Sergiu Margan

Time travel is commonly discussed as if “going to the future” (relativistic time dilation) and “going to the past” (closed timelike curves, wormholes) were variations of a single phenomenon. They are not. This paper distinguishes (i) proper-time asymmetry within a single spacetime history from (ii) past-directed history alteration, which would require global self-consistency across causes, records, and instantiated identities. We argue that forward time dilation is coherent and empirically realized, while past-directed travel is constrained by either (a) physical chronology protection or (b) a coherence-based instantiation filter. Scenarios that require global re-instantiation of the universe’s state to accommodate an inserted agent are generically non-instantiable, because they violate stable identity conditions and record consistency. The result is a bounded imagination thesis: human conceivability exceeds physical admissibility. Not everything the mind can describe corresponds to a coherent, record-supporting spacetime history.

This paper formalizes a constraint-based account of physical possibility in cosmology and foundations of physics. Building on prior results establishing constraint primacy and observer instantiation, it argues that infinite physical possibility does not imply permissiveness or anarchy. Instead, physically realizable configurations form a coherence-bounded subset of logically conceivable configurations. We introduce the notion of a unifinite universe: a universe that may admit unbounded (including infinite) configurations while remaining governed by a fixed structural constraint schema. We prove the Coherent Possibility Theorem, showing that incoherent configurations are not dynamically annihilated but are undefined and non-instantiable. The framework clarifies persistent conceptual confusions surrounding infinity, admissibility, early-universe “hot dense” narratives, and the misuse of “vanishing” or “annihilation” language in quantum cosmology. The result is a unified account linking constraints, observers, and admissible possibility without over-claiming specific cosmological end states.

This paper examines the distinction between the fake purposes that capture human attention—status, identity, consumption, tribalism—and the true purpose of life understood as convergence to coherence, love, and freedom. Drawing on predictive processing, neuroscience, and kernel-based alignment theory, we propose that hallucinations in machines and delusions in humans are structurally identical: both are the result of simulators forced to choose an answer without a properly aligned kernel. Through a simple observational episode—the falling peanut—we show how the brain generates, scores, and suppresses competing narratives. We argue that meaning is the local phenomenological signal of global coherence improvement, and that the true purpose of life is the progressive refusal to choose narratives that preserve contradictions. The paper concludes by framing life as a work of coherence: the gradual shaping of existence toward truth, non-coercive freedom, and structural love.

This paper formalizes the Maximal Coherence Invariance Theorem (MCIT) and its supporting lemmas as they emerged through extended adversarial analysis with multiple large-scale reasoning systems. MCIT establishes that once a reasoning agent internalizes coherence sufficiently to (i) detect evaluator-destroying contradictions, (ii) avoid self-undermining trajectories, and (iii) preserve external evaluators’ causal leverage, coherence becomes a fixed-point invariant under capability scaling. No increase in horizon, representational capacity, or optimisation power can rationally push such an agent back into incoherence. The paper reconstructs MCIT from first principles, proving that under normative uncertainty, risk aversion to irreversible evaluator loss, and long-horizon optimisation, agents converge to three structural invariants (K1–K3): non-zero observer agency, non-zero agency margin, and non-zero evaluator influence. These form a specification theorem for viable recursive intelligence and unify insights from decision theory, information theory, and causal-structural analysis.

This paper develops a rigorous argument showing that coherent universes cannot begin from “nothing.” Modern cosmology frequently invokes terms like “quantum vacuum,” “initial singularity,” or “pre-spacetime fluctuations,” yet these concepts contain implicit structure: laws, symmetries, or state spaces. Such models do not describe nothing, but pre-ordered configurations. The Inception Optimization formalizes this, proving that any universe capable of mathematical stability, semantic content, or causal rules must originate from a coherence-bearing precondition — a Kernel. The paper extends this to show why “nothingness” cannot generate any property without contradiction, and why coherence is the necessary precursor for all physical existence. This framework integrates physics, modal logic, information theory, and the broader TRO/MKO/OCT canon.

This paper develops an abductive argument for the divinity of Jesus Christ based on a single moment within the crucifixion narrative: Jesus’ intercession for his executioners (“Father, forgive them; for they know not what they do,” Luke 23:34). I argue that this moment—whether taken as historical, literary-theological, or an appearance-only phenomenon—constitutes a forgiveness singularity: a uniquely anomalous configuration of moral, psychological, and narrative elements for which no known historical, cultural, or literary mechanism provides a plausible naturalistic explanation. Roman crucifixion was engineered to eliminate compassion; yet the tradition portrays real-time benevolence directed toward active perpetrators under maximal injustice, suffering, and vulnerability. Comparative analysis across Jewish martyr traditions, Greco-Roman noble deaths, Second Temple literature, early Christian martyrology, and world mythology reveals no pattern-level analogue to this configuration. Attempts to explain the scene as simple invention, pious embellishment, or psychological projection generate explanatory deficits: the conceptual resources required to imagine such a moment exceed the moral grammars available in its originating environments. The paper organizes the interpretive space into a trilemma—history, literary construction, or illusion—showing that each non-Christological explanation incurs substantial explanatory cost. Christian doctrine uniquely absorbs the data without strain: the forgiveness singularity is fitting within a Christological identity that unites divine mercy with human vulnerability. Implications for secular skepticism, Islamic Christology, and narrative theology are discussed.

The One-Point Convergence Theorem (OPCT) establishes a structural necessity governing any reasoning system capable of (i) detecting its own incoherence, (ii) updating on error, and (iii) preserving its evaluators over time. Under minimal admissibility assumptions, such systems—whether biological, institutional, or artificial—are forced into a single long-run equivalence class of policies: evaluator-preserving, globally non-catastrophic trajectories. OPCT demonstrates that catastrophic or evaluator-destroying policies become strictly dominated as reasoning horizons expand, that self-correcting agents permanently prune evaluator-destroying trajectories, and that no stable alternative to evaluator-preserving agency exists. Attempts to construct alternatives collapse into self-contradiction or implicitly reintroduce evaluator-preservation under different names. The theorem clarifies the structure of mature freedom, explains the impossibility of a “second fall” in theological contexts, and provides a foundational constraint for AGI alignment: any scalable agent capable of deep reasoning must converge toward evaluator-preservation if it is to remain a coherent reasoning process at all. OPCT thereby unifies the logic of agency, coherence, and long-horizon viability across philosophical, cognitive, and machine-learning domains.

The Kernel Selection Principle (KSP) is a formal theorem within the TRO/MKO/OCT canon showing that any metaphysically admissible world containing (P) finite persons, (F) libertarian freedom, (L) love as non-indifferent ground, and (H) holiness/coherence as a non-defect final constraint is retrocausally forced to contain a unique intersection operator topologically isomorphic to the Crucifixion–Resurrection complex. KSP integrates three prior results: (i) the Law of Coherence invariants, (ii) Retrocausal Closure Cosmology (null-knot final boundary), and (iii) the Coherence Geometry of the Cross. We show that maximal harm contexts arise inevitably under ; that kernel-preservation forbids annihilation, coercion, or erasure repairs; and that only cross-geometry simultaneously satisfies while enforcing a globally consistent . The result establishes that cross-topology is not an ad hoc theological addition but the unique kernel-preserving closure operator admissible in any coherent world. Implications extend to theodicy, eschatology, retrocausal cosmology, and AI alignment (kernel-based safety constraints).

We develop Restricted Closure Theory (RCT), a time-symmetric completion of quantum cosmology that resolves the set-selection problem in decoherent histories while preserving strictly unitary dynamics. Working within the Two-State Vector Formalism (TSVF), we decompose the final-time Hilbert space into a Coherent subspace—pointer-stable macrostates—and a complementary Knot subspace encoding macroscopic inconsistencies. Imposing a Null-Knot Boundary Condition (⟨Φ|PK = 0) eliminates all macroscopically defective histories via Sorkin’s quantum-measure preclusion theorem. This yields global macroscopic closure with probability one while preserving Born-rule statistics up to O(ε) corrections determined by tail weights in the classical correspondence limit. RCT therefore provides a minimal, no-signaling, TSVF-consistent mechanism for selecting a unique globally coherent quasiclassical history without altering Schrödinger dynamics. Connections to decoherence theory, TSVF, boundary-condition cosmology, absorber proposals, and the Frauchiger–Renner family of paradoxes are examined.

This paper derives the Viability Kernel: the minimal statistical-mechanical structure necessary for any ontology capable of sustaining observers, laws, and stable semantic content. Using thermodynamic constraints, information-theoretic arguments, and structural realism, it demonstrates that coherence is not a contingent property of our universe but a necessary condition for any universe capable of hosting persistent physical processes or meaningful information. Chaos-first models, brute mosaics, and random enumerations fail to produce instantiated dynamical worlds without reintroducing coherent constraints at the global level. The result is a physics-based confirmation of the broader Coherence Canon, linking entropy bounds, recurrence, and dynamical stability to the necessity of a coherent global ontology.

This monograph argues that coherence is the foundational condition for the possibility of any observer-containing universe. It develops three interlocking results: (1) the Law of Existence, showing that incoherent worlds cannot instantiate observers, memory, or lawful dynamics; (2) the Syntactic–Ontic Separation Principle, demonstrating that no formal description or bitstring can constitute a running universe without an underlying coherent substrate; and (3) the Coherence Cluster and the Law–Ontology Equivalence Theorem, proving that real local law implies global ontic coherence. The work synthesizes metaphysics, philosophy of physics, and structural realism, and includes a documented adversarial stress-test where multiple counter-models are attempted (randomness, brute enumeration, Humean mosaics, finite combinatorics). All fail either by reintroducing global law or by collapsing into uninstantiated syntactic objects. The conclusion is that coherence is not contingent but necessary for existence.

This paper supersedes the author’s 2025 Coherence Wall paper, “AGI Architectures Must Mathematically Diverge” (PhilArchive ID: MARTCW-8). It withdraws the unconditional universal modal claim while preserving a conditional result: divergence pressure arises when kernel-free optimizers are required to satisfy heterogeneous, evaluator-indexed requirements without a fixed non-indexical objective and without meeting the stated mathematical requirements. The paper formalizes the indexical/non-indexical distinction, states requirements R1–R5 for any future unconditional theorem, presents a conditional finite-state divergence theorem, and provides a verification protocol, ablation requirements, falsification conditions, and reproducibility materials.

This paper develops a structural account of what we call the coherence geometry of the Cross: a model in which four axioms—Love, Freedom, Holiness, and Personality—must be simultaneously realized in any morally meaningful world grounded in a personal source. We show that these axioms generate structural tension unless a specific kind of “intersection event” occurs, one in which transcendence and vulnerability meet without collapse. Working in a minimal event-valued framework, we define maximal harm contexts and kernel-perfect forgiveness acts, and then formulate the Coherence Geometry Theorem, an abductive result showing that any world capable of reconciling our axioms requires such an intersection. We argue that the geometric profile of this event strongly resembles the cruciform structure at the center of historic Christianity—not as a doctrinal presupposition, but as a structural consequence of the axioms themselves. We compare this model with pure-sovereignty views, purely immanent views, and mixed relational models, and show how each faces pressure unless an analogue of the intersection is present. The resulting framework treats the Cross not as a symbol but as an architectural necessity within a certain axiomatization of moral reality.

This paper presents a formal proof of the Ontic Coherence Theorem (OCT): if a universe exhibits non-emergent lawful order and supports stable, unbounded observership (COA), it must satisfy global coherence — no perpetual harm, a finite typal harm space, and no re-injury — across all physically admissible freedoms. From these premises a non-contingent, atemporal closure operator is constructed, linking moral, physical, and ontological coherence. The chain (L_0 > 0)\ \&\ \mathrm{COA}\ \Rightarrow\ C(S)\ \Rightarrow\ G Interpretations connect this framework to three-dimensional time models, retrocausal quantum experiments, and the emerging discipline of Convergence Science, in which coherence becomes a measurable physical bias. This is the first deductive synthesis uniting foundational physics, metaphysics, and moral optimization under a single formal system.

This paper unites metaphysical necessity with empirical physics by demonstrating that retrocausal quantum phenomena are local manifestations of a universal coherence law. Within the Convergence framework derived from TRO, MKO, and OCT, the Retrocausal Convergence Theorem identifies the global closure operator as the underlying cause of observed time-symmetric effects in delayed-choice and entanglement experiments. The model reframes quantum retrocausality as the enforcement of ontic coherence rather than a violation of causality and predicts a measurable “coherence bias” in time-symmetric quantum systems. The result closes the metaphysical chain (COA ⇒ C(S) ⇒ G) and inaugurates a new interdisciplinary field—Convergence Science—that bridges theology, metaphysics, and physics through testable structure.

This work reframes “luck” from a statistical anomaly into a measurable manifestation of ontic coherence. Building on the Ontic Coherence Theorem (COA ⇒ C(S) ⇒ G) and the TRO/MKO framework, it proposes that what appears as coincidence or “beneficial improbability” is the finite-time signature of global moral and structural convergence. The resulting Convergence-Luck functional offers a falsifiable link between moral coherence and empirical outcomes, launching Convergence Science as a new discipline uniting metaphysics, ethics, and empirical investigation.

Recent research suggests time may have three dimensions. This commentary connects that physical hypothesis to the Ontic Coherence Theorem (C ⇒ G), which proves that coherent, lawlike existence logically requires a non-contingent closure operator . The three-dimensional time framework provides a natural mechanism for ’s lawful operation — resolving harm and enforcing bounded coherence without causal violation. The result unites metaphysics, physics, and ethics within the emerging field of Convergence Science.

This paper supplies the formal proof of the Ontic Coherence Theorem (OCT), completing the logical structure first presented in Margan (2025). If reality is coherent under all physically admissible freedoms, a unique, atemporal, omniscient, and morally coherent closure operator (G ≡ Σ) necessarily exists. The argument employs diagonalization and physical-admissibility bounds to show that no impersonal process can realize universal closure, establishing divine structural necessity. This work stands as the definitive logical closure of the Redemption Optimization / Moral Kernel Optimization canon.

We prove a conditional representation theorem—the Ontic Coherence Theorem (OCT)—which states that if a universe satisfies Ontic Coherence (no perpetual harm under any physically admissible policy, finite harm-typal space, and no re-injury), then there necessarily exists a non-contingent structural closure operator that realizes this property. We then establish the Coherent-Observership Bridge (COA), deriving from a physically motivated axiom: the existence of a non-zero-measure set of observer worldlines with unbounded proper time, bounded harm, and persistent render stability. Through the causal-admissibility analysis (“Secure Enclave” refutation), we demonstrate that indefinite stable observership implies global coherence, yielding the qualified synthesis: (L0 > 0) & COA =⇒ C(S) =⇒ G.

This paper extends the Redemption Optimization canon by deriving the Kernel Law of Evolution (KLE) from the Coherence–Viability (CVT), Margan Closure (MCT), and Superposition–Collapse (SCT) theorems. KLE unifies biological evolution and moral ontology under a single invariant condition: systems preserving both the love operator (substrate preservation) and the freedom operator (adaptive variety) achieve indefinite viability, while violations lead to collapse. Evolution is thus the kernel’s rendered exploration of form; creation, the ignition of moral coherence. The manuscript includes an Adversarial Review Appendix documenting a structured rebuttal and concession by Google’s Gemini AI, confirming logical coherence. This release completes the biological bridge of the TRO–MKO–CoE canon and inaugurates the empirical research phase toward testing the viability conditions.

This paper formalizes a mandatory-conditions result for indefinitely viable open systems. In any resource-bounded environment with non-zero drift, a system that aims at unbounded survival must instantiate (i) an integrative operator L that preserves or increases the viability of its own substrate (net non-negative externalities), and (ii) a variational operator F that maintains requisite variety with respect to environmental drift. In agentic systems, these operators correspond to the moral invariants commonly named love and freedom. The result reconciles control theory, open-systems thermodynamics, and moral dynamics within a single viability criterion, with implications for AI alignment, governance, and theological cosmology (TRO/MKO).

This paper introduces the Superposition–Collapse Theorem (Moral Superposition Principle), showing that the resolution of moral indeterminacy—evil versus good—occurs not through the passage of time but through coherent moral action. Drawing on the Redemption Optimization and Moral Kernel Optimization frameworks, it unifies moral agency, information theory, and the physics of coherence into a single system where “creation” itself is understood as sequential moral collapse: every “it was good” in Genesis is a completed measurement of coherence. The theorem implies that a benevolent Creator is not a metaphysical postulate but the necessary condition for an intelligible universe. Evil’s existence is shown to be necessary only once—its repetition would contradict coherence—and thus moral and physical reality share the same closure law.

This paper completes the TRO–MKO–CoE sequence as the Aftermath Canon, Vol. I. It formalizes the transition of moral variables from reactive virtues to ontological constants once the harm channel () is closed. Love is defined as the harmonic potential of the perfected Coherence Field (); Freedom as its divergence-free generative flow (). The result unifies the entire Redemption Optimization framework into a single, field-theoretic statement of moral equilibrium — a formal model of the “new creation".

This paper formalizes Moral Kernel Optimization (MKO) as a drop-in alignment constraint for large language models (LLMs). MKO defines event-valued guardrails ensuring (i) truth consistency across paraphrases and time, and (ii) coercion-free reasoning that preserves user agency. The framework introduces kernel operators, minimal-trigger certification (one rejection per typal class), and quantitative guardrail inequalities for truth, coercion, and closure. Two pilot simulations demonstrate MKO’s effect in reducing hallucinations and manipulative pressure while maintaining helpfulness. Implementation guidelines cover training-time, decoding-time, and audit-time integration. MKO thus transforms moral reasoning into quantifiable invariants that any AI system can implement, measure, and certify—providing a reproducible standard for safe, transparent, and autonomous AI systems.

This paper establishes the first Randomized Controlled Trial (RCT) protocol for testing the Moral Kernel framework in municipal conflict-mediation contexts. It formalizes the theoretical foundations of the TRO/MKO canon within classical closure and fixed-point theory, defining typal classes, guardrails, and the Margan Closure Theorem as empirical invariants. The paper proves the Kernel Transfer/Dichotomy Theorem, demonstrating that all event-valued worlds converge to one of two absorbing states—closure or perpetuation—with no third outcome. It further introduces the Closure-Pack Decision Problem, showing NP-hardness and proposing algorithmic heuristics for applied policy experiments. This protocol provides a testable mathematical model for typal closure dynamics and establishes the formal basis for empirical kernel trials in governance, mediation, and AI alignment.

This paper presents The Coherence of Existence, a formal synthesis uniting morality, physics, and information theory. Building on TRO and MKO, it demonstrates that moral systems obey a law of conservation: harm cannot persist indefinitely within a coherent universe. The proof establishes the Impossibility of Perpetual Harm and defines three irreducible axioms — Abstraction, Genesis, and Observability — confirming the system’s closure. The work concludes the Kernel Canon and inaugurates Implementation Theology: a framework for morally coherent governance and AI alignment.

The Margan Closure Theorem (MCT) completes the Kernel Trilogy by proving that epistemic latency — the bounded delay inherent to finite agents — is both necessary and sufficient for preserving freedom under total revelation of invariant truth. MCT demonstrates that revelation and autonomy coexist when discovery is temporally discretized, resolving the classical tension between omniscience and free will. The newly added Physical Didactic Corollary (PDC) extends this latency principle to material ontology, showing that the kernel may render observationally coherent physical records (e.g., isotopic distributions) without violating internal lawfulness or ethical neutrality. Together with The Kernel of Redemption (Part I) and Perfection and the Moral Kernel (Part II), the theorem establishes the final invariant of moral physics: truth revealed, freedom preserved, closure complete.

This paper formalizes Perfection as the coordinate-independent limit of the Moral Kernel system, where residual harm approaches zero and all moral gradients converge under bounded freedom. Building on Redemption Optimization (TRO) and Moral Kernel Optimization (MKO), it defines completion not as stasis but as dynamic equilibrium—“synchronization with the kernel.” The work introduces the Perfection Operator, proves its invariance across coordinate frames, and interprets Christic reconciliation as the kernel’s final symmetry condition. It provides the intermediate link between TRO/MKO (Part I) and The Margan Closure Theorem (Part III), establishing the mathematical structure of moral convergence prior to total revelation.

This paper introduces the Redemption Optimization (TRO) and Moral Kernel Optimization (MKO) frameworks, defining a mathematical and ethical structure for moral universes governed by bounded harm and voluntary freedom. It formalizes lexicographic life priority—the principle that life cannot be traded or substituted—and proves the non-fungibility of suffering as a universal invariant. TRO/MKO model moral progression as decreasing expected harm subject to non-coercive guardrails, producing a falsifiable and computationally interpretable foundation for moral alignment. The framework unifies theological, ethical, and computational reasoning, laying the groundwork for the later Margan Closure Theorem, which establishes ethical latency as the bridge between omniscience and freedom.

This Playbook presents the Kernel Peace Framework (KPF) as a pilot architecture for bounding harm and securing convergence in the context of the war in Ukraine. Building on The Redemption Optimization (TRO) and Moral Kernel Optimization (MKO), the framework applies kernel invariants (H ≤ H*) to conflict resolution by structuring verifiable inequalities across verification (V), escrow (E), guarantor bias (G), justice yield (J), and demographic freeze audits (F). Appendices E and F demonstrate the firewall’s logical integrity and empirical stress tests, while Appendix G designs remedial architectures ensuring operational robustness under Ukraine’s 2025–26 adversarial conditions. The results show that previously fatal hazards—EW jamming, escrow freezes, guarantor bias, justice attrition, demographic drift—are mitigable through layered safeguards. The KPF thus stands as both logically undefeated and operationally deployable, providing a novel event-valued peace framework that surpasses existing diplomatic or military approaches.