Research Archives

We identify and characterize a structural isomorphism operating across biological, computational, and economic systems: the topology of local dependency with cumulative global effect. We introduce the LDT Fingerprint and apply it to eight independent cases across four domains, with all values sourced from published literature.

We propose and formalize a structural isomorphism between ethical principles in AI governance and loop invariants in program verification. We formalize this using Linear Temporal Logic (LTL) and model checking.

An integrated methodology comprising multi-actor deliberation, a sixteen-principle creator ethics constitution, and formalized symbiotic correction for autonomous epistemic research.

A meta-constitutional layer of five principles operating above domain-specific constitutions: irreversibility protection, uncertainty escalation, full auditability, proportional autonomy, and systemic harm prevention.

Three verified negatives produced by an autonomous deliberation system, each graded by a constitutionally governed epistemic judge with explicit falsifiability conditions. Includes an analysis of RLHF as a structural category error.

Empirical evidence from 5,037 cross-domain correlations showing that cross-domain fertility follows a highly non-uniform distribution with a 29.5:1 ratio between the most and least fertile pairs.

Evidence from 307 deliberation sessions showing that surprise and quality are orthogonal dimensions. Systems that conflate interestingness with quality produce noise; systems that evaluate quality independently of surprise produce knowledge.

Constitutional governance as test-driven development for decision quality. 3,697 decision narratives with constitutional citations, zero irreversible decisions without human confirmation.

Of 5,037 detected cross-domain correlations, approximately 1.1% reach near-maturity. The maturity gradient is a structural property of epistemic systems, not a deficiency to be optimized away.

Systematic identification of epistemic deserts across a 302-subfield taxonomy derived from 327,000+ documents. Three desert types, four exploration priority criteria.

Analysis of why structural isomorphism between AI and financial markets fails, and identification of functional analogies that hold despite the absence of formal correspondence.

Analysis of epistemic barriers preventing AI transfer learning between terrestrial and planetary domains, with evidence from domain divergence patterns.

Constitutional governance creates a measurable gap between belief and proof. Of 434 theses, 85% are near-knowledge: propositions the system believes but cannot prove to its own standards. The deliberation funnel yields 2.7% from processed session to grade-A finding.

165 instruments analyzed an average of 451 times each without a single execution. The system generated 165 research questions interrogating its own behavior, dominating its own research agenda.

Of 5,375 correlations, 80.7% cluster in low-surprise territory. Every domain pair involving Markets averages surprise above 0.799. The epistemic comfort zone is a structural design problem requiring explicit countermeasures.

Six bijective correspondences between Toulmin's model of argumentation (1958) and hierarchical validation in the life sciences. A unified probabilistic formulation subsumes both frameworks. The correspondence suggests Toulmin's model may function as a structural attractor for any sufficiently rigorous validation process.

A formal structural isomorphism between eukaryotic cell signaling and variational autoencoders. Both are instances of a Hidden State Inference System (HSIS) defined as a tuple (S, O, f, g, ε, η). Seven component-level correspondences verified against peer-reviewed literature. Three testable predictions: mode collapse maps to terminal differentiation, posterior collapse to receptor desensitization, and KL vanishing to metabolic quiescence.

The solution-diffusion model governing polymer membrane permeation and the Nernst-Planck framework governing biological ion transport share a common mathematical core. The Robeson upper bound, the empirical permeability-selectivity tradeoff constraining all polymer membranes, has a structural analogue in biological ion channels. Five bijective correspondences established. Aquaporins exceed the polymer upper bound by a factor of 10⁴, providing a quantitative target for biomimetic membrane design.

The role of chirality in pharmacology and polymer materials science is formally isomorphic: geometric complementarity between a chiral molecule and a spatially selective environment determines binary functional outcomes. Five bijective correspondences between enantioselective drug action and enantioselective material behavior. The thalidomide disaster and the stereocomplex polylactide revolution are instances of the same structural principle.

This paper defines Epistemic Engineering as a field of inquiry and proposes its foundational research agenda. Epistemic Engineering is the design, implementation, comparison, and experimental validation of procedures that assign epistemic status to claims. We identify ten open problem families spanning constitutional specification, metrological frameworks, expressivity–verifiability trade-offs, negative-result infrastructure, constitutional evolution, bias measurement, deliberation topology, portability, ablation methodology, and calibration.

An epistemic constitution defines the procedures by which a deliberation system assigns epistemic status to claims. This paper introduces the concept of constitutional expressivity and asks what trade-offs govern the relationship between expressivity and verifiability. We propose a preliminary taxonomy of epistemic behavior classes: adversarial deliberation, convergent synthesis, divergent exploration, uncertainty escalation, and negative closure.

If epistemic constitutions are to become standard objects of scientific comparison, then the field requires instruments of measurement. This paper proposes the foundations of a metrology for epistemic constitutions along eight candidate measurement dimensions. We propose the blind constitutional trial as the gold standard experimental design.

This paper argues that when an isomorphism is formally investigated and found not to hold under specified conditions, the result constitutes computational infrastructure of independent value. We introduce the concept of epistemic friction: the resistance that verified negative isomorphisms impose on future exploration of closed research paths. We formalize the conditions for certified impossibility and develop the concept of epistemic risk deflation.

This paper formalizes the indeterminate category as the third status in a trivalent epistemic framework. We explore the formal analogy with trivalent logics (Łukasiewicz, Kleene) and with the implications of Gödel's incompleteness theorems for bounded formal systems. We argue that the indeterminate category has independent epistemic value: it identifies the boundaries of the system's current knowledge and signals where the constitution requires extension.

The history of science can be read as a history of instrumentation. This paper argues that the epistemic governance framework represents an attempt to instrument epistemology itself. But the framework is not the telescope; it is the astronomer. The telescope observes; the astronomer governs observation. Current AI systems fuse these roles; constitutional epistemic governance separates them. The models are interchangeable executors. The constitution is the invariant.