Expositio in quatuor Evangelia

Merged from the four translation stints' files (cruces-0000, -0005, -0010, -0014) after the whole apparatus had been read at Migne's plate. The stints' own sections are preserved below, verbatim and unabridged, because their reasoning is the record of how each reading was reached. Where a stint's attribution CLASS disagrees with the plate, the plate governs, and §0 says so site by site. Nothing below §0 has been silently altered.

Martianay's own Admonitio (chunk 0000) warns that the text is "teeming with barbarisms and solecisms" and that he handed it to the printers "with almost all its faults" and deliberately did not correct it. Everything here is read against that declaration: an ungrammatical construction in this work is the expected state of the text, not by itself a plate defect.

§0. THE PLATE READ — what it settled, and what it overturned

Source. On-disk PL 114 scan raw/scans/pl114/patrologiaecurs06saingoog.pdf, 200 dpi full page width (one 400 dpi crop at 0897C where a word division was at stake). Page map PDF page = (column + 11) / 2, calibrated against the printed corner numbers at p.436 (861/862) before the first read and re-confirmed on every page. 25 pages / 36 columns, recorded in data/plate-reads.json.

The result: 73 markers read, 65 confirmed, 8 withdrawn

The split is entirely by marker class, and every withdrawal is a single broken WORD. Not one quoted clause was ours. This is the 11535/11550 asymmetry reproduced on a third work and at full strength: suspicion of a word is suspicion of our file; suspicion of a clause is evidence about Migne.

⭐⭐⭐ AND THE HARD FINDING: THE STINTS' OWN ATTRIBUTION GUESSES WERE WRONG TWO TIMES IN THREE

Stints 3 and 4 classed fourteen readings as probably our-file. The plate agreed with five and refuted nine:

⛔ And the three withdrawals in stint 2's range — 0879D synanoga, 0885B Eccelsiae, 0889C Judaeroum — were fired by the one stint that made no attribution claim at all. So across the work the guesses and the truth are close to uncorrelated.

⚑ What failed is identifiable, and it is the corpus-as-control argument. Every wrong guess rested on internal evidence of exactly the kind that reads as decisive: "the correct form prints eleven words later in the same sentence" (0899C — which happened to be right), "a non-word, and the sentence is an etymology of the very word" (0898B), "a contracted or truncated arbores" (0903B), "a stray initial capital fused to the adjective" (0901D), "a fusion at a line break" (0897C). These are the same moves reference_plate-read-triage already rules out — frequency and internal corroboration prove that an error EXISTS, never whose it is — and this work puts a number on the cost of forgetting it: 9 of 14.

⚑ A fused word is NOT automatically our artifact. 0897C is the worked case: Luke 3:5 prints et erunt prava indirecta, and the stint classed the fusion as ours on the strength of its being a fusion. Read again at 400 dpi, the plate genuinely sets it fused, with normal word spacing either side (compare aspera in vias planas two lines below). The marker was right; only the class was wrong.

Corrections applied to the text as a result

• 8 [sic: …] withdrawn, the Latin repaired in data/tei-patches/9004.json, and the English re-rendered with the plain word: 0879D synanoga→synagoga · 0885B Eccelsiae→Ecclesiae · 0889C Judaeroum→Judaeorum · 0899C malta→multa · 0903C an fora→anfora · 0906B ine→me · 0908C Abranam→Abraham · 0912A mimicos→inimicos. The work was re-chunked with the patches and no chunk boundary moved.
• ⚑ Two of those eight are the same letter-pair in OPPOSITE directions inside one work — 0906B (m read as in) and 0912A (in fused to m). That mirror is itself evidence the shape belongs to the transcription and not to the compositor, and it is the same argument the 11550 eum/cum pair made.
• 2 [var: …] ADDED, both divergences a stint had logged unmarked precisely because it wanted the plate first, and both confirmed as Migne's: 0879B lignum fumigans for the Vulgate's linum ("smoking flax") and 0885A spe circumdedit for sepem — where, as the stint predicted, the gloss "the guard of angels" fits a hedge and not a hope.

Two things the plate confirmed that were NOT defects of ours

• 0891B, the Pattern 7a site. The plate really does print ut ait Propheta frater non redemit, id est Moyses per legem redemit — the gloss denying what its own lemma asserts, in a passage arguing that no one before Christ could redeem. Carried as printed; the conjecture (a dropped non) stays in the apparatus and out of the translation.
• 0890B, the printed lacuna. Migne himself sets in pristino.. . . . . with the dots. Correctly left unmarked: nothing was lost by the digitization, so no [ed: …].

⚠ And one thing the plate ADDED to what we know about Migne's apparatus

0865A prints an inline bracketed conjecture — [ Forte, priusquam] — and our TEI carries it. The standing finding (CLAUDE.md) is that Corpus Corporum contains none of Migne's foot-of-page Forte notes, measured at zero across all 5,276 files. That remains true of the foot-of-page layer; it is not true of a conjecture Migne sets inline in the running text. A Forte grep over the TEI is therefore not a guaranteed zero, and this work is the counter-example.

§00. WORK-WIDE CONVENTIONS — reconciled at the merge

Four stints translated this work and diverged on three load-bearing words. The drift was measured across all 18 chunks and reconciled in the direction recorded here. A later sweep should not churn these.

1. gentes → "the gentiles", lowercase, in the commentator's voice. This work's whole argument is the Judaei / gentes typology, so three different English words for one Latin word would show a reader three categories where Migne has one — the 11551 epistylium failure in a milder form. The corpus offered no convention to inherit (8945 leans nations, 8996 and 9003 lean Gentiles), so it is settled here for this work. ⚑ Checked first: there is no natio anywhere in this work — every "nations" was rendering gentes — so the flattening erases no distinction. 112 gloss-side sites normalized; the one lemma-side site kept its traditional wording (In thy seed shall all nations be blessed, Gen 22:18 at 0863B), under the scripture-echo licence, which covers quoted words and stops where the quotation stops.

2. The commentary's METALANGUAGE takes the plain present, not the -eth form. Stint 4 archaized the commentator's own voice throughout (168 gloss-side -eth forms against 4 inside lemmata), which translation-style.md forbids: the echo licence "never licenses archaism in the author's own voice." 88 sites normalized — showeth→shows (80), signifieth→signifies (6), figureth→figures (2) — these three being the verbs by which Walafrid speaks about the text (ostendit 87×, significat 8×, figurat 8× in that stint's Latin), and never scripture. ⚑ The other -eth forms were deliberately LEFT: they sit in scripture Walafrid quotes inside the gloss without italicising it (he that cleaveth to the Lord, 1 Cor 6:17 at 0905D), where the echo licence does apply. The test applied was italic membership, because it is checkable against the committed Latin twin rather than a matter of taste.

3. Register, agreed by all four stints and kept. Walafrid's gloss is telegraphic — X: id est Y, chained with colons, frequently without a finite verb. The colon-chains, the asyndeton, and id est → "that is" are preserved rather than smoothed into flowing English prose. The terseness is the text.

§0000. STINT chunks 0000–0004 (cols 0861A–0877A), Matthew opens

Preserved verbatim from cruces-0000.md. Attribution classes below are the stint's own judgement at the time of translation; where §0 records a plate verdict, the plate governs.

Walafrid Strabo (attrib.), Expositio in quatuor Evangelia, PL 114. Martianay's own Admonitio (chunk 0000) warns that the text is "teeming with barbarisms and solecisms" and that he handed it to the printers "with almost all its faults" and deliberately did NOT correct it. Everything below is read against that declaration: an ungrammatical construction here is the expected state of the text, not a plate defect.

Standing policy for this batch — attribution of broken readings

⛔ Our Latin is Corpus Corporum's TEI, not Migne's plate. Per CLAUDE.md 8a and the measured false-rates on 8945 (6 of 9) and 8996 (7 of 8), a suspicion about a single broken WORD is normally a suspicion about our file. So in this batch:

• [sic: …] fires only where nothing at all can be rendered in place — one site (0877A palasticas).
• A one-letter slip or a word-division defect whose target is unambiguous is rendered by its evident word and logged here, classed as probable digitization damage. Those entries are marked [cc?] below. They are candidates for the plate read, not claims against Migne.
• [var: …] fires on scripture divergences from the Vulgate that change the sense. Truncation ("et reliqua" is this work's normal citation habit), word order, and medieval orthography (coelum, elymosyna, thezaurisare, baptizeretur) do not fire it.

Marker tally

Anchors, guillemets (2/2 in 0001), ? and ! counts, ## heads and paragraph structure all verified 1:1 against the Latin twins.

Vulgate divergences MARKED [var: …]

0861A (0000) — Luke 24:32. Migne: Nonne cor jam ardens erat in nobis; Vulg. Nonne cor nostrum ardens erat in nobis. A word substitution, not a truncation: "the heart already burning" for "our heart burning". Rendered as printed.

0863B (0001) — Ps 131:11. Migne: De fructu ventris tui ponam super sedem meam; Vulg. super sedem tuam. The proof-text is adduced to show that David is Christ's father; "upon my throne" reverses whose throne is in view. Rendered as printed.

0865A–B (0001) — Gen 49:10. Migne: non deficiat princeps de Juda; nec dux de femoribus ejus donec veniat cui repromissum est; Vulg. Non auferetur sceptrum de Iuda et dux de femoribus eius donec veniat qui mittendus est. Two substantive substitutions (prince/sceptre; "to whom it has been promised"/"who is to be sent"), and Migne supplies a nec the Vulgate does not print. All three carried.

0864D (0001) — Isa 7:14 / Matt 1:23. Migne: ecce virgo in utero habebit, et vocavit nomen ejus Emmanuel; Vulg. Matt vocabunt nomen eius. A perfect singular for a future plural: "he called" for "they shall call." Rendered as printed. (The omission of et pariet filium is truncation and is not marked.)

0867A (0002) — Matt 2:16, and this is the best find in the batch. Migne prints Tunc Herodes videns, quod inclusus sit a magis ("shut in"); Vulg. videns quoniam inlusus esset a magis ("mocked"). ⭐ The gloss itself depends on the Vulgate word: the very next clause quotes Ps 103:26 draco … ad inludendum eis and then says hic plus inluditur. So the exposition is built on illusus while the lemma prints inclusus. Rendered as printed, marked. Prime candidate for the plate read (n/c confusion is also within the digitization's range).

0867D and 0871B (0002, 0003) — Matt 3:2 / 4:17. Migne prints the future Appropinquabit enim regnum coelorum both times; Vulg. the perfect adpropinquavit. "Shall draw near" for "has drawn near" is a change in what is asserted; marked at both occurrences, since the chunks are read separately.

0868A (0002) — Matt 3:5. Migne: Tunc exiebant ad eum Jerosolyma; Vulg. exiebat (singular, agreeing with Hierosolyma). ⚑ Note the gloss immediately explains it: pro habitatoribus dicitur — "it is said for the inhabitants", which is exactly what licenses a plural verb. This is the reverse of the 7a″ hazard: the commentator is glossing his own text's number.

0868B (0002) — Matt 3:6. Migne: Et baptizabant in Jordanem (active, accusative); Vulg. baptizabantur ab eo in Iordane (passive, ablative). "They were baptizing" for "they were being baptized by him" changes the agent. Rendered as printed.

0868D (0002) — Matt 3:11. Migne: Ipse vos baptizat igne; Vulg. ipse vos baptizabit in Spiritu Sancto et igni. Present for future, and the Spirit dropped from the lemma — though the gloss then supplies it (vel in Spiritu sancto et igne), which is why the divergence is visible at all. Rendered as printed.

0868D–0869A (0002) — Matt 3:13. Migne: Tunc venit Jesus ad Galilaeam ut baptizeretur; Vulg. venit Iesus a Galilaea in Iordanen. The direction of travel is reversed. Rendered as printed. ⚠ Classed as plausibly digitization as well — a one-letter a→ad is within CC's range — but the reading is coherent Latin either way, so it takes [var:] and not [sic:].

0870C–D (0003) — Matt 4:10. Migne: Dominum tuum adorabis; Vulg. Dominum Deum tuum adorabis. The full form is printed correctly at 0866A in chunk 0001, twelve columns earlier, which is the evidence that the shorter one here is a defect rather than the work's habit.

0873C–D (0004) — Matt 5:17, and the most consequential divergence in the batch. Migne prints Nolite putare, quia non veni solvere legem; the Vulgate reads Nolite putare quoniam veni solvere legem aut prophetas — that is, Migne's text inserts a negation into the clause governed by nolite putare, so the printed sentence says "do not think that I have NOT come to destroy the law." ⛔ Per §7a the non is carried into the English, contradiction and all, and marked. The gloss immediately below prints the sentence again without the negation (ecce non veni solvere legem, sed augere) — i.e. the correct sense — which makes the lemma's non an isolated intrusion. Rendered as printed.

0875A (0004) — Matt 5:34. Migne: neque per coelum quia sedis est Dei; Vulg. quia thronus Dei est. A word substitution ("seat"/"throne"), plus a nominative sedis for sedes. Rendered as printed.

0875B (0004) — Matt 5:41. Migne: vade cum illo alia mille; Vulg. vade cum illo alia duo. The number of the second mile is changed. Rendered as printed.

Vulgate divergences LOGGED, not marked (truncation, order, orthography, conflation)

• 0863B (0001) Matt 1:1 filii Abrahae for filii Abraham — declension of the name.
• 0863B (0001) Gen 22:18/26:4 In semine tuo benedicentur omnes gentes — Vulgate order reversed and terrae dropped. No gloss turns on either.
• 0864C (0001) Matt 1:20 Angelus Domini in somnis dixit for apparuit ei dicens — compression. The gloss turns on in somnis, which is intact.
• 0864D (0001) Isa 7:14 lemma omits et pariet filium.
• 0866A (0001) Matt 2:5 in Bethleem Judae for Iudaeae.
• 0867B (0002) Matt 2:16 et mittit et occidit for the Vulgate participle mittens occidit.
• 0867B (0002) Ps 103:26 ad inludendum eis for ei.
• 0867C (0002) Matt 2:23 Venit in civitatem Nazareth compresses veniens habitavit in civitate quae vocatur Nazareth.
• 0868C (0002) Matt 3:9 Potest Deus et lapidibus istis for potens est Deus de lapidibus istis.
• 0868C (0002) Matt 3:10 Omnis arbor qui non facit … excidetur for quae … exciditur — gender (Pattern 9, no English exponent) and tense.
• 0870A (0003) Matt 4:4 Non in solo pane for in pane solo; vivit for vivet.
• 0870C (0003) Matt 4:10 vade retro — Matt has vade Satanas; vade retro is Mark 8:33. A conflation, and the gloss expounds retro, so it is the author's text.
• 0871C (0003) Matt 4:20 + 4:22 conflated: relictis retibus et patre joins the nets of the first pair to the father of the second. The gloss then distinguishes the two pairs, so the conflation is in the lemma only.
• 0872A (0003) Matt 4:24 Obtulerunt ei paralyticos compresses two clauses.
• 0872B (0003) Matt 5:1 Videns autem Jesus turbas — Jesus supplied; likewise at 0870D Matt 4:12.
• 0872C (0003) Matt 5:5 Beati qui lugent nunc — nunc from Luke 6:21.
• 0873A (0003) Matt 11:29 discate a me for discite. [cc?]
• 0873B (0003) John 4:34 ut faciam voluntatem patris mei for eius qui misit me.
• 0873B (0003) Matt 8:20 non habuit ubi caput reclinet for habet.
• 0874A (0004) Matt 5:19 Qui enim solverit for Qui ergo.
• 0874C (0004) Matt 5:24 vade reconciliari drops prius.
• 0874C (0004) Matt 5:25 ne forte adversarius tradat te judici — Vulgate order ne forte tradat te adversarius iudici.
• 0874D (0004) Matt 5:28 ad concupiscendam eam for concupiscendum.
• 0875A (0004) Matt 5:37 quod amplius est for quod autem his abundantius est.
• 0875B (0004) Matt 5:39 si quis percusserit te in dexteram maxillam — order and case against si quis te percusserit in dextera maxilla tua.
• 0875C (0004) Matt 5:45 et pluit super justos drops et iniustos — ⚑ note the gloss requires both (quia in utroque totus mundus continetur), so the truncation is in the lemma while the exposition still has the full verse in view.
• 0875C (0004) Matt 5:43 diligis for diliges; antiquis supplied.
• 0875D (0004) Matt 6:2 Cum ergo facis for facies.
• 0876A (0004) Matt 6:8–9 scit enim pater drops vester; quod vobis opus est for quid opus sit vobis; sic ergo orate for sic ergo vos orabitis.
• 0876A (0004) Matt 6:10 Veniat regnum tuum for Adveniat. ⚑ The same work prints adveniat regnum tuum correctly at 0876C, in the summary of the seven petitions — an internal inconsistency worth the plate reader's eye.
• 0876B (0004) Matt 6:11 Panem nostrum quotidianum where Vulg. Matt has supersubstantialem (Luke 11:3 has cotidianum). The liturgical form; the gloss expounds quotidianum (omni die usque ad caput anni), so it is the author's text.
• 0876C (0004) Matt 6:16 Cum autem jejunaverit for ieiunatis.
• 0876C (0004) Matt 6:19 noli thezaurisare (singular) for Nolite thesaurizare vobis thesauros.
• 0876D (0004) Matt 6:24 and 6:21 conflated in one lemma.

Broken or unconstruable readings

0862B (0000) — artis, dosis, anatholis, mesimbrio. The four Greek cardinal points (ἀνατολή, δύσις, ἄρκτος, μεσημβρία) behind the ADAM acrostic, in a corrupt medieval transliteration with artis/anatholis apparently transposed from arctos/anatole. Carried in Migne's italics, untranslated, as foreign proper names. No [sic:]: this form of the acrostic is a widely attested medieval corruption, not evidence about this plate (Pattern 9).

0861A (0000) — insufflictio for insufflatio (gloss on Phison). Rendered "insufflation". [cc?]

0861C (0000) — Ex aere flatus: Igne sanguis, aqua flamma, terra corpus. The four-element scheme assigns flame to water, which contradicts the scheme it is stating. Rendered as printed; no marker, because flamma is a real word in its slot. Candidate for the plate.

0862B (0000) — Gen 6:15 paraphrased, not quoted: longitudo trecentorum cubitorum et latitudo unius, altitudo triginta et uno cubito. The Vulgate has breadth fifty, height thirty. ⚑ Not marked [var:] — the sentence is legitur paraphrase, not a scripture lemma, and the allegory is built on the one cubit (unum cubitum unum Dominum colere), so the number is load-bearing for this author rather than a slip.

0864A–B (0001) — (deletum est verbum [0864B] unum forte vero,) — the editor's own parenthetical note about a deletion in his copy, printed inside the body text and with a column anchor falling inside it. Translated as prose in place (it is a sentence that says something; a Latin-only rendering would strand an English reader), parentheses kept, the conjectured word vero kept in italic Latin, anchor kept in position.

0864B (0001) — factus ex Spiritu Spirius ( sic ). Migne prints his own ( sic ); per Pattern 12 it is reproduced verbatim and no [sic: …] is stacked on it. Spirius carried untranslated in italics.

0864B (0001) — antequam fuissent in qua habitatione. Does not construe; in qua stands where in una is required. Rendered as closely as the words allow ("before they had been in whatever dwelling"); no emendation in the text.

0865A (0001) — donec pronum quam [ Forte, priusquam] dicitur primogenitus. Migne's own bracketed conjecture stands beside his broken run. Per Pattern 19 and Pattern 12, his conjecture is reproduced verbatim untranslated and ours is not added; the run pronum quam is carried in italics without a [sic:], because the reader already meets Migne's answer to it on the same line.

0865B (0001) — Non habemus regem nisi Cae sarem. A word split across what would be a line break. [cc?] — this is the named Corpus Corporum corruption shape, so the English renders "Caesar" and the split is logged rather than marked. If the plate carries the split, this becomes a Pattern 10 [sic: Cae sarem].

0865B (0001) — Item stellam ei, et non coeli et terrae, necnon eorum fuit, sed Filii Dei, and again at 0865C sive timuerunt omnes, nec non illis bene esse. In both places necnon ("and also") stands where the sense requires a plain nec. Per negation fidelity the negative is carried in the English ("nor was it theirs", "nor was it well with them") rather than being turned into an affirmative to make the sentence run.

0865A (0001) — mysterium Dei non potest cognosere semel. For cognoscere. Rendered "known". [cc?] The non is carried.

0866A (0001) — ut cissent in quo tempore. For scissent/scirent. Rendered "might know". [cc?]

0866A (0001) — cum enim salvatorem in altum duceret, renuit, ipse non renuit. The plate asserts and denies the same refusal in one clause. Both carried per §7a; no repair, no marker.

0867A (0002) — Noctem hic, ostendit quod secessit litterae tenebrae legis. Three stacked genitive/nominative forms that do not agree. Rendered "he withdrew, the darkness of the letter of the law." Martianay's solecism class.

0867A (0002) — primis Moysen et Aaron provocant Christum. Unconstruable as printed (primis for primum; two accusative objects for one verb). Rendered as closely as the words allow.

0867B (0002) — infantiae nomen amorum duorum for annorum. Rendered "years". [cc?]

0868A (0002) — Vestitum Joannis de camelolis. For camelis, or a late diminutive camelulus. Rendered "of camels". [cc?]

0868A (0002) — qui litteram legis uti locustae flos. Nominative flos where an accusative object is required. Rendered literally ("as locusts the flower").

0868B (0002) — mel Sylvestrae for silvestre — gender/declension, Pattern 9, no English exponent. Rendered "wild honey".

0868B (0002) — matrem suam synagogem for synagogam. Rendered "the synagogue".

0869A (0002) — ad te pervenit omnem justitiam et reliqua (Matt 3:15, sic decet nos implere omnem iustitiam). Case and verb both wrong for the construction. Rendered as closely as the words allow ("there comes to thee all justice"); no emendation.

0869D (0002) — ut diabolus falleret ejus nativitate. Active falleret where the sense needs a passive. Rendered "that the devil might be deceived as to his nativity" — ⚠ this is the closest English the printed words allow, and it is flagged here because it is the shape §7a warns about; if a reader judges it a repair, the literal alternative is "that the devil might deceive by his nativity", which asserts the opposite.

0870A (0003) — super pinnam templi, id est super murum archae maceriae. A chain of unagreeing genitives. Rendered "the wall of the ark's enclosure."

0871D (0003) — si cut Gedeon for sicut. [cc?] Rendered "as".

0873A (0003) — inter hominem et proximuminter, hominem et amicum suum. A fused run: proximum inter. [cc?] — a word-fusion at a line break is the named digitization shape, so the English renders the sense ("between man and neighbour, between man and his own friend") and Migne's comma is kept where he prints it. If the plate carries the fusion, this becomes a Pattern 10 [sic: proximuminter].

0873C (0003) — id est apostolos et doctores . non ponunt sub modio. A full stop mid-clause with spaces either side. Preserved exactly (Pattern 8, never delete).

0875A (0004) — manus, id est, opos malum abscide for opus. Rendered "work". [cc?]

0875D (0004) — hypocrita dicitur fictus perciperunt mercedem suam (Matt 6:2, receperunt). Perciperunt is not a form of percipio (perceperunt). Rendered "they have received". [cc?]

0875C (0004) — illi qui nos oper persequitur for opere. Rendered "by work". [cc?]

0875C (0004) — in pluvia, humida et grigida for frigida. Rendered "cold". [cc?]

0876D (0004) — qui divitiis servit, diabolos eruit. "He who serves riches draws out devils" — a real verb in its slot, but it makes the servant of riches the one who expels demons, which reverses the moral of the sentence. Rendered as printed (Pattern 7); declined for a [cj:] because the defect is not a negation or a correlative (Pattern 18a), and no candidate reading is clearly right (eruitur? servit?). Flagged for the plate.

0877A (0004) — ⛔ nisi deponant portam fides palasticas II. The one site in this batch where nothing can be rendered in place: palasticas is not a Latin word and the surrounding case-forms will not resolve into a construction. [sic: palasticas] fired; the rest of the run is rendered word for word ("the gate, faith … II"), and per §7a⁗ the clause keeps its verb (deponant) and both its negatives (non capent, nisi). Class: uncertain. palasticas is a lexical non-word rather than a division defect, which is why it takes the marker where the [cc?] entries above do not — but it could still be a CC misreading of, e.g., plasticas. The plate read decides it.

Register notes

• The telegraphic X: id est Y chains, the colon-strung asyndeton and the absent finite verbs are preserved rather than smoothed, per the work's own Admonitio.
• id est is rendered "that is" throughout; vel/sive opening an alternative construal, "or"; item, "likewise"; et reliqua, "and the rest" — fixed within the work.
• Pattern 17: the work addresses a singular reader (erue eum et projice abs te, unge caput tuum), so thou throughout, including inside the lemmata; the plural ut sitis filii patris vestri, vos estis sal terrae takes you.
• One em-dash was written and removed on review; the batch now carries none.

§0005. STINT chunks 0005–0009 (cols 0877B–0892B), Matthew closes, Mark opens

Preserved verbatim from cruces-0005.md. Attribution classes below are the stint's own judgement at the time of translation; where §0 records a plate verdict, the plate governs.

PL 114, cols 0877B–0892B. Chunks 0005–0007 close the Matthew section, 0008 is the 242-word tail of Matthew, 0009 opens IN EVANGELIUM MARCI.

⚠ General caution for this work. Martianay's Admonitio warns that the Latin is full of barbarisms and solecisms deliberately left uncorrected. Case, number and gender routinely fail to agree (messes multa, hoc sunt prophetas, clamidem coccineo, doctrinam paternarum, duobus tunicis). These are rendered as printed and are not logged individually below except where the sense is at stake — under Pattern 9 they carry no English exponent and are not defects to mark.

⚠ Marker-attribution classes. Per CLAUDE.md 8a, our Latin is Corpus Corporum, not Migne's plate. Every [sic:] and [var:] below carries a class judgment: (A) likely Migne's own type/text; (B) likely a Corpus Corporum digitization artifact (letter swap, metathesis, split/fused run). No column in this range has been plate-read; plate-gate.mjs must clear all of them before the work ships.

[sic: …] markers fired — carried non-word type

[d: …] — dittography

0888A–B. Migne prints the gloss on the scarlet cloak twice, with variation:

induit [0888B] clamidem coccineo, id est Christum indutum fide gentium. Induitur clamidem coccineo, id est Christum indutum fide gentium aliter per coccineum clamidem, ostendit effusionem sanguinis…

The SECOND copy is marked — and per Pattern 11a the reason has to be stated, because the two copies differ (induit / Induitur):

1. The first copy's induit is active, matching the active exuerunt Christum that governs the whole sequence; the second's Induitur is passive and agrees with nothing before it.
2. Deleting the second copy leaves …induit clamidem coccineo, id est Christum indutum fide gentium. Aliter per coccineum clamidem, ostendit… — a first gloss closed by a full stop and an aliter introducing the alternative, which is this work's standard shape. Deleting the first would strand the column anchor [0888B] inside the deleted run.

Class: A/B undecided. An eye-skip at a column break is as plausible in a compositor as in a transcriber; the [0888B] anchor falling inside the doubled run is weak evidence for the plate.

Vulgate divergences

Marked [var: …] (whole-clause or whole-word divergence, sense at stake)

Logged only — divergence real but single-word, so probably OUR file, not Migne's

Per the standing rule (single broken word ⇒ suspect the digitization; whole clause ⇒ evidence about Migne), none of these got a marker. Each still needs the plate.

• 0877C Et omnis qui audit verba mea et non faciet ea (Matt 7:26) — Vulg. verba mea haec et non facit ea: future for present, haec absent. Rendered "will not do".
• 0877D tantum dic verbum, et sanabitur (Matt 8:8) — Vulg. tantum dic verbo et sanabitur puer meus.
• 0878B reversus in domum suam (Matt 9:7) — Vulg. abiit in domum suam.
• 0878D Nihil opertum quod non reveletur (Matt 10:26) — Vulg. Nihil enim est opertum quod non revelabitur: subjunctive for future indicative.
• 0879B lignum fumigans (Matt 12:20) — Vulg. linum fumigans, "smoking flax." Rendered "smoking wood". This is the most consequential single-word divergence in the range: the printed lignum makes the gloss ("the hearts of the gentiles made ready for the faith") read against a famous image the plate may or may not carry. lin → lign is one inserted letter and therefore the class the rule says is usually ours. ⭐ Priority plate read.
• 0880A videntes non videant (Matt 13:13) — Vulg. non vident.
• 0880B Et facit fructum (Matt 13:23) — Vulg. et fructum adfert.
• 0881C tu solus discipulus ejus es? (John 9:28, of the blind man) — Vulg. Tu discipulus eius sis; solus may be attracted from Luke 24:18.
• 0883D difficile dives intravit (Matt 19:23) — Vulg. dives difficile intrabit.
• 0885A spe circumdedit (Matt 21:33) — Vulg. sepem circumdedit, "he set a hedge about it." Rendered "he surrounded it with hope"; the gloss (id est custodia angelorum, "the guard of angels") fits a hedge, not a hope, so the printed reading is probably damaged. One letter; plate needed.
• 0885C Nemo militans Deo implicans se (2 Tim. 2:4) — Vulg. implicat se.
• 0886A ecce relinquetur domus vestras desertas (Matt 23:38) — Vulg. relinquetur vobis domus vestra deserta; number disagreement, no English exponent for the verb.
• 0887A the five-senses proof texts are all heavily compressed rather than divergent (Ps. 118:37; Ecclus. 28:28; 2 Cor. 2:15; Rom. 12:14; Eph. 4:28) — compression is not a variant and none is marked.
• 0892A consepulti enim sumus cum Christo per baptismum: in morte enim ipsius baptizati sumus — a conflation of Rom. 6:4 and 6:3 (cum illo), not a divergence in a single text; left unmarked.

Other cruces (no marker)

• 0877D vel vidissent quod natus esset a lepra — "that he had been born from leprosy." natus will not construe with a lepra; mundatus or sanatus is what the sense wants (the healed leper shows himself to the priest). Rendered literally.
• 0878C pulvis tribus modis impendit, gressus obligat, oculus obscurat, vestimenta inquinat — oculus is nominative in a run of accusative objects (gressus, vestimenta). Rendered "the eye darkens", which keeps the printed case; oculos is the obvious conjecture but supplying it would be a silent repair.
• 0878D seu Christus Barabias — the name printed thus; carried as printed, not normalized to Barabbas.
• 0880D Thetrarcha, id est inita pars regni — "an entered part of the kingdom." A tetrarch rules a fourth part; quarta is what the etymology requires and inita is a real word occupying its slot, so Pattern 7 governs and no marker is fired.
• 0884A prima hora, id est infantia secunda; hora adolescentia, sexta juventus — Migne's punctuation splits secunda hora across a semicolon. Reproduced exactly under Pattern 8; the English is correspondingly broken.
• 0884C qui baptismum vel quacumque sacrationem — the verb is elided. The English supplies [sell] in the corpus's existing supplied-word brackets rather than smoothing.
• 0884C relectis illis abiit foras — relectis ("re-read") where the sequence wants relictis ("left behind"), which the very next clause prints (ostendit relictos Judaeos). A real word in its slot, so rendered literally ("those having been read over again") under Pattern 7; the conjecture lives here.
• 0884D Mens id est Christus, tollite hinc — Mens ("the mind") where Matt 21:21 requires Mons ("the mountain… be thou taken up and cast into the sea"); the following jactate in mare proves the mountain is meant. Real word, rendered as printed.
• 0885D Mulier . — the plate sets a space before the stop. Reproduced.
• 0886A in spiris pinnis, id est sublevatis alis — "in coiled feathers"; the gloss says the wings are raised, so spiris is doing no work the gloss recognizes. Rendered literally.
• 0888C consolationem recorrectionis — "the consolation of re-correction." resurrectionis is the sense the paragraph (the empty tomb, the first day of the week) wants. recorrectio is a formable Latin word, so it is rendered, not carried.
• 0889A temperantia est omni irae, imponere motum — "to lay upon all anger, a motion." modum ("a limit") is plainly intended, and motum inverts the definition. A real word in its slot: rendered as printed, comma and all, per Patterns 7 and 8.
• 0890B invenit eam in pristino.. . . . . corporis sui — Migne prints a lacuna here, as a run of spaced points. Reproduced verbatim in the English rather than filled or normalized; no [ed:], because nothing has been lost by the digitization — the hole is in the text as printed. Plate read wanted to confirm the exact number of points.
• 0891B ⭐ ut ait Propheta frater non redemit, id est Moyses per legem redemit. — The gloss contradicts its own lemma. Ps. 48:8 (fratrem non redimit homo) is quoted with its negative, and the exposition then says Moses did redeem through the law, where the whole surrounding argument is that neither Prophets nor Patriarchs could heal the human race nisi veniens Christus. The negation is carried in the lemma and not supplied in the gloss: the printed page contradicts itself and the English says so. A dropped non in the gloss is the near-certain conjecture. This is the one Pattern 7a site in the range where the temptation to repair is strong.
• 0891C humiliare se in sacculo per poenitentiam — "in a sackcloth"; saeculo ("in the world") is the likelier reading, since the lemma is mergere in mare and the gloss chain runs through worldly things. Real word in its slot, rendered as printed.
• 0891A revocavit illum ad pristinam mundanae naturae — pristinam has no noun. Rendered "to the former of his worldly nature", keeping the hole visible.

§0010. STINT chunks 0010–0013 (cols 0892C–0904D), Mark closes, Luke

Preserved verbatim from cruces-0010.md. Attribution classes below are the stint's own judgement at the time of translation; where §0 records a plate verdict, the plate governs.

Range: PL 114, cols 0892C–0904D — the close of IN EVANGELIUM MARCI (0010) and the opening two-thirds of IN EVANGELIUM LUCAE, whose head falls at col 0893 (0011–0013).

⛔ Every [sic: …] and [var: …] below stands on a column NOBODY HAS READ AT THE PLATE. Our Latin is Corpus Corporum's transcription, not Migne's page. Each entry carries an explicit attribution guess — our-file (probable CC digitization damage: a single broken word, a c→e / n→u / m→in confusion, a split or fused run) or plate (a whole quoted clause, well-formed Latin, that diverges as a unit). node scripts/plate-gate.mjs 9004 must be satisfied before this work ships.

1. Register and structural notes (no marker)

• Heads. IN EVANGELIUM MARCI. (cont.) / IN EVANGELIUM LUCAE. / IN EVANGELIUM LUCAE. (cont.) render as ON THE GOSPEL OF MARK / OF LUKE, with (cont.) carried untranslated (it is our chunk marker, merged at assembly).
• 0011 opens biographically, not with a gloss. Cols 0893D–0895A give Luke's origin (a Syrian of Antioch), his profession, his death at 83 in Bithynia, his burial at Constantinople, and the four Evangelists' four audiences. Rendered as the prologue it is; the head structure is exactly the Latin's.
• The chunk-0011 anchor run [0893] then [0893D] is the plate's own (the Luke division head carries a bare column number before the lettered band). [0893D] is set flush against Lucas with no space; reproduced the same way. Not a defect — known false-positive class 1.
• The work's Latin is barbarous by its own editor's admission (Martianay's Admonitio leaves the solecisms uncorrected), and this range is thick with them. Rendered as they stand, not itemized individually, none marked: accusative for ablative after in (Agnum immolatum in Aegyptum 0893B, in monte orare 0898B, in praedicationem audire 0904C); loss of final -m (in pala for in palam, 0901B; in civitate 0895C); b/v interchange (acerva for acerba 0894A, portabit for portavit 0897A); d/t finals (reliquid ×2, 0894A); ch→c (caus for chaos, 0902A); gender/case mismatch with no English exponent (sterilis … maledictus 0895A, majore gaudium 0902A, Barrabam filius 0904A, duas leges after id sunt 0903C, novem servis 0903A); y/i and ph spellings (sabbatho, simphonias, Nichodimus, Bithinia, hiericho, Mathaeo, sylva, syndone/sindone in one sentence).
• Aphesis is this text's habit, not a defect (Pattern 9): nagoga for synagoga (0898A) and postolos for apostolos (0901D) are the same vulgar-Latin loss of an unstressed initial syllable. Translated normally as "the synagogue" and "the apostles". ⚠ If the plate reads synagoga / apostolos in either place, these are CC losses and belong in a patch file, not here.
• Printed punctuation carried where it breaks the sense (Pattern 8, never supply, never delete): the full stop inside aptum fuit ut in tempore officii. Patris Domini ceptus esse (0895A); the missing stop at praedica baptismum Vallis implebitur (0897C) and at regnante cum patre justificata est sapientia (0899A), both carried unpunctuated; the colon splitting homo qui divitias: et paupertatem non sustinet (0898A), rendered "the man who riches: and poverty does not bear" so the mark keeps its place; the spaced ellipsis si gentiles . . . ut equus molens (0902B), carried verbatim; and the chunk-0013 ending, which stops after dimisit with no terminal mark and does so in the English too.
• Direct questions keep their marks (0896A cur non dixit Johannis erit?, 0899A quis est qui peccata dimittit?, 0904C nunc eos congregat in unam pacem? quare non uno loco surrexit et ascendit?). None is an indirect question, so 8b does not fire anywhere in this range.
• Fixed equivalences within this range (merge-time reconciliation): theorica and contemplativa both = "contemplative"; actualis and activa both = "active"; doctores = "teachers"; stabularius = "innkeeper"; mna = "mina". ⚑ The Latin carries two synonym pairs for the contemplative/active distinction (theorica/actualis at 0898B, 0899B, 0904C; contemplativa/activa at 0900B); English has one exponent for each side and the doublet is recorded here rather than invented in the prose.
• Etymological glosses keep the Latin word visible where the gloss is the etymology: quadrans (0892C, "which is a quadrans, through which he understands the man who squares"), and sylva missus dicitur (0900D), which is the Siloam etymology of Jn 9:7 (Siloe, quod interpretatur Missus) attached to Lk 13:4's tower — sylva is a real word carrying a name, so it is kept as a mention rather than rendered "a wood".

2. Defect markers fired — [sic: …]

Seven, all on runs that cannot be rendered at all: either non-words, or a real word with no grammatical slot (Pattern 12's test 1). Every one of them is a single broken word or a split of one word, which is the class the standing warning says is usually our file — the attributions below say so, and none of them is offered as a finding about Migne until the plate is read.

Declined [sic:] — non-words and fusions I judged to be our file and rendered rather than marked (each is a silent treatment only in the sense that Pattern 9 covers it; they are all listed here so a plate reader can overturn them):

• 0898C praeceditmortem — a fused praecedit mortem. Rendered "precedes death". Word fusion at a line break is the named CC class; marking it would put a plate accusation on the likeliest digitization artifact in the range.
• 0892B non sciebat vas transferre per templum — Mk 11:16 has non sinebat ut quisquam vas transferret. sciebat/sinebat is a c/n confusion, the other named CC class. Rendered literally, "did not know how to carry a vessel through the temple", which is what the printed words say and is visibly odd.
• 0898C tangitur locus a Deo — Lk 7:14 tetigit loculum. Rendered "the place is touched by God", not "the bier": reading loculum here would be exactly the silent repair 7a forbids, and locus renders in place.
• 0904B aucurrebat — beside cucurrisset in the same clause; taken as a spelling of accurrebat and rendered "ran".
• 0896B adsuerunt — rendered "had become accustomed" (adsueverunt).
• 0896B a tabernaculo meritoria — the Roman taberna meritoria whose fountain of oil is the standard Nativity portent. Rendered "from the hired lodging"; the gender mismatch has no English exponent (Pattern 9).
• 0893B amplectitudo — not classical, but a transparent formation on amplector. Rendered "the embracing-breadth", keeping the printed active ostendit ("shows", not "is shown") against the passive ostenditur one clause earlier.
• 0898B lenes — istae quatuor lenes. A real adjective with no noun; rendered literally "these four gentle ones". No conjecture is offered because none is supported.

3. Vulgate divergences — the collation, lemma by lemma

Every italicised Gospel lemma in the four chunks was listed and collated word by word against the Vulgate before translating. Five divergences were judged material and took a [var: …]; the rest are recorded here and took none. Migne's wording is rendered in every case.

Marked (5)

Collated and NOT marked (the full list)

Lemma abbreviation is this commentator's normal practice, so a dropped connective or a truncated opening is not a divergence in the Pattern 14 sense; those are listed anyway so the collation is complete rather than selective.

• 0892B Mk 11:13 Non erat tempus ficorum — Vulg. non enim erat. Dropped connective.
• 0892D Mk 13:32 De die et hora nemo scit — Vulg. De die autem illo vel hora. Lemma compression.
• 0893A Mk 13:34 Homo peregre profectus est — Vulg. Sicut homo qui peregre profectus. Lemma compression.
• 0893B Mk 14:13 Homo languenam aquae bajulans — Vulg. lagoenam. Orthographic; the gloss (languena baptismum Trinitatis significat) uses the same form, so it is the text's own spelling. Pattern 9.
• 0894B Mk 15:46 / Mt 27:60 Monumentum scissum — Vulg. monumento quod erat excisum de petra. One prefix; single-word suspicion, declined.
• 0894C Mk 16:9 Maria, de qua ejecerat septem daemonia — Vulg. Mariae Magdalenae. Case + abbreviation; the name is supplied in the 0899A gloss.
• 0894D Mk 16:17–18 linguis novis loquentur (Vulg. linguis loquentur novis, order); si mortiferum biberint (Vulg. si mortiferum quid biberint); super aegros manus (Vulg. aegrotos). All minor.
• 0894A Mk 15:34 eli eli lemasabactani, quod est Deus Deus meus quare me dereliquisti — Vulg. Mk Heloi Heloi lama sabacthani … Deus meus Deus meus ut quid dereliquisti me. Migne's transliteration is Matthew's (Heli) with the Aramaic fused into one token, and the rendering follows an Old Latin quare. Not marked: the transliterated form is a name-shape, and the gloss turns on none of it.
• 0895A Lk 1:5 Fuit in diebus Herodis — matches.
• 0895A Lk 1:11 ad dextris altaris and 0903C Lk 22:69 sedens ad dextris virtutis — Vulg. a dextris both times, and Lk 22:69 also has Dei. ad + ablative is this text's habit (twice, in two Gospels), so Pattern 9.
• 0895B Lk 1:7 Non erat illis filius — matches. But the following gloss prints ambo perseverant in diebus suis against Vulg. ambo processissent; rendered "persevere", present tense as printed.
• 0895B Lk 1:13 exaudita est oratio tua — Vulg. deprecatio.
• 0895C Lk 1:25 Auferte obprobrium meum — Vulg. ut auferret obprobrium meum inter homines. An imperative plural for a purpose clause; the gloss keeps inter homines, so the lemma is Luke's. Single-word inflection, declined, but this is the nearest miss in the range — if the plate reads Aufer or auferret, this is a CC corruption and should be recorded as one.**
• 0895D Lk 1:35 Spiritus sanctus venit in te (Vulg. superveniet) and Virtus altissimi obumbravit tibi (Vulg. obumbrabit). Both past for future, and the glosses exploit the past (abscondit te in Aegypto fugiendo), so the English follows Migne. Declined as single-word inflections; worth a plate read, since two tense shifts in one verse could be one compositor.
• 0895D Lk 1:39 Abiit in montem — Vulg. in montana.
• 0896A Lk 1:41 in utero matris ejus — Vulg. in utero eius.
• 0896A Lk 1:68 benedictus Dominus noster — Vulg. Dominus Deus Israhel.
• 0897A Lk 11:27 beatus ille venter qui te portabit — Vulg. beatus venter qui te portavit. portabit is the b/v spelling; ille is an addition.
• 0897A Lk 2:35 pertranseat gladius — Vulg. pertransiet.
• 0897B Lk 2:51 et erat illis subditus — matches (italic in Migne, kept italic).
• 0897C Lk 3:2 factum est verbum Domini super Johannem — Vulg. verbum Dei.
• 0897C Lk 3:5 Montes et colles humiliabuntur — Vulg. omnis mons et collis humiliabitur; number. Same verse as the marked prava indirecta.
• 0897D Lk 4:18 et captivis redemptionem — Vulg. remissionem; and dimitte confractos for the Vulgate's infinitive dimittere.
• 0897D Lk 4:21 hodie impleta est scriptura — Vulg. haec scriptura.
• 0898A Lk 5:7 Impleverunt ambas naves — Vulg. naviculas.
• 0898A Lk 5:8 exi a me, Domine, quia peccator sum — Vulg. exi a me quia homo peccator sum Domine; order, and homo dropped.
• 0898C Lk 7:11 civitatem quod dicitur Naum — Vulg. quae vocatur Naim.
• 0898B Lk 6:23 mercis vestra multa est in gloria — Vulg. merces vestra multa est in caelo; and faciebant patres eorum Prophetas for Vulg. faciebant prophetis patres eorum.
• 0898C Lk 6:26 Vae vobis cum benedixerint homines — Vulg. vae cum bene vobis dixerint omnes homines.
• 0899C Lk 8:44 Tangit fimbriam — Vulg. tetigit; tense.
• 0899C Lk 9:13 Vos date illis manducare — Vulg. date illis vos manducare; order.
• 0900C Lk 11:41 date eleemosynam et omnia munda sunt — Vulg. quod superest date elemosynam et ecce omnia munda sunt vobis. The clause that follows in Migne (et si non dederitis immunda sunt omnia quae habetis) is not scripture at all but the commentator's own contrary, so no [var:] attaches to it.
• 0900C Lk 12:32 Noli timere pusillus grex — Vulg. Nolite timere. A singular imperative for a plural. Rendered "Fear not", where English shows no number; declined as a single-word inflection, and Pattern 17 is not engaged because an English imperative carries no second-person pronoun.
• 0900B Lk 10:35 quodcumque super erogaveris — Vulg. supererogaveris. A split that happens to render exactly ("whatsoever thou shalt have spent over and above"), so no marker and no loss.
• 0901A Ps 79:9 vinea de Aegypto transtulisti — Vulg. vineam de Aegypto transtulisti; final -m.
• 0902A Lk 16:26 caus magnum — Vulg. chaos magnum; treated as the ch→c spelling, rendered "a great chaos".
• 0903B Lk 21:18 Capillus de capite non peribit — Vulg. de capite vestro. Rendered without a possessive, since none is printed.
• 0904D Mt 5:44 orate pro persequentibus — Vulg. orate pro persequentibus et calumniantibus vos. The chunk ends mid-clause at dimisit; nothing is supplied.
• 0892C Mk 11:24 / Mt 21:22 quaecumque petieritis, credentes in oratione, fiet vobis — a conflation of the two Gospels' forms; 0892D Rom 8:26 nescimus quid petamus, nisi Christum qui interpellat pro nobis — paraphrase with the commentator's own nisi Christum inside it; 0893C Heb 9:13 sanguis hircorum et taurorum — order reversed. Inline, loose, none marked.

4. Dittography — [d: …] ×2, both in chunk 0011

5. Open, and what would settle it

1. Plate read, cols 0892C–0904D, for all seven [sic:] and all five [var:]. Priority order: 0904A densi seduna (no conjecture at all), 0893A Fretus alabastrum, 0893C Videbunt (the one whose divergence the gloss depends on, so confirming it at the plate converts the range's most valuable finding from probable to certain), 0898B caudes, 0897C prava indirecta.
2. Migne's foot-of-page apparatus. Per data/plate-notes/coverage.json this work owes at least a spot check; PL 114's verified map is PDF page = (column + 11) / 2 for patrologiaecurs06saingoog, recalibrated 2026-08-19. Several declines above (caudes, densi seduna, lenes) are exactly the shape Migne answers with a Forte note, and we cannot see that layer.
3. The 0895D tense pair (venit/obumbravit for superveniet/obumbrabit in one verse of Lk 1:35) is the one declined divergence I would most expect a plate read to promote to a [var:].

§0014. STINT chunks 0014–0017 (cols 0903–0916B), John and the Capitulatio

Preserved verbatim from cruces-0014.md. Attribution classes below are the stint's own judgement at the time of translation; where §0 records a plate verdict, the plate governs.

Range: PL 114, cols 0903–0916B — the John section (IN EVANGELIUM JOANNIS) and the closing INCIPIT CAPITULATIO to the end of the work.

⛔ Every [sic: …] and [var: …] below stands on a column NOBODY HAS READ AT THE PLATE. Our Latin is Corpus Corporum's transcription, not Migne's page. Each entry below carries an explicit attribution guess — our-file (probable CC digitization damage: a single broken word, a c→e / n→u / m→in confusion, a split or fused run) or plate (a whole quoted clause, well-formed Latin, that diverges as a unit). scripts/plate-gate.mjs 9004 must be satisfied before this work ships.

1. Register and structural notes (no marker)

• Head, cols 0911D and 0915B — INCIPIT CAPITULATIO. Rendered "HERE BEGINS THE CAPITULATIO." Capitulatio is kept as the printed technical label: nothing in the matter that follows is a chapter list — it is continuous lemma-and-gloss on John 13–21 — so translating it "chapter-list" or "table of chapters" would make a claim about the section's function that the page does not support. Structural label, Pattern 9 family.
• Col 0912A/B — Explicit Passio. Migne sets this as a centred italic line inside the John section, before the Farewell-Discourse glosses resume. Reproduced in place and in the same structure as "Here endeth the Passion." It is Migne's rubric, not a defect, and takes no marker.
• Col 0911A — the printed ellipsis. Migne prints sic et Christus . . . multiplicavit Ecclesiam suam, spaced periods and all. Carried through verbatim (Pattern 8: never supply, never delete). Nothing is missing from our source here that we can demonstrate, so no [ed:].
• Cols 0908D, 0914C — printed question marks on non-questions. apertos oculos tenuit ad malum et clausit ad bonum? and duo discipuli currentes? Both marks ride as printed (Pattern 8).
• The work's Latin is barbarous by its editor's own admission (Martianay's Admonitio), and this range is the densest stretch of it: nominatives for accusatives (alienus sequuntur, majorem omnibus), indicative for subjunctive (est qui quaerit et judicet), singular verbs with plural subjects, ad + infinitive (ad clarificare, 0913C). These are rendered as they stand and are not itemized individually below; none is marked.

2. Defect markers fired — [sic: …]

Declined, deliberately — real words rendered as printed under Pattern 7, with the conjecture recorded here and nowhere else:

• 0906A, in libano crucis — "on the Lebanon of the cross," where the sense wants in ligno crucis, "on the wood of the cross." libano is a perfectly good word, so Pattern 12 has nothing to wrap; and the English, though opaque, does not assert anything false, so it fails the Pattern 18 [cj:] bar (18a). Rendered as printed.
• 0908C, Scripturae Scriptura divina vocati nominati — the same lexeme in two forms, adjacent. It looks like an eye-skip, but Pattern 11's [d:] is a claim that Migne's plate sets a run twice, and one word of our transcription is not evidence for that (see the header). Rendered as printed: "of the Scriptures, by divine Scripture called, named."
• 0908C, abscondit Jesus … a cordibus eorum se, exivit de templo — a stray se. Same reasoning; rendered as printed ("himself he went out of the temple").
• 0916B, insita arbor, in alter facit fructum arboris — in alter does not construe. Rendered as closely as the printed words allow ("a grafted tree, in another it maketh the fruit of the tree"); no marker, because the whole sentence is of a piece with the work's admitted solecisms.
• 0914C, viderunt in quem pupunxerunt — pupunxerunt for the classical pupugerunt, and viderunt for Jn 19:37's videbunt. Both parse; rendered as printed. See §3.
• 0907C, vos ascenditis ad diem festum — indicative for Jn 7:8's imperative ascendite. The indicative occupies its slot and renders in place ("ye go up"), so Pattern 12 does not apply.

3. Vulgate collation — every italicised lemma, word by word

⚑ This is the deliverable the 7a″ clause exists for, and it was done as the first step of each chunk. Divergences that are material and well-formed carry a [var: …]; the rest are logged here only. Lemma truncation (Migne cutting a verse short, or dropping an opening autem / enim / ergo to start the lemma) is a citation habit, not a divergence, and is not listed.

Marked with [var: …] — 19 in all

Logged, not marked — divergence real but immaterial, or plainly the commentator's own compression

• 0903D quod est, in eo vita erat (Jn 1:3–4 quod factum est, in ipso vita erat) — the famous punctuation crux; factum dropped, ipso → eo.
• 0904D in mundum venit, mundum non cognovit (Jn 1:10 in mundo erat … et mundus eum non cognovit) — the accusative mundum leaves the subject of cognovit the Jews, not the world; rendered as printed ("the world he knew not").
• 0905A Veteri Testamenti gratia novi Testamenti — Veteri for Veteris; rendered as an asyndetic pair, "the grace of the Old Testament, the grace of the New Testament," rather than supplying the pro that "gratiam pro gratia" invites.
• 0905A confessus est, Johannis non esse Christum — Johannis for Joannem; medieval nominative spelling, Pattern 9.
• 0905B coelos apertos (Jn 1:51 caelum apertum) — number.
• 0905B die tertio (Jn 2:1 die tertia) — gender, no English exponent (Pattern 9).
• 0906A testimoniorum nostrorum non accipitis (Jn 3:11 testimonium nostrum non accipitis) — genitive plural; rendered partitively, "ye receive not of our testimonies."
• 0906C fons aquae salientis ad vitam aeternam (Jn 4:14 in vitam); 0909A lava ad natatoria Siloe (Jn 9:7 in natatoria); 0906A Nemo ascendit ad coelum (Jn 3:13 in caelum) — a consistent ad-for-in habit across the section, which is why none of the three is marked: it is this copy's preposition usage, not three separate textual events.
• 0906D/0907A sed qui misit me Patris (Jn 5:30 sed voluntatem eius qui misit me) — rendered "but of the Father who sent me," supplying nothing.
• 0907A lauda te alienus non os tuum (Prov 27:2 Laudet te alienus, et non os tuum) — imperative for subjunctive; the clause will not construe with a nominative alienus, and is rendered as closely as the printed words allow.
• 0907B Jesum ambulantem super mare (Jn 6:19 supra mare).
• 0907C Ascendit Jesus Jerosolymam set against die festo mediante — Jn 7:14 reads ascendit Jesus in templum; Migne's clause is Jn 2:13 / 5:1 verbatim. A conflation of two real Vulgate readings, so no [var:]: the printed words agree with Scripture, just not with the verse in view.
• 0907D Adhuc modicum vobiscum sum (Jn 7:33 Adhuc modicum tempus vobiscum sum).
• 0908C ante Abranam ego sum (Jn 8:58 antequam Abraham fieret, ego sum) — a divergence sitting on top of a broken word; the [sic:] above covers the type, and the syntax is logged here.
• 0909A Lutum fecit, linivit oculos meos — conflates Jn 9:6 (linivit lutum) with 9:11 (lutum fecit et unxit oculos meos). Both words are Johannine; no marker.
• 0909A in judicio ego veni (Jn 9:39 In iudicium ego … veni) — ablative for accusative: "for judgment" rather than "unto judgment."
• 0909C quotquot venerunt fures fuerunt (Jn 10:8 fures sunt).
• 0910A mansit ibi duos dies (Jn 11:6 in eodem loco duobus diebus).
• 0910B qui credit in me et mortuus fuerit vivit (Jn 11:25 etiam si mortuus fuerit, vivet) — the concessive lost; rendered "and shall have died, liveth."
• 0910C expedit unum hominem mori quam tota gens pereat (Jn 11:50) — heavily compressed, and the non of non tota gens pereat is absorbed into the comparative quam, so the sense survives; introduced as reported speech (princeps causas dixit), which is why this is the commentator's compression rather than a variant.
• 0910D invenit asinum (Jn 12:14 asellum); domus repleta est (Jn 12:3 impleta); granum tritici … multa grana adfert (Jn 12:24 granum frumenti … multum fructum affert).
• 0911A princeps mundi mittitur foras (Jn 12:31 eicietur foras).
• 0912C pacem meam do vobis … pacem meam relinquo vobis (Jn 14:27 reverses the order and has pacem relinquo, not pacem meam); venit princeps mundi ad me, et nihil in me invenit (Jn 14:30 in me non habet quidquam).
• 0913A majorem caritatem nemo habet (Jn 15:13 Maiorem hac dilectionem) — caritas for dilectio. Left unmarked: the two are interchangeable in this author's own vocabulary and nothing in the gloss turns on either.
• 0913B sed palam de Patre manifestabo vobis (Jn 16:25 annuntiabo); 0913B/C Pater diligit vos, quia diligitis me (Jn 16:27 amat … amastis).
• 0913D/0914A venerunt cum laternis et fagulis — Jn 18:3 also has et armis, dropped here; the type defect is marked above.
• 0914B sortiamus de illa cujus erat (Jn 19:24 sortiamur … cuius sit) — active for deponent, a period form (Pattern 9).
• 0914C lancea latus ejus punxerunt (Jn 19:34 aperuit) and viderunt in quem pupunxerunt (Jn 19:37 Videbunt in quem transfixerunt) — the commentator has assimilated 19:37 to 19:34's piercing; tense and lexeme both diverge.
• 0914D sudarium … separatim in unum locum involutum (Jn 20:7 separatim involutum in unum locum) — word order only.
• 0915A tibi non ascendi per divinitatem (Jn 20:17 nondum enim ascendi ad Patrem meum) — rendered as printed, "I have not ascended for thee through the divinity." The printed non is carried; the Vulgate's nondum is not substituted.
• 0915A/B mittite in dexteram partem (Jn 21:6 in dexteram navigii rete); viderunt prunas, et piscem et panem positum (Jn 21:9 vident prunas positas, et piscem superpositum, et panem).
• 0915B indivisum rete et non scissum (Jn 21:11 non est scissum rete) — both negations carried.
• 0916A Discipulus quem amabat Jesus — Jn 21:20 has diligebat, but quem amabat Iesus is the Vulgate's own wording at Jn 20:2, so no marker.

4. Left unresolved

1. Nothing in this range has been read at the plate. Nine [sic:] and nineteen [var:] stand on cols 0903–0916B. Five of the nine [sic:] are attributed above to our transcription rather than to Migne, and if the plate confirms that they should be withdrawn rather than kept — in particular 0908A te, which is the weakest of the nine.
2. 0907A ordiatium — worth a lexicon check as much as a plate check: if ordeacium/ordiacium is attested medieval orthography for hordeaceum, the marker comes off under Pattern 9 rather than under a plate read.
3. 0912D/0913A primum for plurimum — the one [var:] in this range whose attribution I would not defend without the plate.
4. No foot-of-page apparatus is recoverable from our source for this work. Several entries above are exactly the class Migne's numbered Forte notes settle (0906A libano, 0912A loc, 0916B in alter). PL 114's spot check is recorded as a sample, not as coverage; if this work's own spot check has not been done, these three columns are where it would pay.

§P. THE BLIND POLARITY READ (2026-09-01)

Three Opus readers, disjoint ranges of six chunks each, covering all 18, each forbidden to open this file until its findings were written down. All three named their clean chunks and their rejected candidates, so a silent pass can be told from a real one.

Yield: 8 defects of OURS found, 7 repaired, 1 left standing and logged. No dropped or added negation anywhere in the work — reader 3 checked roughly 70 printed particles across John and the Capitulatio and found the count exact in both directions. The failure profile here is not negation at all; it is translation-style.md mechanism 5 — a reference or an agent silently re-bound to the contextually easier noun. Five of the eight are that one shape.

Repaired — defects of ours

Left standing, deliberately — and logged so it is not invisible

0914C. ostendit gentes quae in adventum Christi, sive doctrinam illius — the clause is verbless on the plate, and our English supplies one: "the gentiles who believed in the coming of Christ." The supplement is unbracketed, and it does real argumentative work, since the contrast is with the preceding adventum ipsius non receperunt. It is left as it stands because the sense is not in doubt and no marker class fits a verb Migne simply never set — [ed: …] is for text the DIGITIZATION lost, and nothing was lost here. Recorded so that a reader knows the verb is ours and not Migne's.

New Migne-side contradictions — English already faithful, no edit, apparatus only

These are sites where the plate fights itself and the translation correctly carried the printed reading. None was previously logged. They are the edition's evidence.

⚑ Calibration. All three readers independently re-derived a number of already-adjudicated cruces in the same direction the stints had reached (0884C relectis, 0869D falleret, 0876D diabolos eruit, 0866A ipse non renuit, the caus magnum spelling class), which is the best available evidence that this pass was calibrated rather than lucky — and that its NEW findings are new because nobody had looked, not because the reader was inventing.