Papers

The Papers page gathers formal and semi-formal work behind the Credencing framework.

Related Research Papers

The paper cluster behind Credencing

These Academia.edu papers support the public site as a connected research program: formal modeling, two-gap diagnosis, warranted uncertainty, institutional credencing, AI-mediated inquiry, scalar language, Bayesian audit methods, and motivated-cognition diagnosis.

Foundational modelA Formal Framework for Core and Deep RationalityDefines Objective Evidence (E0), Perceived Evidence (EP), Assigned Credence (CA), Deep Rationality (SD), Calculation Error (DE), and Core Irrationality (IC).Two-gap modelThe Two-Gap Theory of Irrationality: Evidence Perception, Belief Integrity, and Diagnostic RepairSeparates E0-EP evidence-perception failures from EP-CA belief-integrity failures, giving the framework its main diagnostic repair map.Confidence disciplineWarranted Uncertainty: The Discipline of Not Knowing in a Credence-First EpistemologyDefends confidence bands, warranted slack, and honest uncertainty so credencing does not collapse into false precision.Biased expert diagnosisIntelligence as Rationalization Engine: Deep Rationality, Core Rationality, and the Biased ExpertExplains how high Deep Rationality can reduce evidence-processing error while still defending an assigned credence the agent is motivated to keep.Institutional extensionInstitutional Credencing: Official Confidence, Evidence Pipelines, and Organizational Belief IntegrityExtends E0, EP, CA, and SD to organizations through evidence pipelines, public confidence, dissent channels, and update trails.AI-mediated credencingAI as Deep Rationality Prosthetic: Evidence Processing, Delegated Credence, and Core Rationality in AI-Mediated InquiryShows how AI can improve evidence processing while leaving final credence vulnerable to motivated uptake, false precision, and delegation errors.Conceptual bridgeCore RationalityIntroduces core rationality as evidence-proportionate credence, distinct from mastery of formal probability, logic, and decision theory.Credence-first entry pointEvidence-Proportioned Belief: A Credence-First Reframing of Faith and RationalityFrames belief as confidence scaled to evidence rather than a binary status or performance of conviction.Credence-first epistemologyCredence First: Against Plantinga's Warrant as an Epistemic UpgradeArgues that graded credencing is a better epistemic target than binary knowledge or warrant, especially when certainty can become miscalibrated.AI over-delegation warningHumanity as a Middle Child in the Age of AI: Patterns of Delegation and Epistemic AtrophyWarns that AI can weaken human judgment when people outsource inference without retaining responsibility for credence.Language and calibrationThe Gravity of Grammar: Binary Inertia and the Distortion of Epistemic CalibrationExplains how ordinary assertion can compress scalar confidence into binary belief-talk, distorting epistemic self-reporting.Linguistic repairBeyond the Binary: Linguistic Evolution and the Transition to Credence-Based DiscourseDevelops a scalar lexicon, Bayesian update protocol, and discussion norms for keeping belief revision explicit and socially usable.Applied Bayesian auditThe Miracle Audit: A Probabilistic Framework for Differentiating Anomalies from Supernatural InterventionsApplies probabilistic discipline to extraordinary claims by separating anomaly, causal link, evidential rigor, and burden of proof.Motivated cognitionThe Architecture of Explanatory Satisfaction and the Affective Veto Against Stochastic RealityConnects explanatory hunger, agency bias, and affective resistance to stochastic explanations with the site's Core Rationality (IC) diagnosis.Operator knowledgeFoundational Choices for the AutodidactDefends transferable reasoning operators: base rates, causal inference, measurement discipline, calibration, and feedback-rich practice.Methodological naturalismThe Asymptote of the Natural: Inductive Density and the A Posteriori Justification of Methodological NaturalismFrames naturalistic inquiry as a Bayesian and risk-management strategy grounded in projectibility, constraint, and historical evidential yield.Ontology and inductionThe Threshold of the Real: Inductive Density and the Emergence of the Human OnticExplains objecthood and category confidence as a function of predictive reliability and cognitive cost.Reification mechanismThe "Ontic Snap" as a Cognitive Phase TransitionModels the moment noisy input becomes a stable “thing,” helping diagnose premature closure and overconfident categories.Category humilityMereological Indigestion: Why We Cannot Divvy Up the BlobChallenges overconfident partitions of reality and supports the site’s caution about reifying useful categories.Pedagogical scaffoldingUseful Fictions Common in Philosophical PedagogyClarifies when simplified teaching models help inquiry and when they become overconfident substitutes for mature understanding.Attribution calibrationAgainst the Great Man: A Systems-Theoretic Critique of Individual AttributionUses a control coefficient to calibrate praise and blame in noisy systems where individual agency may be weak.

View full Academia.edu profile

Purpose

The papers library is the bridge between the public-facing model and the more formal argument behind it. It gives the project a place for definitions, diagrams, objections, replies, revisions, and extensions that would overload the introductory pages.

The public site explains the framework in practical language. The paper trail makes the commitments precise enough to criticize: what the variables mean, what the gaps measure, what the model claims, and where it might overreach.

Current Paper Spine

The central paper is “A Formal Framework for Core and Deep Rationality.” It supplies the basic architecture: objective evidence (E0), perceived evidence (EP), assigned credence (CA), deep rationality (SD), calculation error, core irrationality, and the archetypes generated by different gap patterns.

The new research spine develops that architecture in five directions. “The Two-Gap Theory of Irrationality” makes the diagnostic split explicit. “Warranted Uncertainty” protects the model from false precision. “Intelligence as Rationalization Engine” explains the biased expert. “Institutional Credencing” scales the model to organizations. “AI as Deep Rationality Prosthetic” applies the same distinction to AI-mediated inquiry.

The older supporting cluster remains narrower than before: credence-first epistemology, scalar language, Bayesian audit methods, explanatory hunger, operator learning, methodological naturalism, ontological commitment, and calibrated attribution. Together, these papers treat Credencing as a program for replacing all-or-nothing belief language with calibrated, updateable confidence.

Relation to the Books Shelf

The Books shelf now carries the larger foundations: induction as the center of bounded inquiry, Bayesian bookkeeping as disciplined credence tracking, and ontology as stable but revisable commitment. The Papers shelf should be read as the shorter formal and applied cluster around that larger background.

How the Papers Align With the Site

The public site should be read as the teaching and navigation layer for the papers. The formal pages explain the E0-EP-CA model; the Bayes theorem pages explain how confidence should move; the skills and discussion pages train the operators needed for disciplined updating; and the case studies show what happens when identity, emotion, language, or institutional pressure distorts the update.

The language of the site is intentionally scalar. Claims such as “belief,” “confidence,” “evidence,” and “rationality” should be interpreted as graded and revisable unless a page is discussing a practical action threshold. This keeps the site aligned with the binary-inertia papers, which warn that ordinary grammar can make partial confidence sound like final possession of truth.

The model also keeps competence and integrity separate. Deep Rationality (SD) concerns operator skill: probability, statistics, causal reasoning, measurement, and formal modeling. Core Rationality concerns whether assigned credence remains responsive to perceived evidence. This distinction is the through-line from the Core Rationality paper to the interactive archetypes.

Objections to Develop

The papers section is also where the strongest objections belong. Important objections include whether E0 can be estimated without pretending to have God’s-eye access, whether EP can be separated from CA in messy human cases, whether “core irrationality” overclassifies hard disagreement, and whether Bayesian language can be made practical without false precision.

The newer alignment with the paper cluster adds further objections worth developing: whether scalar language can remain usable in ordinary conversation, whether AI tools can improve calibration without encouraging dependence, whether institutional audits can distinguish internal evidence from official confidence, and whether Bayesian audit rubrics smuggle in contested priors.

Each objection should be treated as a stress test rather than a threat. If an objection forces a cleaner definition, a humbler claim, or a better diagnostic distinction, it improves the project.

How to Read

Read the public Start Here and Core Ideas sections first. Then use the papers to inspect the formal structure, not as a replacement for the interactive model but as its theoretical support.

A good reading order is: “Core Rationality,” the formal framework, the two-gap paper, “Warranted Uncertainty,” “Intelligence as Rationalization Engine,” “Institutional Credencing,” and “AI as Deep Rationality Prosthetic.” After that, move outward to the credence-first, binary-language, Bayesian audit, methodological naturalism, operator-learning, and affective-veto papers. For the deeper background, move sideways to the Books shelf.