RESEARCH METHODOLOGY
How every claim is evaluated, structured, and rated. Applied identically to mainstream consensus and alternative hypotheses alike.
Every Claim Gets a Tier
We don't evaluate topics — we evaluate individual claims. A single document about ancient Egypt can contain Tier 1 facts (the Pyramid of Giza exists and was built ~2560 BCE) alongside Tier 3 speculation (it encodes advanced astronomical knowledge) alongside Tier 4 claims (it was built by extraterrestrials). All three appear in the same document, correctly labeled.
Peer-Reviewed Consensus
Multiple independent lines of peer-reviewed evidence. Scientific or archaeological consensus. Primary sources confirmed and reproducible.
Legitimate Academic Debate
Published by credentialed scholars. Evidence exists, but remains debated among experts. Multiple serious positions present.
Possible but Unverified
Coherent hypothesis without sufficient primary evidence. Popular-author level. Interesting but not proven. Treat with caution.
No Credible Support
No verifiable evidence, or actively contradicted by available evidence. Included for completeness, not endorsement. Flagged with [DEBUNKED] where applicable.
Key rule: Tier 4 claims are never deleted — they are labeled and retained. Understanding why something is wrong is as valuable as knowing what's right.
Not All Sources Are Equal
Every document receives a weighted bibliography score. Higher-quality sources earn more points. The score determines the [N/5] confidence rating visible on every document header.
| Source Type | Points | Examples |
|---|---|---|
| Peer-reviewed journal article | 3 pts | Nature, Science, PNAS, PLOS ONE, specialized journals |
| Academic press book / primary text | 2 pts | Oxford UP, MIT Press, Cambridge, translated ancient texts |
| Popular book / website / other | 1 pt | Bestseller non-fiction, documentary, general-audience website |
Exceptional
Strong
Moderate
Limited
Minimal
The Same Evidence, Three Framings
Ancient texts and traditions are evaluated through three non-exclusive interpretive lenses. Most phenomena can be read through more than one. We present all applicable lenses rather than declaring one correct.
Tradition as mythology, archetype, and cultural psychology. The Jungian and Campbellian reading — universal psychological patterns expressed through narrative. Typically Tier 1–2 when limited to psychological claims.
Tradition as describing real historical events, entities, or technologies. A Sumerian flood narrative as a record of an actual flood. Higher-risk interpretation — most often Tier 2–4 depending on corroborating evidence.
Tradition as encoding knowledge in allegorical or symbolic form. A serpent representing the regenerative principle in medicine. Often Tier 1–3 — the strongest cases have cross-cultural symbolic consistency confirmed by independent scholars.
Every Document Uses the Same Template
From ancient Sumerian texts to quantum physics to oceanography, every document follows an identical structure. This makes cross-referencing possible and ensures no topic gets special treatment.
- Header Block Machine-readable metadata: source count, weighted score, [N/5] confidence rating, primary tier, last updated, keywords, category tags, cross-references to related documents.
- Quick Summary 3–6 sentences capturing the topic and its evidential status. Written to be readable standalone.
- Tier 1–4 Sections Claims organized by evidence strength — Verified, Credible, Speculative, Dubious. Named scholars appear in bold. Exact dates, measurements with units, institutional attribution required.
- Counter-Arguments The strongest real, published objections to the document's primary claims. If no credible counter-argument exists (e.g., for settled consensus), an honest "no scholarly dispute" note replaces this section. No invented objections.
- Bibliography Minimum 5 entries. Chicago style. DOIs and ISBNs where available. Every entry in canonical 12-slot schema, verified against academic databases where possible.
- Cross-Reference Index Links to related documents with explanations of how they connect. 34,596+ keywords tracked across the corpus for systematic cross-referencing.
Applied Without Exception
These rules are enforced by tooling, not just intention. The pipeline checks compliance automatically on every document.
- 1Never fabricate sources. Only cite real, verifiable publications.
- 2Never delete data. Debunked claims get [DEBUNKED] tags and remain visible.
- 3Always present real counter-arguments. If none exist, say so honestly.
- 4Always apply tier ratings. No unrated claims in finalized documents.
- 5Always use source confidence. [N/5] based on weighted bibliography score.
- 6Always flag uncertainty explicitly — "UNCERTAIN" or "UNKNOWN" when warranted.
- 7Always date claims. Publication year for every cited source.
- 8Always note corrections. Updated claims are dated, not silently replaced.
- 9Never present belief as fact. Even popular claims are rated at their actual tier level.
- 10Apply the same scrutiny to claims we're inclined to believe as to claims we're inclined to reject.