Methodology: Citation Standard and Verification Protocol

How content in this repository is sourced, cited, verified, and flagged. Every new file must be born compliant with this standard; existing files are being brought into compliance by a systematic verification pass (see the per-collection verification-log.md files as they appear).

1. File format

Every content file follows the established repository format:

# Title

> Two- or three-sentence summary of the file's content.

**Tradition**: Pan-Orthodox | Russian Orthodox | ROCOR | Greek Orthodox / Athonite | (etc.)

[Body with inline citations]

## See Also

- [Related file](relative/path.md) — one-line reason for the link

## References

- Full bibliographic entries for every work cited inline

Rules:

  • Relative links only. Within a collection: ../02-section/file.md. Across collections: ../../orthodoxy/.... Never absolute filesystem paths; never bare URLs to other files in this repo.
  • Files in -ru collections link only within the Russian mirror (fall back to the English file only when no Russian twin exists).
  • Every content file must be listed in its section README’s navigation table, and every section in the collection’s master README.

2. Citation standard

2.1 Patristic and classical sources

Cite by work + internal division first (stable across editions), then the standard scholarly reference, then the translation actually consulted:

(Athanasius, *On the Incarnation* 54 [PG 25:192B]; trans. Behr, PPS 44a)
(Basil, *On the Holy Spirit* 27.66 [PG 32:188]; trans. NPNF II/8)
  • PG/PL (Migne) volume:column references are added only when verified against an actual column reference — never guessed. A citation without PG/PL is honest; a wrong one is poison.
  • Where a critical edition supersedes Migne (CCSG, GCS, SC, DOML), cite it instead or alongside.
  • Public-domain translations (NPNF I/II, ANF) are preferred as verification anchors because they are web-checkable; modern translations (SVS Popular Patristics, DOML) cited where used.

2.2 Scripture, councils, canons

  • Scripture: book chapter:verse; note LXX numbering where it differs (especially Psalms).
  • Councils: council, year, and the specific definition or canon number (e.g., Chalcedon 451, Definition; Quinisext 692, Canon 55).
  • Canons: number plus the collection consulted (e.g., The Rudder [Pedalion], trans. Cummings — noting that St. Nikodemos’ commentary is not the canon text itself).

2.3 Secondary scholarship

Author, Title, place: publisher, year, and page/chapter where the claim is specific. Wikipedia and OrthodoxWiki may be used to locate sources but never as the terminal citation. The same applies to orthodoxhistory.org and similar aggregators: follow them to the primary document and cite that.

2.4 Compiled sayings and oral tradition

For figures known mainly through compiled sayings (St. Paisios’ Spiritual Counsels, recorded counsels of elders), cite volume and section. Where the earliest written attestation is late or the transmission chain is weak, say so in the text — provenance is a topic, not an embarrassment.

3. Verification protocol

Used both for the retroactive verification pass over existing content and as a birth-checklist for new files.

  1. Inventory. List every [Verify] tag in the file, plus every date, statistic, direct quotation, and attribution — even untagged ones.
  2. Research. For each item, find at least two independent sources, at least one primary or scholarly.
  3. Disposition. Exactly one per item:
    • Confirmed — remove the tag; add or upgrade the citation.
    • Corrected — fix the text, cite the correct source, log the correction.
    • Disputed — convert the tag to a structured note naming the positions:
      [Disputed: source A gives X (cite); source B gives Y (cite)]
      Contested statistics get named scholarly ranges, never a single number.
    • Apocryphal / unattested — if a quotation cannot be located in any primary source, either delete it or keep it explicitly labeled: “widely circulated but not found in any primary source.”
      A [Verify] tag may be removed only under Confirmed or Corrected; otherwise it must be converted — never silently dropped.
  4. Patristic quotations are located in PG/PL or a named critical edition before the tag is cleared; an unlocatable quotation falls under the apocryphal rule.
  5. Hagiographical claims are never “verified” as historical fact. They are pinned to their earliest written attestation (e.g., “first recorded in the 1867 Valaam Life”). Miracle accounts are reported as what the sources attest, with the source named.
  6. Log. One line per disposition in the collection’s verification-log.md: date · file · claim · disposition · sources.

4. Handling disputed and sensitive topics

4.1 Confessional standpoint

This knowledge base is academic from within the Orthodox Church — not from a Roman Catholic, Protestant, or confessionally disinterested vantage point. The academic apparatus (citations, dispositions, attestation-pinning) governs how claims are sourced, not where the work stands. Concretely:

  • Where the Orthodox Church has a dogmatic or conciliar position against another confession — the Filioque, papal supremacy and infallibility, the Immaculate Conception (1854), purgatory as satisfaction, created grace — Orthodox teaching is presented as the teaching of the Church, and the other position as what it is: a doctrine the Church has examined and not received. The heterodox position is still reported accurately, in its own terms, with citations — accuracy toward an interlocutor is not neutrality between the positions.
  • Historical events entangled with these disputes (the Photian councils, 1054, the Fourth Crusade and the sack of 1204, Lyons 1274, Ferrara–Florence 1438–39, the Unia of Brest 1596) are narrated with full academic sourcing but assessed from the Orthodox side: Florence is a council whose union the Church rejected and whose sole non-signer is honored as a confessor, not a “reunion regrettably lapsed”; 1204 is an atrocity against the Church, whatever its strategic context.
  • Heterodox scholarship is cited freely and gratefully where it is good scholarship; using a Roman Catholic historian’s critical edition does not import his ecclesiology.

4.2 Both-sides treatment — scope

The both-sides rule exists for disputes the Church herself has not settled, and for questions of historical fact where evidence, not confession, decides. It applies to:

  • disputes within Orthodoxy (Possessor/Non-Possessor, the ecumenism debate, calendar questions, ROCOR–MP glorification categories, contested receptions of particular figures);
  • source-critical and factual questions (datings, attestations, authorship, casualty figures, relic identifications) — here the standard is strictly evidential, and Orthodox tradition’s claims are pinned to their attestations like any other (§3.5);
  • live jurisdictional disputes (e.g., the status of the Macedonian and Montenegrin questions, or the overlapping jurisdictions of the diaspora): every contested claim carries both-jurisdiction sourcing and a dated “as of” note.

Where this library’s own jurisdiction has taken a position on a live jurisdictional question, that position is stated as the library’s own rather than surveyed from outside — but it is fenced by the **Tradition**: label so the reader knows whose judgment he is reading, the opposing case is reported accurately from its own documents and at its strongest, and the dated “as of” discipline still applies. The Ukrainian autocephaly question is treated this way: the narrative of events in orthodoxy/11-world-orthodoxy/ documents both positions and is labelled Pan-Orthodox, while the ecclesiological judgment in ecclesiology/09-the-ukrainian-question/ is labelled Russian Orthodox / ROCOR. A rupture of communion between churches is never reported as though it placed a party outside the Church, unless the Church has said so conciliarly.

It does not mean neutrality between Orthodox dogma and its denials (§4.1). Further rules:

  • Distinguish dogma (conciliar definition) from theologoumenon (theological opinion) explicitly; the toll-houses treatment in seraphim-rose/04-soul-after-death/aerial-realm.md is the model.
  • Polemical framing is avoided — the confessional standpoint of §4.1 is stated soberly, never triumphally; where a collection’s perspective is tradition-specific, the **Tradition**: label says so.

5. Quality gates (run at the end of every work phase)

  1. Format compliance — title, blockquote summary, Tradition label, See Also, References present in every content file.
  2. Navigation accuracy — every file reachable from the root index via README navigation tables; no orphans.
  3. Link integrity — every relative link target exists; zero absolute filesystem links.
  4. Citation presence — every quotation and factual paragraph carries an inline citation.
  5. [Verify] hygiene — new files contain zero bare [Verify] tags (structured Disputed/attestation notes are the permitted forms); verification-pass files have all dispositions logged.
  6. EN/RU structural parity — for every mirrored pair, the file trees are identical, the disposition counts match per file, and the ## section-heading counts match per file. The heading check is the cheap one and it catches a class the others cannot see: a Russian mirror silently missing an entire section still passes format compliance, link integrity and disposition parity. A scan of all 5,967 mirrored pairs in July 2026 found six such files. Disposition keywords stay English in Russian files so that notes remain greppable across both languages.

6. Caveat

The collections were generated with AI assistance. This methodology exists precisely because of that: it makes the confidence level of every claim legible. Nothing here is a substitute for catechesis under a priest or for consulting the primary sources directly.