Liber I Esdrae

Translated 2026-08-23, one translator, 8 chunks (7,986 Latin words), on Opus. verify-english.mjs clean: 8 chunks, columns, notes and sections aligned, no duplicate paragraphs (one warn only — chunk 0000's ratio 1.60, discussed at §6).

Apparatus fired: 9 [var:] · 1 [cj:] · 0 [sic:] · 0 [d:] · 0 [ed:] · 0 [nt:] · 138 [n:] (exact parity with the Latin twins, verbatim, in order). Structure: PROLOGUS + CAPUT PRIMUM → CAPUT X, eleven heads, all present, all translated. Guillemets: zero in either language (this book's lemmata are italic, as Ruth's are). Question marks: 3 in the Latin, 3 in the English, band for band. Italic spans: 271 in the Latin, 272 in the English. All 271 stand in the same places; the one extra is the italic carry of mala at 0695B (§5), the single Latin word in this work with no English rendering. Asterisks inside [var:] and [cj:] content are ours and are not spans.

Ruth (8968) §1 governs unchanged: VERS. n.-- passes through verbatim, the lemma is Englished, a fragmentary lemma is rendered as the phrase it stands for, the attribution sigla are ordinary [n: …] notes with contents unanglicized and Migne's own internal spacing reproduced character for character ((BEDA ubi supra. ) with the space, and at 0708C (BEDA ubi supra.) without it — both as printed). 8978 §6 / 8970 §1(b) govern the abridgment formula: usque ad → as far as, in its own italic span, no verb ever supplied to complete a resumption tail.

§1 · THE PROLOGUE HEAD WITH NO BODY — the two questions the launch brief named

Chunk 0000 opens with ## B. HIERONYMI IN ESDRAM ET NEHEMIAM PROLOGUS., then a single note [n: ( Vide tomo VIII operum. )], then ## CAPUT PRIMUM. and the gloss. No prologue text.

(a) Pattern 15 — the note stays [n:]. It is a locator, not a statement.

The call: [n: ( Vide tomo VIII operum. )] is a citation and is NOT converted to [nt:].

Judged by content, per Pattern 15, not by length. The whole content after Vide is a place-reference — a volume number in an edition (tomo VIII operum, i.e. Migne's Jerome). It contains no finite verb making an assertion, no proposition, nothing an English reader loses by meeting it in Latin. It is a see-under, the same speech-act as (ID., ibid.) or (BEDA, ubi supra. ), and those are the paradigm [n:].

⭐ This is not my judgment alone: the corpus has already ruled it, three times, on three sibling books of this same recension and volume, and I found the ruling rather than re-deriving it.

8948's cruces.md §3 states the reason from an actual plate read, and its §3 of the guillemet section contrasts the two cases inside one book: its opening Vide note stays [n:], its closing note at 0748C — editorial prose, a statement that the old copies carry no exposition past that point — was converted to [nt:]. Two notes four words apart, decided opposite ways by content. 9005 (Baruch), the precedent the launch brief named, converted its note for the same reason and said so explicitly: that note "names no place a reader can go", where 8948's "names a place (inter opera … tom. VIII)". Mine names one.

The rival reading, and why it lost. Vide is a second-person imperative addressed to the reader, and one could argue that any note which tells the reader to do something is editorial prose. It lost on two counts. First, Pattern 15's line is locator vs. statement, not indicative vs. imperative — mood is not the test, and taking it as the test would convert every cf., vide and ubi supra in the corpus. Second, and decisively, the corpus has three shipped instances of this exact note-type in this exact position in this exact recension, all [n:], one of them plate-verified; ruling against them here would make I Esdrae the only book of the four to translate its cross-reference, for no reason a reader could see.

(b) Absent from the PLATE, not from our digitization. No [ed:].

I believe the prologue's text is absent from Migne's plate, and the evidence is on the page. The note is not a stray: it is doing the work of a see-under. Migne prints the head so the book's structure is complete, and then sends the reader to Jerome's own works rather than reprinting the prologue. A digitization that lost a paragraph would not have invented a redirection notice pointing at where the lost paragraph is printed instead.

8948 settles it as a class from the plate, and I am quoting its finding rather than guessing at mine: p. 375 of that volume prints, in order, the rule · LIBER ESTHER. · the Jerome prologue head · (Vide inter opera B. Hieronymi, tom. VIII.) in parentheses beneath it · a short rule · CAPUT PRIMUM. — and only then text. That is the same layout our chunk 0000 carries, in the same volume of PL, for the same series of prologue heads. 8948 logged it as "no lacuna is logged, none exists."

⛔ So [ed: …] is NOT marked, and Pattern 13's own boundary is why: [ed:] is for demonstrated source loss, and nothing here demonstrates loss. I can show that the plate carries a redirection; I cannot show that it carries a prologue.

⚑ What would settle it beyond the parallel, and it is cheap: a read of cols 0691–0692 (PL 113), which is one page image. If the plate prints the two heads with only the parenthetical note between them, the record is closed with a positive rather than an analogy. I do not need it to ship — nothing in the English or the apparatus turns on it — but it is worth having, and per brief §4 a checked negative is worth recording precisely because three of that class came back positive. I am asking for it as corroboration, not as a blocker.

§2 · THE FULL LEMMA COLLATION — all 271 spans, agreements included

Collated against sources/vulgate/clementine-flat.txt. Migne's book is Esr; Neh is his II Esdrae and no lemma in this work resolves there — every verse lemma in all ten chapters keys to Ezra, and I checked that rather than assuming it, because the prologue head names both books.

⚠ The launch pre-scan's ✓ marks are substring hits and several are false comfort. Two of this work's nine [var:] sit on spans the pre-scan marked ✓: Aedificemus vobis matched because vobis is a substring of vobiscum, and Omnis multitudo quasi vir unus matched because those four words occur together at Esr 3:1 and 3:9 — just not at Esr 2:64, which is the verse being glossed. That is exactly the trap the brief's header warned of, and it fired twice.

2.1 · DIVERGENT — marker fired (9 [var:])

2.2 · DIVERGENT — logged, no marker, and the reason in each case

The bar I used: a [var:] is a public claim, so it is spent on divergences that are more than word order, more than a connective, more than orthography, and more than a lemma's own clipping. Twenty-five divergences failed that bar and are recorded here instead. Every one of them is rendered as Migne prints it in the English.

• 0693A · 3 Jo 1:8 — Migne Nos ergo debemus hujusmodi suscipere / Vulg. …suscipere hujusmodi. Word order only.
• 0695B · Ps 35:7 — Migne Homines et jumenta salvabis Domino / Vulg. …salvabis, Domine. Dative for vocative. ⚠ It turns on one letter, and sources/vulgate/README.md says a one-letter call should say so rather than trust this copy to that precision. Rendered "thou wilt save for the Lord", which is what the printed case means.
• 0696B · Apc 2:5 — Migne Memor esto unde excideris / Vulg. Memor esto itaque unde…. A connective dropped at the head of a quotation.
• 0696D · Mt 28:10 — Migne Ite, nuntiate fratribus meis / Vulg. ite, nuntiare fratribus meis. The Clementine's own form is the odd one (infinitive for imperative); a [var:] here would read to a reader as if we were flagging a typo of ours.
• 0697A · Hbr 13:14 — Migne Non habemus / Vulg. Non enim habemus. Connective, head of quotation.
• 0697C · Esr 3:5 — Migne's lemma In Kalendis, quam in universis, etc. / Vulg. tam in calendis quam in universis solemnitatibus. The clip strands the quam.* This is the abridgment cutting a lemma, not a text variant; rendered "On the Kalends, as on all, etc.", which is the correlative the fragment opens (Ruth §1.2). Kalendis/calendis* is orthography and the gloss expounds the word at length in that spelling.
• 0698A · 1 Mcc 4:47 — Migne secundum illud quod prius fuit / Vulg. …quod fuit prius. Word order only.
• 0698D · 1 Ptr 4:8 — Migne mutuam in vobismetipsis continuam charitatem habentes / Vulg. mutuam in vobismetipsis caritatem continuam habentes. Word order plus charitas/caritas.
• 0698D · Col 3:14 — Migne Super omnia autem habete charitatem / Vulg. Super omnia autem haec, caritatem habete. Word order, and haec dropped. Nothing asserted differently.
• 0699A · Ps 28:5 — text agrees verbatim; Migne's citation reads (Psal. II). Reproduced as printed. Citation repair belongs to data/citation-corrections.json at index time, never to the running text (Ruth §2, 0534C).
• 0701A · Mt 10:34 — text agrees verbatim; Migne's citation reads (Matth. XVI) for Matt. 10. Same treatment.
• 0702A · Ps 128:4 — Migne Dominus justus concidet cervices peccatorum / Vulg. …concidit…. Future for perfect. ⚠ One letter; see the Ps 35:7 entry. Migne's citation reads (Psal. XII.) for Ps 128. Rendered "will cut", as printed.
• 0702B · 1 Thes 4:15 — Migne in adventu Domini / Vulg. in adventum Domini. One letter, ablative for accusative, sense identical.
• 0702D · 3 Reg 7 — Migne quotes Fecit atrium, etc. citing (III Reg. VII). No verse of 3 Rg 7 contains that phrase; the nearest Clementine matches are Ex 38:9 Fecit et atrium and 2 Par 4:9 Fecit etiam atrium sacerdotum. A two-word catchword plus a citation misprint; rendered as printed, no marker, because I cannot say which verse he meant and a [var:] would have to name one.
• 0703A · Esr 6:8 — Migne Sed a me praeceptum est quod / Vulg. Sed et a me praeceptum est quid oporteat fieri. et dropped; quod/quid is one letter; the lemma is clipped at quod with an etc. anyway.
• 0703C · 1 Tim 2:1–2 — the densest cluster in the work and still below the bar: Migne Obsecro ergo primo omnium fieri obsecrationes et orationes, postulationes gratiarumque actiones pro omnibus hominibus / Vulg. Obsecro igitur primum omnium fieri obsecrationes, orationes, postulationes, gratiarum actiones, pro omnibus hominibus. Four differences, all connective, inflectional or punctuational, none of which changes what is asserted. Logged in full so a later reader can weigh it himself.
• 0706A · Esr 7:10 — Migne's lemma Esdras autem paravit / Vulg. Esdras enim paravit. Declined deliberately, following the Judith (8960) precedent the brief names: an autem-for-tamen lean there was reported as wrong, and the standing rule is that a connective swap does not change the sense and does not earn a public claim.
• 0706D · Esr 7:13/14 — Migne's cuicunque for the Clementine's cuicumque is n/m orthography (Pattern 9). ⚠ Separately: Migne heads it VERS. 14.--, but ut cuicumque placuerit is Esr 7:13; his VERS. 14 words (Et septem consiliatorum, Missus es) then follow without a new address. His numbering, reproduced as printed; the Glossa's verse addresses are not ours to renumber (brief: "Migne's own printing is not yours to tidy").
• 0707B · Ps 79:15 — Migne Domine Deus virtutum, convertere, respice de coelo / Vulg. Deus virtutum, convertere, respice de caelo. A vocative Domine added — Ps 79:20's opening contaminating 79:15. Rendered with the Domine.
• 0704D · Ps 115:7–8 — Migne Dirupisti, Domine, vincula mea / Vulg. Dirupisti vincula mea. Same class as the entry above, same treatment.
• 0709C · Esr 8:28 — Migne Vos sancti Domini estis / Vulg. Vos sancti Domini, et vasa sancta. A copula supplied and the verse cut. The copula is what English needs anyway, and nothing is asserted that the Vulgate does not assert.
• 0710B · Esr 8:35 — Migne Obtulerunt holocaustomata Deo / Vulg. obtulerunt holocautomata Deo Israel. holocaustomata is an attested spelling of the same word (Pattern 9); Israel falls to the etc.
• 0710D · Jac 1:18 — Migne Voluntarie genuit nos verbo, etc. / Vulg. Voluntarie enim genuit nos verbo veritatis. Connective dropped, remainder under etc.
• 0694A · Esr 2:58 — Migne heads the gloss VERS. 58.-- Nathinaei filii, but Esr 2:58 reads omnes Nathinaei, et filii servorum Salomonis; the sequence Nathinaei … filii stands at Esr 2:43 (Nathinaei : filii Siha). The second lemma of the same gloss, Omnes Nathinaei, is 2:58 exactly. Rendered as printed ("The Nathinaei, the sons, etc."). No marker: I cannot tell whether Migne is clipping 2:43's words under a 2:58 address or printing a variant of 2:58, and a [var:] would have to choose.
• 0699C · Esr 3:8 — Migne Et omnis qui / Vulg. et omnes qui venerant de captivitate. Singular for plural in a two-word catchword; the same singular stands at Esr 1:5 (omnis cujus Deus suscitavit spiritum). Rendered "And every one who."

2.3 · CHECKED AND IN AGREEMENT — the verbatim list

Recorded because a findings-only report cannot be told apart from a report by an agent that never looked. Each of the following was read word for word against the Clementine verse it addresses and agrees with it exactly (allowing only Migne's punctuation and the etc./usque ad clipping).

Esr 1 (chunk 0000) — 1:2 Et ipse praecepit mihi ut aedificarem ei domum · 1:3 Quis est in vobis · 1:3 Ascendat in Jerusalem · 1:4 Et omnes · 1:4 Adjuvent · 1:5 Et surrexerunt · 1:5 Principes · 1:5 Ut ascenderent · 1:6 Adjuverunt manus eorum in vasis · 1:6 Et jumentis · 1:8 Protulit autem · 1:8 Et annumeravit · 1:9 Et hic est.

Esr 2 (chunk 0001) — 2:1 Hi sunt · 2:1 Et reversi sunt in Jerusalem · 2:1 Unusquisque in civitatem · 2:2 Numerus · 2:58 Omnes Nathinaei · 2:61 Et de filiis sacerdotum · 2:65 Et in ipsis · 2:66 Muli eorum · 2:68 Sponte obtulerunt in domum Dei ad · 2:69 Secundum vires · 2:69 Auri solidos · 2:70 Habitaverunt ergo sacerdotes et Levitae · 2:70 Universusque Israel. Also Ps 138:16 Imperfectum meum viderunt oculi tui (verbatim) and Apc 3:11 Tene quod habes, ut nemo accipiat coronam tuam (verbatim).

Esr 3 (chunk 0002) — 3:1 Jamque venerat mensis septimus · 3:2 Et surrexit Josue · 3:2 Et fratres ejus sacerdotes · 3:2 Et aedificaverunt · 3:3 Bases suas · 3:3 Et obtulerunt · 3:4 Feceruntque · 3:4 Solemnitatem tabernaculorum · 3:4 Per ordinem · 3:6 A primo die · 3:7 Dederunt autem · 3:7 Latomis · 3:7 Caementariis · 3:7 Sidoniis Tyriisque (both occurrences) · 3:7 Ut deferrent ligna · 3:7 Joppe · 3:8 Anno autem · 3:8 In Jerusalem · 3:8 A viginti annis · 3:9 Stetitque · 3:10 Steterunt sacerdotes in ornatu suo · 3:11 Et concinebant in hymnis · 3:12 Flebant. Also 1 Cor 6:19 Membra vestra templum sunt Spiritus sancti and Ps 95:12 Exsultabunt omnia ligna silvarum (both verbatim).

Esr 4–5 (chunk 0003) — 4:1 Audierunt autem hostes Judae et Benjamin · 4:4 Populi Judae · 4:5 Conduxerunt · 4:6 In regno autem Assueri · 4:14 Laesiones · 4:15 Quoniam urbs · 4:16 Trans fluvium · 4:17 Verbum misit rex ad Reum Beelteem · 4:17 Habitatores Samariae · 4:23 Itaque exemplum · 5:1 Prophetaverunt · 5:2 Zorobabel · 5:2 Et Josue · 5:3 In ipso · 5:4 Ad quod respondimus · 5:6 Exemplar epistolae · 5:8 Lapide impolito · 5:8 Et ligna ponuntur in parietibus, opusque. Also Mt 10:34 and Lc 12:49, both verbatim (see 2.2 for the citation misprint on the first).

Esr 6 (chunk 0004) — 6:1 Tunc Darius · 6:3 Ut ponant fundamenta · 6:5 Reddantur · 6:6 Nunc ergo · 6:8 Ut de arca · 6:9 Vitulos et agnos · 6:9 Secundum ritum · 6:10 Orentque · 6:11 A me ergo · 6:12 Deus autem · 6:14 Juxta prophetiam · 6:14 Et Artaxerxe · 6:15 Mensis Adar · 6:16 Fecerunt autem · 6:17 Et obtulerunt · 6:17 Hircos caprarum · 6:17 Pro peccato · 6:18 Et statuerunt · 6:18 Sacerdotes · 6:19 Fecerunt autem filii · 6:22 Et fecerunt solemnitatem · 6:22 Septem diebus · 6:22 In laetitia. Also Ps 133:1 Qui statis in domo Domini, in atriis domus Dei nostri (verbatim).

Esr 7 (chunk 0005) — 7:1 Post haec autem · 7:1 Post haec · 7:7 Et ascenderunt · 7:9 Quia in primo · 7:9 Juxta manum Dei · 7:12 Artaxerxes rex · 7:13 A me decretum · 7:14 Et septem consiliatorum · 7:14 Missus es · 7:14 Quae est in manu · 7:15 Et ut feras · 7:15 Cujus in Jerusalem · 7:16 Et omne argentum · 7:19 Vasa quoque · 7:21 Et a me · 7:22 Sal vero · 7:24 Vobis quoque · 7:25 Tu autem · 7:28 Et ego confortatus. Also Ps 56:8 Paratum cor meum, Deus, paratum cor meum, Ps 39:8 In capite libri scriptum est de me, Is 45:1 Haec dicit Dominus Christo meo Cyro, Ps 11:7 Eloquia Domini eloquia casta…purgatum septuplum, and 1 Jo 2:2 Ipse est propitiatio pro peccatis nostris — all five verbatim, which is worth saying in the chapter that also carries two of the work's largest divergences.

Esr 8–9 (chunk 0006) — 8:1 Hi sunt ergo principes · 8:15 Congregavi autem · 8:17 Et misi eos ad Eddo · 8:21 Et praedicavi ibi jejunium · 8:25 Appendique eis argentum, et aurum, et vasa consecrata · 8:29 Vigilate et custodite · 8:29 Donec appendatis · 8:31 Promovimus · 8:32 Et mansimus · 8:34 Descriptumque · 8:35 Sed et qui venerant · 8:36 Et elevaverunt populum · 9:1 Non est separatus · 9:3 Scidi pallium · 9:4 Convenerunt (both occurrences) · 9:5 Scisso pallio · 9:5 Curvavi genua. Also Ps 44:17 Constitues eos principes super omnem terram and Pr 31:23 Nobilis in portis vir ejus, both verbatim.

Esr 10 (chunk 0007) — 10:2 De filiis Aelam · 10:2 Et nunc si est poenitentia in Israel · 10:4 Surge · 10:6 Filii Eliasib · 10:6 Panem non comedit · 10:9 Ipse est · 10:9 Et sedit omnis populus · 10:10 Et surrexit Esdras · 10:16 Et sederunt in die · 10:18 Et inventi sunt · 10:18 Et fratres ejus · 10:19 Et dederunt. Also 2 Ptr 2:20 Si enim refugientes coinquinationes mundi and 2 Cor 7:1 mundemus nos ab omni inquinamento carnis et spiritus (the tail ut participes resurrectionis esse valeamus is the gloss's own continuation, not 2 Cor 7:1, and is rendered as printed inside the italic span).

§3 · usque ad — DECIDED PER OCCURRENCE, and two are not the formula

80 occurrences, the highest count in the batch, and the launch brief named this as a trap. Counted and adjudicated one at a time:

• 78 are the abridgment formula — 77 printed usque ad in their own italic span, plus one at 0712B printed usque ad: Si enim refugientes coinquinationes mundi, where the colon and the resumed words stand inside the same italic span. That one is Englished whole, span structure preserved: as far as: If indeed, fleeing the defilements of the world,.
• 2 are ordinary Latin, and both are in chunk 0006:
• 0709B, inside a resumption tail: usque ad et per successores eorum usque ad finem aedificanda est — "even to the end". ⚠ This is the shape most likely to be swallowed: an ordinary usque ad sitting a few words downstream of a formula usque ad in the same sentence.
• 0710C: quos exemplis in bona conversatione instituere vel verbis usque ad coelestium perceptionem praemiorum promovent — "even to the receiving of the heavenly rewards".

No verb was supplied to any of the 78 resumption tails, per 8970 §1(b). Where a tail lands mid-clause, the English lands mid-clause. Four tails are visibly broken by the excision and are left broken; see §4.

§4 · THE ABRIDGMENT'S OWN WRECKAGE — four tails rendered as they stand

These read badly. They read badly because the formula cut words out of them, and the brief's rule is that the formula removes words without suspending grammar — so each was parsed as strict prose first, and only then left alone.

• 0706A — usque ad perpetuam ad coelestem patriam pervenientibus do. The object of do is on the far side of the cut; only its feminine accusative adjective survived. Rendered "as far as a perpetual one I give to those that come to the heavenly country." Nothing supplied.
• 0699D — usque ad Zorobabel vero de tribu regia orti sunt. A plural verb with one named subject: the excision took the co-subject (Bede's sentence pairs Josue's priestly tribe with Zorobabel's royal one). Rendered "as far as but Zorobabel indeed were sprung of the royal tribe" — the plural kept, no second name invented.
• 0705D, closing CAPUT VI — conversis etiam gentilibus, qui olim resistebant [0705D] et ad adjuvandum ejus statum, confirmandam Christianae religionis pacem per mundum. A verbless tail at the chapter's end. Rendered as it stands.
• 0706A, second — sed chronicorum libri successorem ejusdem Xerxis. putant elided by zeugma from Josephus putat in the previous clause. Rendered "but the books of the chronicles, the successor of that same Xerxes", keeping the ellipsis visible.

§5 · THE PLATE CANDIDATES — READ AND SETTLED (orchestrator, 2026-08-24)

⭐⭐⭐ THE PLATE WAS READ THE SAME SESSION. PL 113, archive.org patrologiaecurs04migngoog on disk, page map calibrated against the printed corner numbers before any crop, every crop taken wider than its target, and the whole line read at each site.

All seven readable non-words were ours. Every one is now patched in data/tei-patches/8951.json and the work is re-chunked, so the Latin column prints what Migne printed; verify-english passes and no boundary moved. Not one [sic:] had been fired at any of them, which is what the ship test is for: seven pages' worth of public claims against Migne's plate were never made, and the repair belonged in our Latin all along.

⭐⭐ 0691D — the normalization class reaches five, and it was found answering a different question

The plate prints (Vide tome VIII operum.) — tome, the FRENCH word, in an editorial note by a French editor pointing his reader to volume VIII of Jerome's works. Our file reads the Latin ablative tomo, which construes perfectly and is therefore invisible to every check this edition runs. This is the first normalization found in a NOTE rather than in the body, and the first where the tidying changes the LANGUAGE of the word.

⚑ It was found while settling §1(b) — whether Migne prints the prologue's text at all. He does not, and that is now a POSITIVE READ rather than an analogy from 8948: the plate gives the drop title ESDRÆ LIBER PRIMUS, the prologue head, this note, a rule, and then CAPUT PRIMUM with the exposition beginning. §1's two adjudications both stand, and the [ed:] was rightly declined.

⭐ 0697A magistorum — SETTLED at a second witness, and the suspicion withdrawn

Our archive.org copy could not show this word (see below), so it was recorded as unreadable and deliberately not guessed. A second witness answered it in one page fetch: Gallica bpt6k5505319w, f349 (f = (column + 1) / 2, calibrated first), which carries the full margin and opens col. 697 with magistorum quasi ferramentis fabrorum praepa-rata. Our file agrees with the plate exactly. No patch; the suspicion is withdrawn.

⚑ Worth keeping as a checked negative for a second reason: magistorum is an ordinary genitive plural and never belonged on a list of non-words at all. The non-word class, whose plate reads are 19 for 19, stayed unbeaten — this was a misfiled candidate, not a counter-example.

Why it could not be read in our own copy

PDF p. 354 (cols 697/698) of this scan is clipped at the left margin: the first characters of most left-column lines are cut off, so the plate reads rum quasi ferramentis fabrorum praepa- where the word under test is magistorum. Only its final rum survives. No patch and no guess — an illegible plate is recorded as illegible, exactly as an illegible conjecture is. It fires no marker, so shipping asserts nothing about it; a second PL 113 scan would settle it in one page.

⭐ A false suspicion killed by cropping properly

At 150 dpi the word beside septilormem looked like Prophetia, and a second divergence seemed to be standing there. The 450 dpi crop shows Propheta, agreeing with our file exactly. The crop-wider discipline catches the true findings and kills the false ones by the same motion.

The candidates as the translator filed them

⛔ Zero [sic:] in this work, and that is a decision, not an oversight. The 2026-08-23 addendum measured the split 10 for 10: every suspicion aimed at a NON-WORD came back as OUR digitization's corruption, not Migne's plate — which is precisely the class that made seven false [sic:] withdrawable in that batch, each one publicly blaming Migne for Corpus Corporum's error. Six of this work's seven suspect forms are non-words. I therefore rendered the evident word, logged the crux, and fired nothing, exactly as Ruth's significaRuth was handled (a tei-patch, never a marker).

Non-words — plate candidates, worth the read (eight):

(Eight rows, not six; integramente, Judaeo rum and cuncto rum are the word-division cases 7a⁗-b names as the dangerous shape, and I checked each for swallowed force: none of the three carried a verb, a negation or a conjunction — integra mente is an ablative of manner, Judaeorum and cunctorum are genitives of nouns whose governing words survive.)

Real words — Migne's until a plate says otherwise, per addendum A. Rendered as printed:

• 0695B · mala — Mulus ex asino et equa gignitur, mala ex equo et asina. The sentence needs the name of the reverse cross (Isidore's burdo); mala is a real Latin word with no slot here. ⛔ Carried in italics, untranslated, with NO [sic:] — following Ruth §2's societate fieri ruling verbatim: no marker in this corpus means "Migne prints a word that will not construe", and saying so here beats inventing one for a single site. Both rivals named and neither adopted: mula (which would make the sentence false, since a she-mule is also got of ass and mare) and burdo (which is the right animal and no palaeographic path from mala). Worth a plate read; it is a real word, so the corpus-frequency argument is worthless here and I am not making one.
• 0699D · projectum — Non est dubitandum ibi statum Ecclesiae per spiritum sumere projectum. sumere profectum, "to take its increase", is the phrase the sense wants, and profectum/projectum is one letter. Not adopted. Rendered "takes projection through the spirit", which is what the printed word says and reads as oddly as it should.
• 0704A · capiunt — hanc … placidam pacem semper habere capiunt. cupiunt ("they desire") is the obvious rival and is not adopted; capiunt + infinitive does not construe. Rendered "and take that she should always have quiet peace."
• 0701B · honorum — inchoationem et perfectionem honorum Deo … referunt, where bonorum would be the expected word. A real word that construes perfectly ("of honours") in a sentence about referring one's good back to God. Rendered as printed, no conjecture offered.
• 0708D · ebrius — ebrius per gratiam Dei … mente roboratur. "Drunken through the grace of God" is startling and is also a commonplace of the sober inebriation; nothing is wrong with the word. Rendered as printed.
• 0693B · grado — quem in quo grado ministerii ecclesiastici promoveant. Attested medieval declension shift, Pattern 9. Not a defect, not marked, not reported as one.
• 0700C · accerrime — doubled consonant for acerrime. Pattern 9, same treatment.
• 0694C · nihili — ut nihili prorsus bonitatis … remansisse videatur. Genitive form of nihil, attested. Pattern 9.
• 0707D · delectorum — vasa, id est delectorum corda. delecti = the chosen; the word is right as it stands and electorum is not needed. Raised and rejected, recorded so a later reader knows it was seen.

5.1 · The one [cj: …] — 0707B, Indiam

Mittitur Esdras a facie regis et septem consiliatorum ejus, ut visitet Indiam et Jerusalem…

Indiam is a perfectly good Latin word, so Pattern 7 renders it and Pattern 12 has nothing to wrap — and the English then asserts that Esdras is sent to India, which the author did not say. That is exactly Pattern 18's founding shape, so the gloss is additive: that he may visit India [cj: Indiam; read Judaeam, "Judea"] and Jerusalem.

Why Judaeam wins. The gloss is paraphrasing Esr 7:14, whose text Migne has printed as a lemma two lines above on the same column (Missus es), and the Clementine reads ut visites Judaeam et Jerusalem — with Jerusalem surviving in Migne's own sentence as the second member of the pair. The rival, and it is not frivolous: that Indiam is meant, India standing for the ends of the earth, since the gloss goes straight on to the Gentiles (congregandam de gentibus). It loses because the sentence is not allegorising yet — it is reporting what the verse says, in the verse's own words, and the surviving half of the pair is the literal Jerusalem, not a figure of it.

§6 · WHAT MIGNE PRINTS AND WE DID NOT TIDY

• 0698B · hic in pane et vino, ille proprio filio oblato. Tertiam aetatem Melchisedech et Abraham hic in pane et vino, ille proprio filio oblato sacrarunt. Read as ordinary Latin, hic is the nearer (Abraham) and ille the remoter (Melchisedech) — which assigns the bread and wine to Abraham and the son to Melchisedech, the wrong way round in both directions. ⛔ Rendered as printed ("this one in bread and wine, that one by the offering of his own son"). No reversal, silent or marked: the pronouns are on the page, the demonstrative reference is the author's or his compositor's, and correcting it is the silent-repair class Pattern 7a exists to forbid. A reader with the Latin opposite sees exactly what we saw.
• 0695D · ille … liberatus est, ita hic … ascendunt. A singular subject with a plural verb across a sicut … ita pivot. English has an exponent for that, so Pattern 12's clause (2) does not apply and the disagreement is rendered rather than marked: "so this latter … they ascend as it were to the walls of their own virtue or city."
• 0701B · VERS. 16-- — Migne prints two hyphens with no period after the numeral, alone among this work's fifty-odd verse addresses. Reproduced character for character. (Compare Ruth §2's VERS.9, which was patched — that one was demonstrably the TEI's; this one I have not checked against the plate and am not patching on suspicion.)
• 0711C · the ellipsis inside a lemma. Sic ergo orante Esdra, et implorante Deum et flente . . . . . . . Flevit populus. Seven spaced points, inside the italic span, marking Migne's own excision within the verse. Reproduced exactly, spacing included.
• 0710D → CAPUT IX. The foreign-wives material of Esr 9:1–2 is glossed at 0710C–0711A before Migne prints CAPUT IX., which then opens at VERS. 3. His arrangement; the order is reproduced and not reordered.
• 0708C · the measures. Minimum, librarum quinquaginta medium, librarum sexaginta duo summum, librarum centum et viginti. Corus triginta modii: batos, qui et ephi decima pars cori, id est tres modii. The commas fall between the wrong members — each sits after the ordinal instead of after the weight — so the run reads "The least, of fifty pounds the middle, …". Every numeral is rendered exactly as printed (fifty · sixty-two · a hundred and twenty · thirty · three), and the comma positions are kept, because Pattern 8 governs punctuation as Pattern 7 governs words. A reader who repunctuates gets 50 / 62 / 120 pounds, which is the arithmetic Migne means.
• Numerals, checked one by one. 0706A: Darius annos sex et trigenta, Xerxes annos viginti, Archabanus menses septem, Artaxerxes annis quadraginta — rendered "six and thirty", "twenty", "seven", "forty", i.e. spelled where he spells. 0709A: mille septingentorum prope virorum → "nearly a thousand seven hundred". 0712D: centum anni et amplius → "a hundred years and more". 0697B: quinto decimo die mensis septimi … septem diebus → "the fifteenth day of the seventh month … seven days". 0699C: a viginti annis → "from twenty years". 0708B: sub centenario numero / Centenarius → "under the number of a hundred" / "the number of a hundred". 0701B: tricesimum, sexagesimum et centesimum → "thirtyfold, sixtyfold and hundredfold". No numeral in this work is a Roman numeral in the body text, so the "Roman where he prints Roman" rule had nothing to bite on outside the [n:] notes, which pass through untouched.
• Proper names as Migne spells them, never normalized: Egbathanis / Egbathana (not Ecbatanis) · Medena · Horna Jebusaeus (for Ornan/Areuna) · Archabanus (for Artabanus) · Aelam · Nathinaei, kept undeclined in the English as a proper name rather than given Douay's "Nathinites" · Reum Beelteem · Josedec · Zorobabel · Salathiel · Casleu · Joacim · Eliasib · Aggaeus → "Aggeus" · Malachia → "Malachy". Deviation from a familiar English Bible's forms is data, and 7a″ is the reason none of them was conformed.
• 0697B · scenophegia — Migne's Greek term, kept as he prints it and glossed by his own words ("that is, the fixing of tabernacles"). Not italicized: adding an italic span the Latin does not have would break span parity, and the word is not carried type.
• 0700A · Verba Dierum — Migne's title for Paralipomenon. Rendered literally, "the Words of the Days", not replaced with "Chronicles" or "Paralipomenon".
• 0699D · Flebant (BEDA ubi supra ) — a siglum printed inline in the running text, with an italic span inside the parentheses and Migne's space before the closing bracket, rather than as an [n:]. Reproduced exactly, untranslated, structure and spacing included. It is the only attribution in the work that is not a note.

§7 · SENSE CALLS — where two readings construe, said here rather than chosen silently

• 0707D · Nec tantum ex hoc argento vel auro oblationes … emi possunt. Test 2a's named population: a negative with tantum beside it. Two readings construe. (i) The negative on emi possunt, tantum qualifying the money: "Nor out of this silver or gold alone can the oblations be bought." (ii) The negative on tantum … oblationes: "Nor can only the oblations be bought out of this silver or gold." Reading (i) was taken, and the reason is structural, not euphonic: the quia clause that follows gives as its ground that these marvels can be admired but not imitated — which is a statement about our inability to make them our own offering, not about what else besides oblations might be purchased. Reading (ii) leaves the causal clause with nothing to explain.
• 0707D · quae non omnibus in exempla operis possunt ostendi: sed tantum … recitari. Same population, and here the sed settles it cleanly: the pivot is ostendi ⟷ recitari, so the negative is on ostendi with omnibus as its dative, "which cannot be shown to all as examples of work: but only recounted". Checked, not assumed.
• 0696A · Universum Israel, non decem tribus tantum dicit. Negative + tantum again. The host is tantum, and the quae quondam vocabantur Israel ad distinctionem duarum that follows proves it: the point is that "Israel" is not restricted to the ten. Rendered "he means not the ten tribes only".
• 0712C · non scribit Esdras Jesum alienigenam duxisse uxorem, sed quosdam de filiis et fratribus suis. The sed contrasts the objects (Jesum ⟷ quosdam de filiis), not the verbs, so the negative lands on Jesum: "Esdras does not write that Jesus married a foreign wife, but certain of his sons and of his brethren."
• 0695C · quanto eos nec licitis uti considerant. The negative is on licitis, not on the infinitive: "by how much they observe that these use not even lawful things."
• 0699B · unexpressed subject. sed Salomon regali potentia a rege Hiram quae voluit, facile obtinuit: hi autem de captivitate redeuntes … cuncta emebant. Both subjects are printed (Salomon, hi); nothing was supplied. Named here because the surrounding paragraphs supply "they" freely and I checked each: in every case the antecedent is filii transmigrationis, named in the lemma or in the sentence before.
• 0711B · Magna mutatio rerum perfidia principum, multi luxuriam secuti, reducti sunt. perfidia principum is ablative with no preposition and multi … reducti sunt has no conjunction joining it to Magna mutatio rerum. Rendered as a run of appositional clauses with Migne's single colon preserved before conveniunt. Nothing supplied and no conjunction invented — the Ruth effoetandi error (an invented "and") is the named precedent for what not to do here.
• 0696C · cum ablatas sordes vitiorum … illuminamur. ablatas sordes is accusative where an ablative absolute is wanted. English participial phrases carry no case, so there is no exponent for the slip and Pattern 9's logic applies: rendered "the filth of the vices taken away", logged, not marked.

§8 · READ CLEAN — the sections where I found nothing, said so a blind reader knows

Recorded because a flagged passage is not a checked passage, and the converse also needs saying.

• Chunk 0000 entire (0691D–0693C). 13 notes, 5 anchors, one negative (non solum … sed et, scope on pauperibus … ministrare, checked). No divergence in any of its thirteen Esr 1 lemmata. The only thing in it worth an entry is §1, which is not a defect. ⚠ Its word ratio is 1.60, which trips verify-english.mjs's warn: that is the formula's doing — eight usque ad in 456 words, each of which turns two Latin words into three English ones, plus a head expanded from four Latin words to eight. Not padding.
• 0698A–0698C, the five ages. The densest historical stretch in the work (Abel · Noe · Melchisedech and Abraham · David · Josue and Zorobabel · the sixth age), read twice against the Latin. Everything it asserts, Migne asserts — including the hic/ille inversion at §6, which is the one thing in it I would have "fixed" if the rules allowed.
• 0703C–0704B, CAPUT VI's decree and the Artaxerxes question. Migne prints no ? on Quaeritur quomodo … perficeretur., so Pattern 8b was NOT invoked and the indirect question keeps its full stop. The one ? in the chunk is Quis enim explicare valeat…?, a direct question, and it rides as printed. Checking that before reaching for 8b's colon is the whole point of the rule's boundary.
• 0709A–0709C, the vessels. Checked for the illa … haec antecedent chain (the vessels carried to Babylon vs. those made there) — the demonstratives are printed, the referents are unambiguous, no subject supplied.
• All 138 notes compared string for string against the Latin twins: identical, identically ordered, including the two spacing variants of (BEDA ubi supra. ) / (BEDA ubi supra.), the comma-less (BEDA.,ibid.) at 0699A, the bare (ID.) at VERS. 35, the (AUG.) at VERS. 36, and (BED. in Esdr. ) at 0700C.
• Attribution. Every siglum in this book credits Bede In Esdram except the single (AUG.) at VERS. 36 (band 0710B), on a gloss that reads exactly like its Bedan neighbours. ⚠ Reproduced verbatim, uncorrected. This recension's attributions are known to be unreliable and the work page says so; passing a siglum through is not endorsing it. Flagged, not touched.
• Runbook false positives, checked and NOT reported: column-band arithmetic (item 1) — the anchor run 0691D → 0692D skips the whole of col. 0692's A–C bands and 0700B → 0700C is likewise normal; no text breaks mid-word or mid-construction across any anchor in this work, which is the only real signal. Pattern 4 [f:] tails (item 4) — this is a lemma-and-gloss commentary and none was tagged.

§9 · THE FOURTH TEST — read last, with the counts already clean

Every paragraph re-read against the Latin asking only whether it asserts what Migne asserts — not whether everything is there, which the parity checks had already answered. Four sites were live at that pass, and at every one of them a smoother English was available and was refused. Recorded with the rival named, because a crux that names no rival records nothing:

1. 0695D — ille … liberatus est, ita hic … ascendunt. The smooth reading evens the number out to a singular ("so this latter … ascends"), and nothing on the page would show it. Refused: the plural is Migne's and English can carry it, so it stands (§6).
2. 0698B — hic in pane et vino, ille proprio filio oblato. The smooth reading assigns the bread and wine to Melchisedech and the son to Abraham, which is the history — and is not what the demonstratives say. Refused: this is the silent-repair class exactly, and it would have been invisible to every check we run (§6).
3. 0706A — perpetuam ad coelestem patriam pervenientibus do. The smooth reading supplies the feminine noun the excision took ("perpetual rest I give"). Refused: no word is ever supplied to complete a resumption tail (§4), and requiem is a guess (§4).
4. 0704A — placidam pacem semper habere capiunt. The smooth reading is cupiunt, "they desire", which is very likely what was meant and reads perfectly. Refused: capiunt is a real, well-formed word, addendum A rules that class Migne's until a plate says otherwise, and the conjecture lives in §5 instead of in the text.

⚑ All four are smooth English over rough Latin, which the brief names as a suspect rather than a success. None of the four is a question of what the English contains — anchor parity, note parity, band punctuation and the negation host check were all clean before this pass and all clean after it, and not one of them can see any of these four.

§10 · THE BLIND POLARITY READ (2026-08-24) — five sites, ALL FIVE NEW, all five repaired

A reader fenced from this file read all eight chunk pairs and reported before opening it. This was the richest of the batch's four blind reads: five defects on our side, none of them in this file under any heading, and three of them in stretches the file does not name at all.

⚑ Not one of the five is a negation defect. Every printed negative in this work was walked to its host and found correct — by the translator and independently by the reader. What got through was, in every case, a referent or a relation that English cannot leave implicit: an object mistaken for a subject, a reflexive pointed at the wrong person, an equation flattened into a statement, a real word quietly replaced by the word the context expected. Counts and parity passed at all five.

10.1 · 0701A — the sentence contradicted itself, and the parse was the reason (repaired)

Quia nequeunt sufferre, qui regnum diaboli conantur impugnare, timent bella concitari in Ecclesia, ne haereses et fallaciae eorum … debellentur…

The English made the men who assail the devil's kingdom the subject of nequeunt and timent — and then gave those same men "their heresies and deceits" to fear the loss of. A page that says the assailants of the devil's kingdom fear that heresy will be destroyed asserts the opposite of the gloss. The relative clause is the unexpressed-antecedent OBJECT of sufferre: the heretics cannot endure those who assail the devil's kingdom, and so fear war in the Church, Christ's sword being the thing that would destroy them (duce illo qui ait: Non veni pacem mittere, sed gladium). Now "Because they cannot endure those who strive to assail the kingdom of the devil, they fear…"

10.2 · 0700D — confidentem rendered as confitentem (repaired, and the plate agrees)

Populum Juda, id est confidentem et glorificantem… was Englished "the confessing and glorifying people". Confidens is trusting; "confessing" is confitentem — and the standard Juda etymology (confessio) is exactly what makes that the reading a translator expects, which is why it slipped in as a silent conjecture with the printed word disappearing behind it.

⭐ The plate settles it: Migne prints confidentem. I read this line at 450 dpi this session — it is the same line as the integramente patch, four words along, and the whole line was read because that is the rule. Now "the trusting and glorifying people." ⚑ Two defects of two different kinds on one line of one column, found by two different passes.

10.3 · 0696A — an accusative-and-infinitive reversed (repaired)

…qui de captivis geniti ad ipsas se paterna successione pertinere didicerant. The English had "had learned that these belonged to them" — possession running the wrong way. Se is the accusative subject, ad ipsas the goal: they had learned that they themselves belonged to those cities. Now rendered so. The whole passage is about men returning to a patrimony that claims them, not about property they claim.

10.4 · 0701D — a reflexive that made the man seek himself (repaired)

Unusquisque etiam cum Deum invocans non aliud quaerit quam ipsum… was Englished "seeks nothing other than himself". Ipsum is Deum, four words earlier; the English made the seeker the thing sought, which is the reverse of the gloss and leaves the naming (Salathiel, I asked God) unintelligible. Now "seeks nothing other than him." ⚑ Potest eum Salathiel vocare is left as it stands: the referent is ambiguous in the Latin too, and resolving it would be our choice, not Migne's.

10.5 · 0693C — the lemma word dropped out of its own gloss (repaired)

Lemma: Adjuverunt manus eorum in vasis. Gloss: Manus aedificantium socii, qui aedificare nequeunt datis pecuniis adjuvant… Migne asserts an identification — the hands of those building ARE the associates — and that identification is the only reason the lemma word is being glossed. The English read "The companions of the builders, who are not able to build, help with monies given", which is a true statement about companions and says nothing about hands: the word under gloss had vanished. Now "The hands of those who build are the associates, who, not being able to build, help with monies given…"**

10.6 · Two of Migne's own, found by the reader and previously unlogged — no change

Both are rendered literally, which is correct; they are recorded so the ledger is complete, beside the projectum and capiunt entries §5 already keeps.

• 0707C ut per hoc patrem cum sanctis habeant — "that through this they may have a father with the saints". partem cum sanctis, "a part with the saints", is the phrase the sense wants and is one letter away.
• 0708B ea quae in dextra indicis … sunt gaudia — "the joys which are on the right hand of the index". Almost certainly judicis; this book prints ad dexteram judicis at 0700C for exactly this idea.

⚠ Neither is patched and neither carries a marker, because both are real words and the plate has refuted that class 5 times out of 5 this batch. They are Migne's until a plate says otherwise.