The full methodology behind Divinity Atlas: how it separates what sources establish from what a tradition holds, how it sources and audits every claim, and an open invitation for scholars to test it.

Draft 1.1, August 17, 2026. Licensed under a Creative Commons Attribution 4.0 International license (CC BY 4.0). Cite as: Leite, C. (2026). Divinity Atlas Methodology White Paper (Draft 1.1). Divinity Atlas.

Download the paper as a PDF or as a Word document.

Abstract

Divinity Atlas is a public reference system for religions, mythologies, philosophies, and esoteric traditions. Its central methodological problem is not merely how to store many claims, but how to represent claims that are differently established, differently held, disputed, culturally restricted, or genuinely unresolved. The system separates the evidentiary status of a claim from the stance a tradition takes toward it; treats sources, facts, relationships, uncertainty, and dissent as structured records; and subjects changes to auditable write procedures and executable quality gates. This paper describes that framework, its operational controls, and its public challenge and case process. It also states the framework's present limitations. Divinity Atlas has not yet undergone broad external academic validation, and internal second-pass review is not presented as equivalent to scholarly peer review. The system is offered as a transparent, testable methodology for public knowledge representation and as an invitation to scholars, librarians, and digital-humanities researchers to evaluate and improve it.

How to Respond

Three ways to take part: challenge a specific claim, read the public case docket, or volunteer to review.

Every claim on the atlas can be challenged in place, on the page where it appears. A signed-in reader's own open and resolved challenges, and the appeals raised from them, are gathered on the reader challenges page; an appeal that names a specific, checkable source and what should be verified can become a public case. Every case, whether still open or already decided, is public on the case docket, along with the evidence examined, the reasoning behind the judgment, and any dissent recorded against it.

Scholars who want to review under the protocol below can sign up at the reviewer sign-up page.

Scholars who have completed a review and agreed to be named for it are acknowledged on the reviewers page.

For anything else, including a question this page does not answer, write to us through the contact page.

What We Ask Reviewers to Do

The full protocol is Appendix B of the paper. In plain language:

• The review unit. One entry, or one specific claim or relationship inside it, read with its full context: its citations, its status and stance labels, its notes, and the version it was reviewed at, never a bare row pulled out of a database.
• Seven dimensions. A reviewer is asked to judge identity (is the subject correctly identified), accuracy (does the claim fairly state what the evidence supports), entailment (does each citation actually support the specific claim made), status (is attested, debated, debunked or unknown the right call), stance (is a tradition's own position represented without treating some other tradition as the default), framing (is the account fair, proportionate and free of avoidable outsider bias), and restriction (does the treatment respect the rights and wishes of the community involved).
• Sampling. Reviewers work from a stratified sample that deliberately oversamples debated, debunked, unknown, restricted, living-tradition and lower-tier-source material, not a set of showcase entries chosen to look their best.
• Dissent preserved. A reviewer who disagrees with the final disposition may publish that disagreement rather than have it quietly absorbed into a single verdict.
• A remediation ledger. Every accepted finding is linked to what it actually produced, a case, a correction, an open question, or a documented decision that nothing needed to change, so a reviewer's work stays traceable after the fact.