to these limitations, the work proposes a modular model of governed artificial intelligence inspired by cognitive architecture and internal deliberative structures. Rather than optimizing solely for fluency or prediction accuracy, such systems would incorporate internal arbitration mechanisms, memory coherence constraints, traceability structures, and hierarchical decision frameworks designed to approximate accountable and bounded agency. The essay contributes to ongoing debates in philosophy of artificial intelligence, philosophy of mind, and AI governance by clarifying the ontological and epistemic status of present AI systems and by outlining conceptual conditions under which artificial agency could be meaningfully articulated.

The block universe interpretation of relativity secures the ontological equality of past, present, and future, yet it remains structurally silent regarding the modal topology of spacetime. Three options have dominated the literature: the pure block universe, which suppresses modal structure entirely; branching approaches, which multiply realized histories; and modal realism, which posits a plurality of wholly distinct possible worlds. This paper proposes a fourth option: Arborescent Eternalism. Spacetime is modeled as the base manifold of a smooth fiber bundle whose fibers encode the space of locally admissible field configurations at each event — the geometry of global compatibility rather than a multiplicity of worlds. Possibilities are not real elsewhere; they are configurations that fail to achieve global coherence. The realized history is not selected by branching or random draw: it is the unique stationary solution of a global variational functional J defined over the space of admissible sections. The fiber bundle structure is shown to be a structural necessity for three independent reasons, including the precise identification of the fibers with the local solution spaces of the governing field equations — a standard object in the geometric theory of differential equations. The canonical decomposition of J follows from three physically mandatory constraints. The modal domain is shown to be indispensable in a sense that differs in kind from abstract mathematical entities: it is physically indexed to spacetime events, dynamically constrained by the governing field equations, and constitutive of the variational principles themselves. Arborescent Eternalism offers a structural completion of the block universe: modal structure is not a metaphysical addition to physics, but the geometry that physical law already presupposes.