The literature

This page lists every source the article rests on, and every source worked up for it and not used, with the reason.  The research files are in research/:  lead-verified.md (everything the lead read in the bytes, with page anchors), three helper dossiers (ancient-testimonies.md, secretaries-and-letters.md, stylometry-literature.md, leads only), and four Deno scripts with their outputs (verses.ts, shared-runs.ts, word-counts.ts, stylometry-validation.ts).

Direction is judged by what a source shows on the article's two main claims:  that New Testament citation practice fits documented ancient practice, and that ancient readers did not accept false authorship.  supports means it supports the traditional or defender's reading of the passage in question, undercuts that it supports the skeptic's, complicates that it cuts both ways or limits both.  Verified is full when every cited passage was read in the source, partial when read by targeted search-inside on a print-disabled scan with anchored page numbers, citation when known only through another work.

Primary texts

morphgnt-2026, eliranwong-2017, openscriptures-2026b Greek New Testament, Septuagint and Hebrew Bible data

method: text data direction: background verified: full SBLGNT via MorphGNT, Rahlfs's Septuagint word list and the Westminster Leningrad Codex (morphhb, commit 3d15126).  Every verse comparison in the article was printed from these files by research/verses.ts;  cited collectively in the note [^note-texts].  The SBLGNT apparatus and Robinson-Pierpont CSV (sblgnt-apparatus, robinson-2005c) supplied the Mark 1:2 variant.

martinez-1999 García Martínez and Tigchelaar, The Dead Sea Scrolls Study Edition (1999)

method: edition direction: supports verified: full 1QpHab VII and XII (pp. 17, 21), II (p. 13), 1QS VIII (p. 89), 4Q174 (pp. 353-355), 4Q175 (p. 357).  Shows pre-Christian Jewish quotation with no source names, a compatible book name over partly restored wording, and pesher with an explicit theory.  The 4Q174 Ezekiel wording partly rests on restorations.

tov-2001b Emanuel Tov, Textual Criticism of the Hebrew Bible, 2nd rev. ed. (2001)

method: handbook direction: supports verified: full Pages 88 and 99:  the Samaritan Pentateuch, 4Q158 and 4QTestimonia carry Deuteronomy 18:18-22 inside Exodus 20, and 4QTestimonia's sequence follows that text form.  Shows that an apparent combination may follow a harmonised scriptural sequence already in circulation;  it does not prove exact identity with a lost exemplar.

jull-1995 Jull et al., radiocarbon dating of scrolls (1995)

method: radiocarbon measurement direction: background verified: full Table 2 (p. 14) and Appendix 1 (p. 18):  handwriting and radiocarbon dates for 1QpHab and 1QS, used only to place the two scrolls before Jesus.

metzger-1971 Bruce Metzger, A Textual Commentary on the Greek New Testament (1971, first edition)

method: textual commentary direction: complicates verified: full Mark 1:2 (p. 73), Matt 27:9 (pp. 66-67), Matt 13:35 (p. 33), Mark 2:26 (p. 79).  Documents copyists' changes to all three hard names, and scribes inserting "the best known prophet."  Edition trap:  the library text filed as metzger-1994 is this 1971 edition (title page and copyright line);  a new YAML records it.  Sigla are OCR-garbled and are not quoted.

jerome-57 Jerome, Letter 57 to Pammachius (395; NPNF)

method: letter direction: complicates verified: full Sections 7 and 9.  States the ancient principle that apostles cared for sense rather than "words and syllables", and leaves Mark 1:2 and 2:26 unsolved.  Stored HTML now beside the existing record.

augustine-1888 Augustine, De consensu evangelistarum 3.7.29-31 (NPNF)

method: harmonisation treatise direction: complicates verified: full Weighs manuscripts without "Jeremiah" and rejects them on the harder-reading principle, considers a slip of memory, argues providence, and proposes the Jeremiah-field composite about AD 400.  An editor's bracketed note sits inside §30.

philo-1854d Philo, On Mating with the Preliminary Studies 9-11 (Yonge 1854)

method: allegorical commentary direction: supports verified: full A Jewish contemporary of Paul allegorises Hagar and Sarah, to a different end from Galatians 4.

eusebius-1890b, eusebius-1890c, mcgiffert-1890d Eusebius, Church History books 3, 6 and 7 (NPNF)

method: history quoting earlier writers direction: complicates verified: full Serapion (6.12), Clement on Hebrews (6.14), Origen on Hebrews and 2 Peter (6.25), Dionysius on Revelation (7.25), Eusebius's own classification (3.3, 3.25).  The Greek of 6.12.3 was checked in eusebius-2016b (TEI, NFD Unicode).  Considered and not used as a separate source in the main article because the McGiffert translation already supports the same claim;  the Greek is quoted in quotes.md.  mcgiffert-1890d has no recorded download source in its YAML;  its wording matches the standard McGiffert text.

jerome-1892, jerome-migne-de-viris-illustribus-latin Jerome, De viris illustribus 1 and 5

method: catalogue direction: complicates verified: full Doubts about 2 Peter "on account of its difference from the first in style";  Hebrews not Paul's "propter styli sermonisque dissonantiam".

hilberg-1912 Jerome, Ep. 120.11 (CSEL 55, pp. 507-508)

method: letter direction: supports verified: full Jerome's secretary theory:  Titus as Paul's interpreter, Mark as Peter's, and different interpreters for 1 and 2 Peter.  English is the lead's;  no public-domain translation found.

pauly-1883 Salvian, Ep. 9 (CSEL 8, pp. 217-223)

method: letter direction: undercuts verified: full The only ancient Christian defence of an apostolic pseudonym found.  Argues the books are obviously recent, that readers judge by the author's name, and that the name is a symbolic dedication, because a plain false name would be a lie.  "Timuit" is Baluze's emendation of the single manuscript's "timothei".  English is the lead's.

wilder-2004 Terry Wilder, Pseudonymity, the New Testament, and Deception (2004)

method: survey of reception direction: undercuts verified: partial Salvian date (p. 194) and conclusions (pp. 217, 220, 253).  Conservative, and agrees with Ehrman that the early church rejected detected pseudepigrapha.  Read by search-inside, printed page = leaf - 14 from scandata.

taylor-1818 Iamblichus, Life of Pythagoras, trans. Thomas Taylor (1818)

method: biography direction: complicates verified: full VP 198 praises Pythagoreans for crediting their work to Pythagoras;  the proem condemns "false and spurious writings".  The Greek (Deubner 1937, deubner-1937, copyright, not stored) was read only by the helper and is not quoted.

holmes-1885b Tertullian, Against Marcion 5.21 (ANF)

method: polemic direction: supports verified: full Marcion "rejected the two epistles to Timothy and the one to Titus";  Tertullian treats them as Paul's.

evans-1964 Tertullian, De baptismo, ed. and trans. Ernest Evans (1964)

method: critical edition direction: undercuts verified: full 17.5:  the presbyter who composed the Acts of Paul "for love of Paul" was deposed.  The apparatus shows "Acta" in Codex Trecensis and missing from Mesnart 1545.  This corrects the prophecy-and-apocalyptic-as-genres dossier, which says the text does not name the Acts of Paul.

tertullian-1885d Tertullian, On Baptism (ANF, Thelwall)

method: translation direction: undercuts verified: full Considered, not used in the article:  its English follows the edition without "Acta" ("the writings which wrongly go under Paul's name"), so Evans's Latin and English are cited instead.  Quoted in quotes.md for comparison.

tregelles-1867, muratorian-ecw-metzger-translation The Muratorian fragment

method: canon list direction: undercuts verified: full Tregelles p. 47 for the Latin "Pauli nomine fincte ad heresem Marcionis";  Metzger's line-numbered English for the Pastorals (lines 59-63) and the forged letters (64-67).  The OCR is now stored beside Tregelles's record.  muratorian-ecw-unattributed-translation was read for comparison only and not used:  Metzger's numbered translation is more checkable.

kuhn-1830, muller-1891b Galen, On My Own Books

method: autobibliography direction: undercuts verified: full Kühn's facing Latin (p. 9) and Müller's Teubner Greek (pp. 91-92):  the reader who rejected a book as "falsely titled" after two lines;  Galen's complaints about others reading his untitled notes as their own and altering his books.  The Greek OCR is damaged and is printed normalised, with that caveat.  Müller's text was found and stored this session.

rolfe-1914f Suetonius, Divus Iulius 56.1 (Rolfe)

method: biography direction: background verified: full Records that the authorship of the Alexandrian, African and Spanish wars was uncertain, "Oppius" or "Hirtius";  the control for the Caesar stylometry study.

perseus-2026 Cicero, Letters to Atticus (Latin, Perseus TEI, commit 0700a6b)

method: letters direction: supports verified: full 3.15.8, 11.2.4, 11.3.3, 11.5.3:  letters composed by Atticus "in my name", with an excuse for missing seal or hand;  13.25.3 dictation to Tiro and Spintharus.  English from Shuckburgh (cicero-1899, Gutenberg vol. 1;  cicero-1900, archive.org vol. 3), named in the source note, cited by letter number.

first1kgreek-2026 Ignatius and Polycarp (Lake's Greek, TEI)

method: letters direction: complicates verified: full Ignatius Phld. 11.2 and Smyrn. 12.1 ("I write through Burrhus", a carrier);  Polycarp Phil. 4.1 (echo of 1 Tim 6:7, 10) and 13.2 (forwarding Ignatius's letters).  Polycarp Phil. 14 ("per Crescentem") survives only in Latin and was not used.

cicero-ad-familiares-16-perseus-latin Cicero, Ad Familiares 16.22 (helper)

method: letter direction: complicates verified: citation Considered, not used:  shows copyists writing Cicero's text and his own inserted passage, but the key English ("I don't write without an amanuensis") is Shuckburgh's gloss of "ipse nihil scribo", and the Atticus letters make the point more directly.

Modern studies of quotation and interpretation

gundry-1967 Robert Gundry, The Use of the Old Testament in St. Matthew's Gospel (1967)

method: text-form study direction: supports verified: partial Pages 123-127 on Matt 27:9-10:  the Zechariah-Jeremiah 19 composite, the 2 Chronicles 36:21 parallel, the rabbinic claim (on a nineteenth-century authority) and ten other explanations.  Printed page = leaf - 20 (scandata, confirmed by his own cross-reference).

wallace-2006 Daniel Wallace, "Mark 2:26 and the Problem of Abiathar" (2006)

method: survey direction: complicates verified: full Five kinds of solution with fair statement of each, Lane's objections to the "passage" reading, the grammar of "in the days of", Morgan's check of the Greek manuscripts, and Wallace's preference.  Saved from the 2010 Wayback capture (bible.org blocks curl).  Gundry's later midrashic view and Casey's error view are quoted in it;  neither was opened directly, so neither is quoted in the article.

ehrman-2005 Bart Ehrman, Misquoting Jesus (2005)

method: popular textual criticism direction: undercuts verified: full Page 9:  his teacher's comment on Mark 2:26.

kaiser-1985 Walter Kaiser, The Uses of the Old Testament in the New (1985)

method: exegesis direction: supports verified: partial Pages 47-52 on Hosea 11:1:  Matthew follows the Hebrew;  corporate solidarity in "my son";  the Lindars link to Num 24:7-8 LXX (reported, not used).  Printed page = leaf - 20.

enns-2005 Peter Enns, Inspiration and Incarnation (2005)

method: hermeneutics direction: complicates verified: partial Pages 133-134, 152-155:  Hosea 11 looks back;  Matthew uses Second Temple methods and reads from the resurrection.  His claim about Jewish readers' reaction is noted as untested.  Printed page = leaf - 2.

turpie-1868 D. M. Turpie, The Old Testament in the New (1868): NT agreement with LXX and MT

method: tabulation direction: complicates verified: citation Considered, not used directly:  the counts belong to the linked article Which Bible Text Gives the Right Date for Creation?, whose memory records the verified figures and the "about a third" misreading to avoid.

Authorship, pseudonymity and secretaries

ehrman-2011 Bart Ehrman, Forged (2011)

method: popular argument direction: undercuts verified: partial Pages 19, 23-24, 36, 39, 137-138, 220-221 and 271 n. 27.  His own line between anonymous books with later false attributions and forgeries;  condemnation of forgery;  the secretary objection.  Offset trap:  printed page = leaf - 12 (scandata and four index entries), not leaf - 11 as the library YAML says;  pages printed elsewhere from leaf - 11 are one too high.

ehrman-2013 Bart Ehrman, Forgery and Counter-Forgery (2013)

method: monograph direction: undercuts verified: citation Considered, not used:  the technical version of Forged, read for a sibling article from a third-party upload whose text was deleted;  the 2011 book makes every point used here and was re-read for this article.

bauckham-1990 Richard Bauckham, Jude, 2 Peter (Word Biblical Themes, 1990)

method: thematic commentary direction: complicates verified: partial Page 44 ("deliberately transparent fiction") and p. 84 ("the fathers" as the first generation).  Edition trap:  the archive.org item recorded as bauckham-1983 (WBC 50) is this 1990 volume;  a new YAML records it and a note was appended to the 1983 record.  The 1983 commentary itself was not read.

guthrie-1988 Donald Guthrie, review of Meade, Pseudonymity and Canon (Themelios 1988)

method: review direction: undercuts verified: full Reports Meade's thesis that attribution asserted authoritative tradition and Meade's concession of deception.

meade-1986 David Meade, Pseudonymity and Canon (1986)

method: monograph direction: supports verified: citation Considered, not used directly:  the only archive copy is mis-scanned (its text is another book), so his view is reported through Guthrie.

richards-2000 E. Randolph Richards, "Silvanus Was Not Peter's Secretary" (JETS 2000)

method: philological survey direction: complicates verified: full Pages 418, 427 n. 45, 432:  the "write through X" formula names the carrier;  secretarial help should still be assumed for NT letters.  Downloaded and stored by the helper.

richards-1991 E. Randolph Richards, The Secretary in the Letters of Paul (1991)

method: monograph direction: complicates verified: citation Considered, not used:  print-disabled;  the helper read only a few passages and the lead did not verify them.  Its classes of secretarial role are reported in the article only through Ehrman's critique.

deissmann-1910 Adolf Deissmann, Light from the Ancient East (1910)

method: papyrology direction: complicates verified: citation Considered, not used:  the helper found the Mystarion letter with a closing in a second hand (pp. 170-172), but the lead did not verify the page, and the NT's own autograph greetings make the point.

Papyrus letters with illiteracy formulas (P.Lond, P.Oslo) and the five papyri in Richards 2000

method: documentary papyri direction: complicates verified: citation Not obtained:  papyri.info answered with a bot challenge, which the helper did not bypass.  Richards's own reading of the five papyri is reported in the helper dossier only.

Stylometry

eder-2010 Maciej Eder, "Does Size Matter?" (DH2010)

method: validation experiment direction: complicates verified: full Cited for the conference version's sample-size tests.  Its corpus-specific figures must be read with Eder 2015's fuller discussion and the shorter-slice verification study by Kestemont et al.;  they are not a universal ban on short-text authorship analysis.

eder-2017 Maciej Eder, "Short samples in authorship attribution" (DH2017)

method: validation experiment direction: complicates verified: citation Considered, not used:  reports under 2,000 words only for texts "that exhibit a clear authorial signal" in English and Polish novels;  the lead did not re-check it, and the 2010 thresholds with their stated limits suffice.

eder-2015 Maciej Eder, "Does size matter?" (DSH 30.2, 2015)

method: validation experiment direction: complicates verified: full Cited in the main article after parent retrieval of the author-hosted preprint linked from https://maciejeder.org/publications/.  The journal site is closed, but the author copy is available.  The abstract, methods and sections 8.2-9 are read directly.  The fuller discussion distinguishes stable corpus-wide attribution from shorter successful individual cases, including a 209-word poem reported from earlier research.  The manuscript title says "short samples";  the published title says "small samples".

kestemont-2016 Kestemont et al., "Authenticating the Writings of Julius Caesar" (2016)

method: authorship verification direction: complicates verified: full A validated success on 1,000-word Latin slices:  Hirtius wrote Book 8 of the Gallic War and the rest of the Alexandrian War.  Author manuscript, own pagination.

stover-2016 Stover and Kestemont, "The Authorship of the Historia Augusta" (2016)

method: authorship verification direction: complicates verified: full Sentence length is invalid for ancient texts (pp. 1-2);  three studies reached three answers;  the six-author structure is "not compatible" with the text (p. 24).

morton-1966 A. Q. Morton and J. McLeman, Paul, the Man and the Myth (1966)

method: sentence-length stylometry direction: complicates verified: partial Page 90:  excluded 2 Thessalonians, Titus and Philemon as too short.  Printed page = leaf - 6.

neumann-1990 Kenneth Neumann, The Authenticity of the Pauline Epistles (1990)

method: discriminant analysis direction: complicates verified: partial Pages 194-195 and 218:  disputed letters classed Pauline but distant;  a Galatians control misclassified;  Pastorals and 1 Peter classed non-Pauline;  attribution decides only when all candidates are known.  Printed page = leaf - 14.

vannes-2018 Jermo van Nes, "Hapax Legomena in Disputed Pauline Letters" (ZNW 2018)

method: length-controlled regression direction: supports verified: full Only 1 and 2 Timothy exceed Paul's predicted rate of once-only words;  Titus, 2 Thessalonians, Colossians and Ephesians fit.  Author's copy with journal pagination.

harrison-1921 P. N. Harrison, The Problem of the Pastoral Epistles (1921)

method: vocabulary counts direction: undercuts verified: full Page 20:  306 of 848 non-proper words not in the ten Paulines;  175 Pastoral hapax legomena.  Counts not length-controlled.

libby-2016 James Libby, "The Pauline Canon Sung in a Linguistic Key" (BAGL 2016)

method: correspondence analysis direction: complicates verified: citation Considered, not used:  a visualisation of text proximity, not an attribution test;  its only relevant findings (the Thessalonian letters lie farthest from the rest;  Titus pairs with 1 Timothy) fit genre as easily as author.  Not re-checked by the lead.

faigenbaum-golovin-2025 Faigenbaum-Golovin et al., word-frequency attribution in the Hebrew Bible (PLoS ONE 2025)

method: attribution with cross-validation direction: complicates verified: citation Considered, not used:  validated on Hebrew source blocks of the Enneateuch, not Greek letters;  outside the article's question.

Not obtained

Burrows 2002 (LLC 17), Savoy 2019 (JASIST 70), Mealand 1996 (JSNT 18.59), Ledger 1995 (LLC 10), Kenny 1986 (A Stylometric Study of the New Testament), Stamatatos 2009, Juola 2008, Kroonenberg 2021.  Routes tried by the helper are in research/stylometry-literature.md.  The article's own validation test ([^note-stylometry]) uses an explicitly documented Delta-style implementation rather than claiming validation from the method's name;  Kenny's finding that only Titus was anomalous is not reported because it was not read.

Other sources worked up and not used

Dionysius of Halicarnassus, On Dinarchus

Read only in fragments by the helper (public-domain Greek OCR too poor);  not used.  It would be a second ancient example of sorting genuine from spurious works by style and chronology.

adams-ehorn Composite Citations in Antiquity (Bloomsbury T&T Clark)

The full collection is not obtained.  The editors' introduction is retrieved and cited as adams-2016 below;  Adams's Homer chapter is retrieved and considered separately.  The remaining chapters may contain additional comparanda for source names.

adams-2016 Adams and Ehorn, "What Is a Composite Citation? An Introduction" (2016)

method: definition and research survey direction: complicates verified: full Cited in the main article for the definition on p. 4 and distinctions among allusion, source lists, memory explanations, earlier collections and intentional combination (pp. 8-13).  The Glasgow accepted manuscript is stored with its printed chapter pagination.  This is not a verified example of naming the more important of two authors.

adams-2016b Adams, "Greek Education and Composite Citations of Homer" (2016)

method: classical citation comparison direction: complicates verified: partial Selected examples and the chapter context were read from the Glasgow accepted manuscript.  Considered and not used in the main article:  the inspected Homer examples combine passages from one attributed author, so they add no direct test of the disputed rule for choosing among different source authors.  The chapter is saved with YAML for further research.