These are the edition's working notes, published as they were written. They record, work by work, where Migne's plate is defective or the reading uncertain: the printed form, what the English does with it, and why. They are not a groomed apparatus. Where a note argues with itself, withdraws an earlier decision, or leaves a question open, that is the record of how the reading was actually reached, and it is left standing — a conjecture that is visible can be checked, and one that has been tidied away cannot.
The English never silently repairs the page. A construable but wrong reading is rendered literally and the conjecture is recorded here, never applied to the text. How the English is produced and checked →
Translated 2026-08-23, one translator, nine chunks, 9,178 Latin words (frontmatter count) → ~13,650 English, 1.47×. verify-english.mjs clean: columns, notes and sections aligned, no duplicate paragraphs.
Apparatus fired: 4 [var:] · 2 [cj:] · 0 [sic:] · 0 [d:] · 0 [ed:] · 0 [nt:] · 188 [n:] (exact parity with the Latin twins). Zero guillemets in either language. Zero em-dashes in the English. Question marks: 15 in the Latin, 15 in the English, and every one of them sits in the same column band as its twin (§5).
Ruth (8968) cruces.md §1 governs everything here — 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 their contents unanglicized. Tobit (8970) §1(b) governs the abridgment formula: usque ad → as far as, and no verb is ever supplied to complete a resumption tail. Nothing in either was re-opened.
§0 · WHAT THIS BOOK MADE ME DECIDE FIRST
The usque ad count, since the brief named it
The Latin prints 74 tokens containing usque. 56 are the abridgment formula and are set in their own italic span, usque ad; all 56 are as far as in the English and none was varied. The other 18 are not the formula and were decided one at a time, from what stands on either side:
| col | printed | reading taken |
|---|---|---|
| 0541A | Usquequo ebria eris? | "How long wilt thou be drunken?" — the lemma's own word (1Rg 1:14) |
| 0543C | a solis ortu usque ad occasum laudant nomen Domini | Ps 112:3's own words, inside a scripture lemma — "from the rising of the sun unto the going down" |
| 0545B | et alii usque ad tempus Christi | ordinary — "and others until the time of Christ" |
| 0545D | lucebat autem usque in mane | ordinary — "but it shone until the morning" |
| 0546A | qui usque hodie sic vocatur | ordinary — "even to this day" |
| 0546A | termini Judaeae tenduntur usque ad Dan | ordinary — "as far as Dan" |
| 0548C | quae usque hodie monstratur | ordinary — "even to this day" |
| 0548C | Ab urbe murata usque ad villam quae erat absque muro | 1Rg 6:18's own words, inside a lemma — "from the walled city unto the village" |
| 0549D | persequuntur eos usque in domum agnitionis: usquequo… | ordinary, twice — "even into the house of acknowledgment: that is, until…" |
| 0550A | quoniam nihil est melius hucusque adjuvat Deus | ordinary — "hitherto" |
| 0550B | et usque hodie ostenditur locus | ordinary — "even to this day" |
| 0545C | de cujus genere usque hodie veniunt | ordinary — "even to this day" |
| 0556A | vexabatur a daemonio usque ad suffocationem | ordinary — "even to suffocation" |
| 0556A | ad quam usque persecutus est Josue | ordinary — "as far as which Josue pursued" |
| 0558A | vestimenta sua a tunica usque ad balteum | ordinary — "from the tunic even to the belt" |
| 0561C | secutus est eam plorans usque Baurim | ordinary (2Rg 3:16's own idiom) — "as far as Baurim" |
⚑ The two dangerous ones are the two that sit INSIDE a scripture lemma — 0543C (Ps 112:3) and 0548C (1Rg 6:18). Reading either as the formula would have cut the verse in half and left the English asserting an abridgment Migne never made. Both were caught by looking up the verse.
The [sic:] that was NOT fired — the ship test applied three times
Three forms in this work are non-words, and the 2026-08-23 addendum's finding is that a non-word is our digitization's error six times out of six, not Migne's. A [sic:] blames Migne in public. None was fired. All three are logged in §2 as plate candidates instead (§2a, §2b, §2c).
§1 · THE LEMMA COLLATION — every italic span, checked against sources/vulgate/clementine-flat.txt
Migne's book is 1Rg in that file, numbering identical, so no Samuel/Kings conversion arises. What follows is the whole inventory, chapter by chapter. A lemma named here with no comment AGREES with the Clementine at its own verse, verbatim, or is a clean clip of it — that is the point of naming it. Everything that diverges carries its divergence.
CAPUT I (0539C–0541D)
Ramathaim Sophim, de monte, (1:1 — Clem. de Ramathaimsophim, de monte Ephraim: Migne divides the name into two words and starts the clip after de; word-division and clip only) · Armathaim, (the gloss's own named variant, "quod in aliis codicibus habetur" — not a Vulgate reading and not collated as one) · Dominus possedit me: (Pr 8:22, verbatim) · Jeroham, · Eliu, · Thohu, · Suph, (all 1:1, verbatim) · Deus meus, Deus meus (Mt 27:46, verbatim) · Unguentum effusum nomen tuum (Ct 1:2 — Clem. Oleum effusum. [var:] FIRED, §3a) · Ephrathaeus, (1:1) · Silo. (1:3) · Sacerdotes Domini (1:3) · Ophni, · Phinees. (1:3) · Venit ergo dies et immolavit Elcana deditque Phenennae, (1:4 — Clem. punctuates Venit ergo dies, et immolavit Elcana, deditque Phenennae uxori suae; commas and the clip only) · Annae autem dedit partem unam tristis, (1:5, verbatim) · Phenennae, · et filiis dedit partes, (1:4 — the gloss's own compression of et cunctis filiis ejus et filiabus, partes; a re-quotation, not a fresh citation) · Dedit eis regiones gentium (Ps 104:44 — Clem. dedit illis; a pronoun synonym, no marker, §4) · dedit partem unam, (1:5) · Tristis, (1:5) · concluserat Dominus vulvam ejus (1:5 — Clem. Dominus autem concluserat vulvam ejus; word order and the dropped autem) · Anna, cur fles? (1:8, verbatim, question mark included) · Dixitque ei. (1:14) · Usquequo ebria eris? (1:14, verbatim) · sacrificium Deo spiritus contribulatus, cor contritum, (Ps 50:19, verbatim; Clem. semicolon for Migne's comma) · Heli ait, Vade in pace, (1:17 — Clem. Tunc Heli ait ei: Vade in pace; clipped and compressed) · Et abiit mulier, (1:18) · Vultusque illius non sunt amplius in diversa mutati. (1:18, verbatim, plural vultus and all) · Facies non concidit amplius, (labelled Alia translatio by Migne — another version, not the Vulgate, and not collated as one) · Et recordatus est ejus Dominus. (1:19) · Vocavitque nomen ejus Samuel. (1:20) · Precorque ut impleat Dominus. (1:23) · Det tibi Deus petitionem, (1:17 — Clem. et Deus Israel det tibi petitionem tuam; reordered, Israel and tuam dropped, §4) · Mansit ergo mulier et lactavit filium suum donec amoveret eum a lacte. Et adduxit. (1:23–24, verbatim across the verse join) · Modio farinae, (Jerome's explicit report of the HEBREW against Clem. tribus modiis farinae — that is the note's whole subject, §4) · Et adoraverunt. (1:28, verbatim — and Migne's verse address for it is VERS. 19, out of sequence after VERS. 23, §2f).
CAPUT II (0542A–0545C)
Dilatatum est. (2:1) · Nunc dimittis servum tuum, Domine, secundum verbum tuum, in pace. Quia viderunt oculi mei salutare tuum (Lc 2:29–30, verbatim; commas only) · Os nostrum patet ad vos, o Corinthii, cor nostrum dilatavum est (2Cor 6:11 — a NON-WORD for dilatatum; §2a) · Non est sanctus. (2:2) · Ego sum qui sum (Ex 3:14, verbatim) · Ut est Dominus. (2:2) · Sed nullus est sicut Dominus, (not a citation — the glossator's own restatement against Clem. 2:2 et non est fortis sicut Deus noster; §2d) · Nolite. (2:3, clipped from Nolite multiplicare loqui sublimia gloriantes; expanded per Ruth §1.2) · Altiora te ne quaesieris, (Sir 3:22, verbatim) · Deus in coelo, et tu super terram: sint pauci sermones tui. (Ecl 5:1 — Clem. Deus enim in caelo, et tu super terram; idcirco sint pauci sermones tui: enim and idcirco dropped, sense untouched) · Vetera, (2:3) · Ipsi praeparantur. (2:3) · Dominus novit cogitationes hominum. (Ps 93:11 — Clem. Dominus scit; a synonym, no marker, §4) · Deus scientiarum. (2:3) · Et infirmi. (2:4) · Miserere mei, Deus, quoniam infirmus sum (Ps 6:3 — Clem. Miserere mei, Domine; a different vocative, no marker, §4) · Saturati prius pro panibus. (2:5 — Clem. Repleti prius, pro panibus se locaverunt, and MIGNE'S OWN GLOSS re-quotes Repleti twice in the next two lines; §2e) · Pleni panibus minorati sunt, et esurientes transierunt terram. (labelled In Graeco ita — the Greek, not the Vulgate) · Et famelici. (2:5) · Sterilis peperit. (2:5) · Sapientia aedificavit sibi domum, et suffulsit columnas septem (Pr 9:1 — Clem. excidit columnas septem. [var:] FIRED, §3b) · Plurimos (2:5) · septem. (the gloss's own Hebrew/LXX comparand) · Pauperem, (2:7) · De pulvere. (2:8) · Nec dabis sanctum tuum videre corruptionem (Ps 15:10, verbatim) · De stercore. (2:8) · quae mihi, / fuerunt lucra, propter Christum aestimavi stercora [detrimenta] (Phil 3:7–8 conflated — Clem. 3:7 haec arbitratus sum propter Christum detrimenta, 3:8 arbitror ut stercora. Migne prints both, his own bracket carrying the second: his apparatus has already adjudicated it, so no marker of ours goes on top — the 10517 ( sic ) principle, runbook false positive 5. The bracket rides verbatim) · Principibus. (2:8) · Sedebitis super sedes duodecim, judicantes duodecim tribus Israel (Mt 19:28 — Clem. sedebitis et vos super sedes; two words dropped, no sense change) · Solium gloriae. (2:8) · Pater omne judicium dedit Filio, et quia regni ejus non erit finis (Jo 5:22 + Lc 1:33 — Migne supplies Pater from Jo 5:22's own first clause and chains Lc 1:33 with his own et quia; a conflation, not a divergent quotation, §4) · Cardines terrae. (2:8) · a solis ortu usque ad occasum laudant nomen Domini (Ps 112:3 — Clem. laudabile nomen Domini. [var:] FIRED, §3c) · Sub quo curvantur, qui portant orbem (Job 9:13, verbatim) · Dominum formidabunt. ×2 (2:10) · Ite in ignem aeternum (Mt 25:41 — Clem. Discedite a me maledicti in ignem aeternum; the received clause abridged to its familiar tag, no sense change, §4) · In coelis. (2:10) · Fines. (2:10) · Et dabit. (2:10) · Imperium. (2:10) · Dabo tibi gentes haereditatem tuam. (Ps 2:8, verbatim) · omne judicium dedit Filio. (Jo 5:22, verbatim) · Regi. (2:10) · Christi sui. (2:10) · Heli. (2:12) · Tunicam. (2:19) · Lac vobis potum dedi, non escam (1Cor 3:2, verbatim, negation included) · Si peccaverit. (2:25) · Patris tui, (2:27) · Aemulum. (2:32) · Projecit Salomon Abiathar ne esset sacerdos Domini, ut impleretur verbum Domini, quod locutus est super domum Heli (3Rg 2:27 — Clem. Ejecit ergo Salomon Abiathar ut non esset sacerdos Domini, ut impleretur sermo Domini quem locutus est super domum Heli in Silo. Four word-level differences and a dropped locative, *all synonymous, and the negation is carried in both (ne esset ‖ ut non esset): no marker, §4) · Ut deficiant. (2:33) · Et animam. (2:35) · Et requievit domus Israel post Dominum (1Rg 7:2 — Clem. et requievit omnis* domus Israel; one word dropped) · Futurum est. (2:36) · Buccellam. (2:36) · Panis quem ego dabo caro mea est pro mundi vita (Jo 6:52, verbatim but for the leading et).
CAPUT III–V (0545D–0548A)
Lucerna. (3:3) · Tinnient. (3:11) · Haec. (3:17) · Et non. (3:18, clipped from et non abscondit ab eo; expanded) · A Dan. (3:20 — Clem. a Dan usque Bersabee, no ad: Migne's clip stops before the point) · Judic um (the etymology of Dan — a SPLIT NON-WORD, §2b) · Bersabee. (3:20) · Et factum. Israel, (4:1 — one italic span carrying two lemmata, kept as one) · Lapidem quem reprobaverunt aedificantes. (Ps 117:22, verbatim) · Philisthiim venerunt in Aphec, (4:1, verbatim) · Juxta lapidem. (4:1) · Aphec. (4:1) · Et tulerunt. (4:4) · Sedentis super. (4:4) · Clamore. (4:5) · Et personuit. (4:5) · De manu. (4:8) · Et arca. (4:11) · Cumque. (4:13) · Cecidit. (4:18) · Mortuus. (4:18) · Uxor. (4:19) · In ipso. (4:20) · Ne timeas. (4:20) · Et vocavit. (4:21 — Clem. prints Et vocabit, a future where the narrative wants a perfect; Migne's vocavit is the ordinary reading. A one-letter divergence, and sources/vulgate/README.md warns that this copy is not to be trusted to that precision: logged, no marker) · Auferetur a vobis regnum Dei, et dabitur genti facienti fructus ejus. (Mt 21:43, verbatim; Migne prints it with no [n:] at all) · ignis patris / piscis tristitiae, (etymologies of Azotus and Dagon — glossed Latin equivalents, Englished, italics 1:1) · Azotum. (5:1 — and the gloss under it, "the place in which Samuel set the stone", belongs to 7:12's lapis adjutorii, not to Azotus; §2g) · Caput. (5:4) · Super. (5:4, i.e. super limen) · Truncus. (5:5) · Quoniam pones eos dorsum (Ps 20:13, verbatim) · Et per. (a truncation, §2h) · omnis caro fenum. (Is 40:6 — Clem. spells foenum; ligature/orthography only) · Circumducatur. (5:8).
CAPUT VI–VIII (0548B–0550D)
Quinque anos aureos, (6:5) · Plaustrum. (6:7) · Duas vaccas. (6:7) · Quae ducit, (6:12) · Bethsames. ×2 (6:12) · Et non (6:12, clipped from et non declinabant neque ad dextram neque ad sinistram — expanded, §2i) · Azotus unum. (6:17) · Ab urbe murata usque ad villam quae erat absque muro, (6:18, verbatim) · Abel. (6:18) · Qui interrogant, interrogent in Abela (2Rg 20:18, verbatim) · Percussit, ×2 (6:19) · Et percussit de populo septuaginta viros: et quinquaginta millia viros, (Clem. 6:19 et quinquaginta millia plebis — but Migne prints this explicitly as In Hebraeo habetur, a report of the HEBREW, not a citation of the Vulgate. Pattern 14 does not reach it; §4) · Miseruntque. (6:21) · Vobis Oportebat primum loqui verbum Dei, sed quia indignos vos judicastis, ecce convertimur ad gentes (Act 13:46 — Clem. Vobis oportebat primum loqui verbum Dei: sed quoniam repellitis illud, et indignos vos judicatis aeternae vitae, ecce convertimur ad gentes. A compression plus a tense change; the sense is the received one, so no marker, §4. Migne's capital O mid-sentence has no English exponent and is normalized, Pattern 9, the Latin column preserving it) · Cariathiarim. (6:21) · Donec inveniam locum Domino, tabernaculum Deo Jacob. Ecce audivimus eam in Ephrata, invenimus eam in campis silvae (Ps 131:5–6, verbatim across the verse join) · Gabaa. (7:1) · Suscipiant montes pacem populo, (Ps 71:3, verbatim) · Eleazarum. (7:1) · Sufficientia nostra ex Deo est, qui idoneos fecit nos ministros Novi Testamenti (2Cor 3:5–6 — Clem. qui et idoneos nos fecit ministros novi testamenti; one word dropped and two transposed) · Annus vicesimus (7:2 — Clem. vigesimus; an attested medieval spelling, Pattern 9, no marker) · Annus vice. (7:2 — the same lemma truncated mid-word; expanded, §2j*) · Masphath. ×2 (7:6) · Hauseruntque aquam, (7:6) · Agnum lactantem. (7:9 — Clem. agnum lactentem; a spelling, Pattern 9) · Betchar. (7:11 — Clem. Bethchar) · Et posuit eum. (7:12) · inter Masphath novam et veterem et vocavit nomen ejus Abenezer, / Abongener, id est lapis adjutorii. (labelled Alia editio — another edition, not the Vulgate, and not collated as one) · Hucusque auxiliatus. (7:12) · Eratque. (7:14) · Galgala. (7:16) · Aedificavit etiam ibi, (7:17) · Factum est, / posuit filios. (8:1) · Abia. (8:2) · Et non ambulaverunt.* (8:3).
CAPUT IX–XIII (0551A–0553D)
Vocabulo Saul. (9:2) · Ecce puer meus electus quem elegi (a conflation, not a divergence: Ecce servus meus is Is 42:1, quem elegi is Is 44:1–2, and Migne's own [n: (Isa. XLIV)] points at the second. §4) · Quis similis Deo in filiis Dei ? (Ps 88:7 — Clem. similis erit Deo in filiis Dei?; Migne turns the clause into a direct question by supplying Quis and dropping erit. The question mark is his, printed OUTSIDE the italic span with his own space before it, and both ride as printed) · Stateris. (9:8) · Demus homini. (9:7) · Olim in Israel. (9:9) · Videns. (9:9) · Clivum, / egredientes ad hauriendam aquam. (9:11 — the column anchor [0551C] falls INSIDE this lemma in Migne and is kept inside it in the English) · Et unges eum ducem. (9:16) · Quare non timuisti mittere manum in Christum Domini (2Rg 1:14 — Clem. manum tuam ut occideres christum Domini; compressed, negation intact, no sense change) · Omnia quae. (9:19) · Assumens ita. (9:22 — Clem. Assumens itaque: Migne's lemma stops mid-word; rendered as the word it stands for, §2k) · Triclinium. (9:22) · De industria, (9:24) · Cumque descenderent. (9:27) · Tulit autem. (10:1) · Et prophetabis cum eis. (10:6) · Quando ergo. (10:7) · Septem diebus, (10:8) · Et venerunt. (10:10) · Et quis. / Et quis pater eorum (10:12) · Non habemus regem nisi Caesarem (Jo 19:15, verbatim) · Samuel (the etymology) · Et respondit. (11:2) · In Bezech. (11:8) · Dispersi sunt. (11:11) · Inimici tui, Domine, peribunt et dispergentur (Ps 91:10 — Clem. Quoniam ecce inimici tui, Domine, quoniam ecce inimici tui peribunt; et dispergentur omnes qui operantur iniquitatem: the verse's doubling and its subject dropped. A compression of a famously doubled line, no sense change, §4) · Messis tritici. (12:17) · Voces. (12:17) · Filius unius anni, (13:1) · Gabaa. (13:2) · Absconderunt se in speluncis. (13:6) · Stulte egisti, nec. (13:13, clipped at the conjunction; expanded, negation carried) · Quod si non fecisses. (13:13) · Et nunc regnum tuum non stabit tibi. (13:14 — Clem. sed nequaquam regnum tuum ultra consurget. [var:] FIRED, §3d) · Sed nequaquam. (13:14) · Quaesivit sibi. (13:14) · Virum juxta cor. (13:14).
CAPUT XIV–XVI (0554A–0556A)
Jonathas. (14:1) · Porro Jonathas. (14:27) · Illuminati sunt. (14:27) · Populus peccasset Domino comedens. (14:33) · Occidite super istud, (14:34) · Et occiderunt. (14:34) · Tunc primum coepit aedificare altare Domino, (14:35 — Clem. tuncque primum; the enclitic dropped at the clip) · Applicate huc universos angulos populi (14:38, verbatim; Migne prints no stop after the italic and none is supplied) · Haec faciat. (14:44) · Qui fecit. (14:45) · Et dixit, (15:1) · Bovem, (15:3) · Dixit Saul Cinaeo. (15:6 — Clem. Dixitque) · Poenitet me. (15:11) · Nonne cum. (15:17) · Ducem te constituerunt, noli extolli, sed esto quasi unus ex illis. (Sir 32:1 — Clem. Rectorem te posuerunt? noli extolli: esto in illis quasi unus ex ipsis. Two content words replaced, a sed supplied, the phrase reordered, and Migne's sentence is a statement where the Clementine's is a question. Weighed for a [var:] and declined: the sense is identical throughout, the whole quotation is loose rather than divergent at one point, and Pattern 14 fires on a divergence that changes the sense. §4) · Melior est, (15:22) · Ego non solum ligari in Jerusalem, sed etiam mori paratus sum (Act 21:13 — Clem. Ego enim non solum alligari, sed et mori in Jerusalem paratus sum propter nomen Domini Jesu. ⚠ This is a test-2a population — a negative in a non solum … sed correlative. The non is on solum, not on the infinitives, and the English carries it there: "I am ready not only to be bound … but even to die.") · Non revertar. (15:26) · Ille autem. (15:27) · Scidit. (15:28) · Et in frusta. (15:33 — Clem. prints in frustra, which is the adverb "in vain"; Migne's frusta, "into pieces", is the correct reading and the Clementine's is the odd one. Logged because the direction matters: this is not our divergence to report against Migne) · Ne respicias. (16:7) · Parvulus. (16:11) · Scientem psallere, (16:16) · Refocillabatur Saul. (16:23).
CAPUT XVII–XIX (0556B–0559A)
Congregantes. / Azeca. (17:1 — one italic span carrying the lemma, its [n:] and a second lemma; kept as one) · fortitudo et decipula. / silentium / gaudium. (etymologies of Azeca and Dommim — glossed Latin equivalents, Englished) · Dommim. (17:1) · Terebinthi. (17:2) · Spurius. (17:4) · Octo filios. (17:12) · David. (17:14) · semetipsum exinanivit, formam servi accipiens (Phil 2:7, verbatim) · Direxerat enim aciem Israel, sed Philisthiim. (17:21 — Clem. sed et Philisthiim) · Plenitudo enim legis charitas est (Rom 13:10 — Clem. Plenitudo ergo legis est dilectio. Weighed hard and DECLINED, §4a) · Plenitudo / legis charitas, (the same, re-quoted) · Sufferentes invicem in dilectione, solliciti servare unitatem Spiritus in vinculo pacis (Eph 4:2–3 — Clem. supportantes invicem in caritate, solliciti servare unitatem Spiritus in vinculo pacis. §4a) · Novi superbiam tuam, et nequitiam cordis tui, quia. (17:28 — Clem. Ego novi) · ut videres praelium descendisti, (17:28, verbatim) · Nunquid non verbum? (17:29 — Clem. numquid non verbum est?; the copula clipped, and Migne's own gloss supplies it back: Quasi diceret, est) · Ibat autem. (17:41) · Et tulit. (17:51) · Caput. (17:51) · De qua. (17:55) · Vivit. (17:55) · Benedixit Naboth Deo et regi. Quia non vivat anima tua, rex, si novi, (3Rg 21:13 has Benedixit Naboth Deum et regem, an accusative for Migne's dative; the second half is NOT scripture at all but the glossator's own antiphrastic paraphrase of 1Rg 17:55 with non supplied, which is the very point he is making. §2l) · Si / non (words discussed as words) · Si introibunt in requiem meam (Ps 94:11, verbatim; Migne's [n:] says Psal. XCII) · Si novi, (17:55) · Et factum est, / et dilexit, (18:1) · Ecce nos reliquimus omnia, et secuti sumus te (Mt 19:27, verbatim) · Percussit Saul, / Saul mille. / David decem millia. (18:7) · Tenebatque. (18:10) · Dixitque. (18:17) · Percussit, / et attulit, (18:27) · Cum / plenitudo gentium intraverit, tunc omnis Israel salvus fiet (Rom 11:25–26 — Clem. donec plenitudo gentium intraret, et sic omnis Israel salvus fieret. Migne recasts the subjunctives as futures and the donec … et sic as cum … tunc; the received sense unchanged. The italic is split around a roman enim in Migne and is split the same way in the English) · In agro, (19:3) · Misit ergo. (19:11) · Pellem. (19:13) · In Ramatha. (19:18) · Misit tertios, / Abiit, (19:21–22, clipped from misit Saul tertios nuntios) · Prophetavit. (19:24) · Nudus. (19:24) · Num et Saul. / Num et Saul inter prophetas? (19:24, verbatim, question mark included).
CAPUT XX–XXVII (0559B–0561D)
Uno tantum. (20:3) · Et requisivit. (20:16) · Et addidit. (20:17) · Nunquid: (20:30, not 20:17; §2m) · Achimelech. (21:1) · frater meus rex, (the etymology) · Via. (21:5) · Dedit ergo. (21:6) · Si me persecuti sunt, et vos persequentur (Jo 15:20, verbatim) · In tabernaculo Domini. (21:7) · obligatus in conspectu Domini, (labelled In Heb. — the Hebrew, not the Vulgate) · Doeg Idumaeus. (21:7) · qui victor coelos ascendit. (a resumption tail, not scripture) · Quod cum. (22:1) · Venite ad me, omnes qui laboratis et onerati estis, et ego reficiam vos (Mt 11:28, verbatim) · Aere alieno. (22:2) · Ipse est enim judex vivorum et mortuorum, et princeps pacis, cujus regni non erit finis. (a composite of Act 10:42, Is 9:6 and Lc 1:33, with no [n:] — the glossator's own chain, not a quotation of one verse, §4) · Dixitque Gad. (22:5) · Et dixit Elias Thesbites de habitatoribus Galaad (3Rg 17:1, verbatim) · Descendet. (23:11) · ad fundam jacta, / excellens sibimet. (the etymology of Ceila) · In circuitu impii ambulant (Ps 11:9, verbatim) · Stultus ut luna mutatur (Sir 27:12 — Clem. nam stultus sicut luna mutatur) · Ziph. (23:14) · florentes / germinantes (the etymology) · Eratque ibi, (24:4) · Dirupit Dominus regnum Israel de manu tua. (a conflation of 1Rg 15:28 Scidit Dominus regnum Israel a te hodie with 28:17 scindet Dominus regnum tuum de manu tua; not a single verse diverging at a point, so no marker, §4) · Nos enim quid sumus (Ex 16:7 — Clem. nos vero quid sumus. ⭐ Checked in the file, not from memory — this is exactly the shape of the 8960 autem case the brief warns about, and here the divergence is real but is one particle with no sense change: logged, no marker) · Vivit, / Vivit Dominus, (24:11) · Pharan. (25:1) · Nabal, (25:3) · Gaudium erit in coelo super uno peccatore poenitentiam agente (Lc 15:7, verbatim but for the leading ita) · Sata. (25:18) · Erit anima, (25:29) · Cum digessisset. (25:37) · Phalti filio. (25:44) · Et venerunt Ziphaei, (26:1) · Dormientem, (26:5) · Agebant praedas de Gessuri, et de Gerzi, (27:8, verbatim).
CAPUT XXVIII–XXXI (0562A–0562D)
Samuel autem. (28:3) · Consuluitque Dominum. (28:6) · Neque per sacerdotes. (28:6) · Neque per urim, (labelled In Hebr. — the Hebrew) · Magos. (28:9) · Quem suscitabo. (28:11 — the attribution (AUG., epist. ad [0562B] Simplicianum, tom. II.) is printed INSIDE the italic span as running parentheses, not as a [n:]; it rides verbatim, unanglicized, with the column anchor still inside it) · Percussit Saul. (29:5) · mille, / decem millia (29:5, the words under discussion) · Cumque Amalecitae. (30:1 — Clem. Cumque venissent David et viri ejus in Siceleg die tertia, Amalecitae impetum fecerant: Migne's lemma names the subject the verse holds back. Same case as the Augustine one for the parenthetical attribution (GREG., lib. XXIX [0562C] Moral.)) · Substiterant. (30:21) · Percusserunt Jonathan, (31:2) · Arripuit itaque. (31:4) · Habitatores. (31:11).
§2 · THIS BOOK'S CRUCES
2a · 0542A · dilatavum — a NON-WORD, and the chunk's own lemma settles it. NO MARKER.
Migne (as we have him) prints 2Cor 6:11 as Os nostrum patet ad vos, o Corinthii, cor nostrum dilatavum est. There is no Latin word dilatavum. The Clementine has dilatatum, and — this is the decisive part and it is internal — the lemma at the head of this very chunk, eleven lines above, is Dilatatum est. (1Rg 2:1), correctly formed. A t read as a v is the ordinary shape of a digitization misread.
Rendered "our heart is enlarged", and NO [sic:] was fired. The 2026-08-23 addendum measured non-word suspicions at six for six as our corruption rather than Migne's, and a [sic:] is a public claim that Migne's plate is defective. ⚑ Plate candidate — PL 113 col. 0542A, and the suspect form is a NON-WORD. If the plate reads dilatatum, this is a data/tei-patches/8952.json fix and nothing else.
2b · 0546A · Judic um — a SPLIT NON-WORD. NO MARKER.
Dan interpretatur Judic um vel judicans. The split run is Jerome's Onomasticon etymology of Dan, iudicium sive iudicans — and the second half, judicans, stands unbroken in the same clause, which is what makes the reconstruction safe. Rendered "judgment or judging", italics 1:1, and no [sic:], for the reason in 2a. ⚑ Plate candidate — 0546A, a NON-WORD split across a space. Note that the alternative reconstruction Judicum ("of the judges") was weighed and lost: Jerome's entry is iudicium, and the pairing with judicans is a noun/participle pair from one root, not a book title.
2c · 0539C · quia fuit a de Ramatha civitate Ephraim — two prepositions stacked. NO MARKER.
Both a and de mean "from" and English has one word for the pair, so there is nothing for the English to show; the rendering is "because he was from Ramatha, a city of Ephraim". This is the Pattern 9 shape (a defect with no English exponent) and it is recorded here so a later reader can tell a seen difficulty from a missed one. ⚑ Plate candidate — 0539C; a stray a before de is the kind of thing a plate settles in one look.
2d · 0542B · Non dicit nisi Dominus: Sed nullus est sicut Dominus — which reading was taken
Two construals both parse:
- "He does not say except the Lord" — nisi Dominus as the glossator's own qualification;
- "He does not say 'except the Lord'" — nisi Dominus quoted as a form of words, contrasted with the italic Sed nullus est sicut Dominus that follows the colon.
Reading 2 was taken, and rendered "He saith not, save the Lord: But none is as the Lord, holy and sanctifying". Why reading 1 lost: the Sed is inside Migne's italic span, so it belongs to the quoted matter, not to his own sentence, and a sed inside a quotation with nothing before it to contrast against is exactly what a "he does not say X, but Y" frame produces. The colon Migne prints after Dominus does the same work. Nothing in the argument turns on it either way.
2e · 0542D · Saturati in the lemma against Repleti in the gloss — logged, NO MARKER
Migne's lemma at VERS. 5 is Saturati prius pro panibus. The Clementine's 1Rg 2:5 is Repleti prius, pro panibus se locaverunt — and Migne's own gloss under that lemma opens Repleti prius Judaei and then Repleti iidem, twice. So his lemma word is not the word he expounds; saturati belongs to the second half of the same verse (et famelici saturati sunt), which is the obvious source of the contamination. No [var:]: saturare and replere are synonyms and the exposition is unaffected — Pattern 14 fires where the divergence changes the sense. Recorded because the pattern (lemma and gloss disagreeing inside one paragraph) is the sort of thing a later reader will want to have been seen.
2f · 0541D · VERS. 19 printed after VERS. 23 — Migne's ordering, reproduced
CAPUT I closes with VERS. 19.-- Et adoraverunt. standing after VERS. 23. The text is 1Rg 1:28 (Et adoraverunt ibi Dominum), so the address is wrong as well as out of place. Reproduced exactly as printed and NOT reordered — the Jude precedent in the brief. Migne does the same thing at 0548C (a 6:12 lemma addressed VERS. 13), at 0547D (a 5:6/5:9 lemma addressed VERS. 7, §2h), and at 0559B (a 20:30 lemma standing in the VERS. 17 paragraph, §2m).
2g · 0547C · Azotum. glossed with 7:12's stone — Migne's, left standing
Azotum. Locus est in quo posuit Samuel lapidem [n: (I Reg. VII)], sicut in sequentibus demonstratur. The place where Samuel set the stone is the lapis adjutorii of 1Rg 7:12, not Azotus; 5:1 reads asportaverunt eam a lapide Adjutorii in Azotum, so the gloss has attached itself to the wrong end of the verse. Rendered as printed, nothing emended, no marker — the note Migne himself prints already sends the reader to I Reg. VII, which is where the difficulty is visible.
2h · 0547D · Et per. — a truncated lemma. Rendered "And he smote.", and why
Migne prints VERS. 7.-- Et per. The gloss under it is entirely about being tormented in posteriora — the hinder parts — which is 1Rg 5:6 percussit in secretiori parte natium, repeated at 5:9 and 5:12. Read as a truncation of Et percussit and rendered as the phrase it stands for, per Ruth §1.2.
The rival, and why it lost: per is a whole Latin word, so the lemma could be read as ending on one (in which case it would be carried, not expanded). But 1Rg 5:7 contains no per at all — its words are super nos, et super Dagon deum nostrum — so on that reading the lemma would point at nothing on the page it addresses, while the smiting reading is what the gloss then spends four sentences on. Migne's verse number is off by one either way (§2f).
2i · 0548C · Et non — expanded to "And they turned not aside."
VERS. 13's lemma is Et non, addressed to 6:13, which has no negative in it; 6:12 has et non declinabant neque ad dextram neque ad sinistram. The gloss reads Id est, nec extollebantur prosperis — "neither were they lifted up by prosperous things" — which is the right-hand side of that non-deviation, so the lemma is 6:12's. Expanded per Ruth §1.2, the printed negative carried onto its own host (the turning aside), not onto the being-lifted-up.
2j · 0549C · Annus vice. — the same lemma twice, one of them cut mid-word
0549C prints Annus vicesimus in the running gloss and then Annus vice. as a fresh lemma four lines later. Per the brief's ruling on 8998, a work that treats one lemma two ways is resolved by EXPANDING the stump, never by contracting its twin: both are "The twentieth year". The Latin column keeps the difference visible.
2k · 0551C · Assumens ita. — 9:22's Assumens itaque, cut inside the enclitic
Rendered "Taking therefore." ita alone would mean "thus", which is not what 9:22 says and not what the gloss expounds (Quod Samuel Saulem in excelsum duxit — the taking). Same class as 2j.
2l · 0557D · Benedixit Naboth Deo et regi. Quia non vivat anima tua… — a negation that is the GLOSSATOR'S
This is a test-2a site and it needed care. Migne's italic run contains two different things: a quotation of 3Rg 21:13 (Benedixit Naboth Deo et regi, where benedixit means its opposite) and then the glossator's own antiphrastic paraphrase of the lemma he is glossing — 1Rg 17:55's Vivit anima tua, rex, si novi, with a non inserted to show what the euphemism means. The whole entry exists to teach that Si is sometimes put for non (he proves it from Ps 94:11).
So the non is not a divergence from any received text and takes no [var:]; it is carried into the English on its own host (vivat), and the paragraph's closing Si novi, id est, non novi is what makes the reading certain. The one real divergence here is Deo et regi for the Clementine's accusative Deum et regem — a case, not a sense, so no marker.
2m · 0559B · Nunquid: — expanded to "Do I not know:", and why
The lemma stands in the VERS. 17 paragraph with no verse address of its own. Its gloss is quasi dicat, Amas inimicum meum adversans mihi, in quo ostendis adulterino te conceptum concubitu — which is 1Rg 20:30, Saul to Jonathas: Numquid ignoro quia diligis filium Isai … in confusionem ignominiosae matris tuae? Both halves of the gloss (loving the enemy, the mother's shame) come from that verse and from nowhere in 20:17.
Nunquid alone is an interrogative particle with no verb and is not English, so Ruth §1.2 requires the phrase it opens. The rival — leaving it as a bare stump — lost because it would print an English word ("Whether") that is not a sentence and would tell the reader nothing; the colon Migne prints inside the italic is kept.
2n · 0561A · Satum est genus mensurae continens modium et semis Polenta est accuratissima farina.
A sentence boundary with no point. semis and Polenta run together, and the capital P shows that a full stop belongs there. Pattern 8 forbids supplying it, so the English runs together too: "containing a peck and a half Polenta is the finest meal." ⚑ Plate candidate — 0561A; note that the suspect here is a missing mark, not a word, so no marker of any kind could carry it.
2o · 0561B · Phalti enim vadens dicitur. quando aufertur; — a point mid-sentence, kept
Migne prints a full stop and then a lowercase quando. Preserved exactly (Pattern 8's "never delete"), giving "Phalti is called going. when she is taken away;". The parallel construction two clauses later (Phaltiel … dicitur: quia custodivit) shows what mark was probably wanted; nothing is emended.
2p · 0540D · qui Christum plus nonnunquam diligit — a masculine relative on a feminine subject
The only singular subject that diligit can take is Ecclesia, and quos carnaliter pariebat two words later is unambiguously feminine. qui for quae is a gender mismatch with no English exponent (Pattern 9): the English reads "who sometimes loves Christ more", nothing is emended, and the Latin column shows the form.
2q · 0547A · antequam primitivo ecclesia … credidit — the same class
primitivo is masculine or neuter where ecclesia is feminine. No English exponent, Pattern 9, no marker: "before the primitive church which was of her believed".
2r · 0539C · nec a se aliter efficitur Deus — an ambiguity kept
Rendered "nor is God made otherwise from himself". The rival construal — taking Deus as the subject of an intransitive "nor does he by himself become otherwise God" — is not excluded, and the sentence is arguing divine immutability on either reading (quia nunquam mutatur immediately before). Nothing turns on it; recorded so the choice is visible.
2s · 0549B · In domo Aminadab, qui interpretatur: pater meus spontaneus, et dicere potest voluntarie scilicet tibi.
This will not construe and is rendered as closely as the printed words allow: "In the house of Aminadab, who is interpreted: my father spontaneous, and he is able to say, namely voluntarily, to thee." dicere potest … tibi has no object and voluntarie has nothing to attach to. Two further facts belong with it: 1Rg 7:1 has Abinadab, not Aminadab; and the etymology given (pater meus spontaneus) is Aminadab's, not Abinadab's, so the substitution is in Migne's source, not merely in his type. Nothing emended, no marker — no marker in this corpus means "Migne prints a clause that will not construe", and inventing one for one site is worse than saying so here (the Ruth societate fieri precedent).
2t · 0556D and 0561B · ipsam legem accepit and the two laudabiliters — checked, and NOT defects
At 0556D deposuit tanquam onera legis, et ipsam legem accepit reads oddly beside the preceding deposuit corporalia sacramenta legis, and at 0561B Gregory says Abigail laudabiliter tacuit … laudabiliter dixit — the same adverb of two opposite acts. Both are checked and left alone: the first is the paragraph's own point (he lays down the law's burdens and takes up the law), and the second is Gregory's own antithesis in the Moralia, not a doubling. Recorded as candidates rejected, so a blind reader knows they were looked at.
2u · 0562C · praemium beatitudinis … praecipiunt
praecipio is a real word and "receive beforehand" construes ("they take beforehand from Christ the reward of blessedness together with the perfect"). percipiunt ("receive") is the obvious suspect and is NOT adopted: the 2026-08-23 addendum's finding is that a suspicion aimed at a real, well-formed word has been refuted by the plate every time it was tested, so the printed word stands and the conjecture lives here.
2v · 0561D · nullius hostis deprecationes perpessi — a [cj:] weighed and DECLINED
"having endured the entreaties of no enemy" is what the printed word yields, where the sense plainly wants depraedationes, "depredations" — the villages were quieti et securi. Declined because deprecatio carries the attested medieval sense "imprecation, curse", on which the clause construes as printed ("they had borne no enemy's imprecations"), and because a real, well-formed word is Migne's until a plate says otherwise. Rendered literally; the conjecture is here and not in the text by decision, not by oversight.
2w · Two spellings Migne varies inside the work, both left as printed
- 0540B Ophni … insania conversionis against 0544B Ophni infamia conversionis. Two different etymologies of the same name four columns apart. Both rendered as printed.
- 0545A Sadoch against 0545B Sodoch, twelve lines apart. Migne's forms are kept everywhere, per Ruth's Lia ruling; no name in this work is normalized.
- 0554A Jonathas (4×) against 0554D Jonathan (1×) and 0562C Jonathan. Kept distinct, form for form.
§3 · THE FOUR [var:], AND WHY EACH ONE EARNED A PUBLIC CLAIM
3a · 0539C · Unguentum effusum nomen tuum (Cant. I) — Clem. Ct 1:2 Oleum effusum
[var: Vulg. Ct 1:2 Oleum effusum, "oil poured out"]. A different noun in a proof-text the gloss is building on: the whole entry is an etymology of Suph as effundens, and it chains effudit → effusum → effusum a suis invisibilibus. The reader meets "ointment" where every Vulgate has "oil", and only the marker tells him the divergence is Migne's and not ours.
3b · 0543A · suffulsit columnas septem (Prov. IX) — Clem. Pr 9:1 excidit columnas septem
[var: Vulg. Pr 9:1 excidit columnas septem, "she hath hewn out seven pillars"]. A different verb entirely — propping up against hewing out. The gloss adduces the verse for the seven, so the verb is not load-bearing; the marker fires nonetheless because the two texts assert different acts and the faithful English would otherwise read as a translator's slip.
3c · 0543C · laudant nomen Domini (Psal. CXII) — Clem. Ps 112:3 laudabile nomen Domini
[var: Vulg. Ps 112:3 laudabile nomen Domini, "the name of the Lord is worthy of praise"]. The most material of the four. Migne needs a finite verb here, because he has bent the verse into a relative clause on orbem Ecclesiae, quae …; the Clementine has no verb at all, only a predicate adjective. The divergence is what makes his sentence work, and hiding it would make his grammar look like ours.
3d · 0553C · Et nunc regnum tuum non stabit tibi. — Clem. 1Rg 13:14 sed nequaquam regnum tuum ultra consurget
[var: Vulg. 1Rg 13:14 sed nequaquam regnum tuum ultra consurget, "but thy kingdom shall no more rise up"]. The argument turns on the printed verb. Eucherius (out of Augustine) writes Stetit ergo et stabit quod in illo significatum est; sed non huic stabit — three forms of stare in one sentence, none of which is in the Clementine's consurget. Render the received text and the exposition stops meaning anything.
§4 · DIVERGENCES CHECKED AND DELIBERATELY LEFT UNMARKED
Every one of these is a real difference from the Clementine. None fires, and the reason is the same in each case: Pattern 14 fires where a divergence changes the sense, and a marker on every loose citation would be spam on every column of a book that quotes this freely.
| col | Migne | Clementine | why no marker |
|---|---|---|---|
| 0540C | Dedit eis regiones gentium | Ps 104:44 dedit illis | pronoun synonym |
| 0541B | Det tibi Deus petitionem, | 1Rg 1:17 et Deus Israel det tibi petitionem tuam | clip + reorder |
| 0541D | Modio farinae | 1Rg 1:24 tribus modiis farinae | Migne labels it In Hebraeo — a report of another text, which is the note's whole subject |
| 0541D | agnos agniculos immaculatos septem | Nm 28:11 agnos anniculos | "little lambs" for "yearling lambs"; a real word, and the argument is about tenths of flour, not the age of lambs. Rendered as printed |
| 0542C | Dominus novit cogitationes | Ps 93:11 Dominus scit | verb synonym |
| 0542C | Miserere mei, Deus | Ps 6:3 Miserere mei, Domine | vocative; probably contaminated from Ps 50:3 |
| 0542D | Saturati prius | 1Rg 2:5 Repleti prius | §2e — synonym, and his own gloss re-quotes Repleti |
| 0543C | aestimavi stercora [detrimenta] | Phil 3:7 detrimenta / 3:8 stercora | Migne's own bracket carries the other reading — his apparatus has adjudicated it; ours does not stack on top |
| 0543C | Sedebitis super sedes duodecim | Mt 19:28 sedebitis et vos | two words clipped |
| 0543C | Pater omne judicium dedit Filio, et quia regni ejus non erit finis | Jo 5:22 + Lc 1:33 | a conflation of two verses, not one verse diverging |
| 0543D | Ite in ignem aeternum | Mt 25:41 Discedite a me maledicti in ignem aeternum | the received clause abridged to its familiar tag |
| 0545A | ne esset sacerdos Domini … verbum … quod | 3Rg 2:27 ut non esset … sermo … quem … in Silo | four synonyms and a dropped locative; the negation is carried in both |
| 0545B | Et requievit domus Israel | 1Rg 7:2 et requievit omnis domus Israel | one word dropped |
| 0548D | quinquaginta millia viros | 1Rg 6:19 quinquaginta millia plebis | Migne labels it In Hebraeo habetur |
| 0548D | sed quia indignos vos judicastis | Act 13:46 sed quoniam repellitis illud, et indignos vos judicatis aeternae vitae | compression + tense; the received sense stands |
| 0549C | qui idoneos fecit nos ministros | 2Cor 3:6 qui et idoneos nos fecit | one word, one transposition |
| 0551B | Ecce puer meus electus quem elegi | Is 42:1 + Is 44:1–2 | a conflation; Migne's own [n:] points at the half he took quem elegi from |
| 0551B | Quis similis Deo in filiis Dei? | Ps 88:7 similis erit Deo in filiis Dei? | he makes the clause a direct question; the mark is his and rides as printed |
| 0551C | mittere manum in Christum Domini | 2Rg 1:14 mittere manum tuam ut occideres christum Domini | compressed; negation intact |
| 0553A | Inimici tui, Domine, peribunt et dispergentur | Ps 91:10, with its doubling and omnes qui operantur iniquitatem | compression of a doubled line |
| 0555B | Ducem te constituerunt, noli extolli, sed esto quasi unus ex illis. | Sir 32:1 Rectorem te posuerunt? … esto in illis quasi unus ex ipsis | loose throughout rather than divergent at a point; identical sense. Migne prints it as a statement where the Clementine prints a question, and his punctuation rides as printed |
| 0555B | Ego non solum ligari in Jerusalem, sed etiam mori paratus sum | Act 21:13 non solum alligari, sed et mori in Jerusalem… | compression; the non stays on solum |
| 0556C | sed Philisthiim | 1Rg 17:21 sed et Philisthiim | one word |
| 0556D | legis charitas est | Rom 13:10 legis est dilectio | §4a |
| 0557A | Sufferentes invicem in dilectione | Eph 4:2 supportantes invicem in caritate | §4a |
| 0557B | Nunquid non verbum? | 1Rg 17:29 numquid non verbum est? | the copula clipped — and Migne's gloss supplies it back: Quasi diceret, est |
| 0557D | Benedixit Naboth Deo et regi | 3Rg 21:13 Deum et regem | case, not sense (§2l) |
| 0558B | Cum plenitudo gentium intraverit, tunc omnis Israel salvus fiet | Rom 11:25–26 donec … intraret, et sic … fieret | mood and tense recast; received sense unchanged |
| 0560B | Stultus ut luna mutatur | Sir 27:12 nam stultus sicut luna mutatur | one particle |
| 0560C | Dirupit Dominus regnum Israel de manu tua | 1Rg 15:28 Scidit … a te hodie + 28:17 scindet … de manu tua | a conflation of two verses |
| 0560D | Nos enim quid sumus | Ex 16:7 nos vero quid sumus | ⭐ one particle, checked in the file rather than inferred — the 8960 autem case's exact shape, and here the difference is real but immaterial |
| 0560A | Ipse est enim judex vivorum et mortuorum, et princeps pacis, cujus regni non erit finis. | Act 10:42 + Is 9:6 + Lc 1:33 | the glossator's own chain, printed with no [n:] at all |
| 0562C | Cumque Amalecitae | 1Rg 30:1 Cumque venissent David et viri ejus in Siceleg die tertia, Amalecitae | the lemma pulls the subject forward |
4a · The charitas/dilectio pair — the sharpest of the declines, and it runs BOTH ways
At 0556D–0557A Migne quotes Rom 13:10 as Plenitudo legis charitas est where the Clementine has dilectio, and eleven lines later quotes Eph 4:2 as Sufferentes invicem in dilectione where the Clementine has caritate. He has each of the two words exactly where the Vulgate has the other.
[var:] was weighed on both and declined on both. The reason is not that the difference is invisible — it is that caritas and dilectio are the standard Latin pair for one thing, the exposition here is about the fulness of the law and not about which Latin word names love, and the form Migne prints for Rom 13:10 is the ordinary patristic citation form of that verse. A [var:] that fired here would have to fire on every Latin father who cites Rom 13:10, which is the definition of noise. What is worth recording is the chiasm, and that is what this entry is for; the divergence is now findable by anyone who wants it.
§5 · THE TESTS, RUN, AND WHAT THEY RETURNED
- Marker counts. 188
[n:]per side, in order; nine##heads per side; column anchors identical in count and sequence.verify-english.mjsclean. - Anchor POSITION (test 2b). Every one of the 93 anchors was placed against the WORD it stands beside in the Latin, not merely inserted into the paragraph. Six of them fall inside a construction and were split there deliberately, matching Migne:
[0551C]inside the lemma egredientes ad hauriendam [0551C] aquam,[0553C]inside non [0553C] stabit tibi,[0556C]inside formam [0556C] servi accipiens,[0562B]and[0562C]inside the two parenthetical attributions, and[0555B]inside the Sir 32:1 quotation. Thirteen more sit hard against aVERS.or againstGalgalawith no intervening space in Migne ([0546A],[0546D],[0547A],[0547C],[0548C],[0548D],[0550A],[0550C],[0552A],[0552B],[0553A],[0553B],[0560C]) and are set the same way in the English. - Punctuation, band by band (test 3). Colons, semicolons, question marks, exclamation marks and guillemets were compared inside every column band separately, twin against twin, and the run returned zero mismatched bands across all nine chunks. It did not start there: the first pass had 21 mismatches in 14 bands — mostly English semicolons and colons supplied where Migne prints commas — and every one was brought back to the plate. ⚠ No comma→semicolon sweep was performed: Migne prints 33 semicolons of his own in this book, so Tobit's exception does not apply here.
- Question marks. 15 in the Latin, 15 in the English, same bands. Two of them are Migne's own oddities and ride as printed:
Quis similis Deo in filiis Dei [n: (Psal. LXXXVI)] ?at 0551B, where the mark stands outside the italic span with a space before it, andNos enim quid sumus [n: (Exod. XVI)]?at 0560D. No?was supplied and none deleted. Pattern 8b was not needed: this book prints no indirect question closed with a mark. - Negation, host by host (tests 2 and 2a). Every printed
non · nec · neque · nisi · ne · nequaquam · nunquam · nullus · nihil · minime · nemowas named to its single Latin host and the host found in the English with the negative still on it. Three sites are the 2a trigger population — a negative on a verb with a quantifier or correlative adjacent — and all three were proved from the sentence's own structure, not from what reads well: 0555B non solum ligari … sed etiam mori (the non is on solum; thesed etiamcorrelative proves it), 0562D aut non utitur aut abutitur (the non is on utitur; theaut … autproves it, and putting it on the second verb would make abuse the thing not done), 0550C non aedificari altare, nisi ubi Dominus elegisset (the non is on aedificari and the nisi restricts it; reading the negative onto elegisset would have the law forbid the one place it commands). - The marker-strip read (test 1). All six markers are additive
[var:]/[cj:]; no marker in this work stands in place of a rendering. Every one of the six was read with its contents deleted and the sentence stands complete, with a subject, a verb, and its object. No[sic:]exists to swallow anything. - The fourth test — does each paragraph ASSERT what Migne asserts. Run last, chunk by chunk, with counts already clean. It returned four repairs, all of them mis-parses that had produced plausible English: 0543A multum esuriendo terram transierunt had been demoted to a participle ("passing over the earth"), leaving the sentence with no main verb; 0543A sed ancillae filii, minorati sunt had been split into a new subject ("but the sons of the bondwoman were made less"), which makes the bondwoman's sons a different people from the Israelites named two words earlier, where Migne is calling them the same people under two descriptions; 0557A hoc est enim in carne gratuitum had been softened to "is given freely" for "is gratuitous"; 0540C "he had not sons" was ambiguous in a way the Latin is not. ⚑ All four read perfectly well before they were found, which is the point of the test.
§6 · UNEXPRESSED SUBJECTS — every one supplied, named, and what named it
The brief warned that this is the book where the class bites, and it does: the narrative shuttles between Anna, Eli, Samuel, Saul and David, and between each of them and the figure he stands for. Each of these is a place where the English prints a pronoun the Latin does not.
| col | Latin | supplied | what decided it |
|---|---|---|---|
| 0539C | quia per misericordiam ad nos veniens signatus in passione ait | "he" = Christ | the same note's Elcana allegorice possessio Dei qui est filius Dei four lines above |
| 0539C | de qua veniens, speculationem docuit, et morte sua possidere fecit | "he" = Christ; "[men]" bracketed as the object of possidere fecit | same antecedent; the object is nowhere on the page and is bracketed rather than asserted |
| 0539C | cum soror ejus gauderet numerositate prolis | "her sister" | ⚠ the antecedent of ejus was cut out by the abridgment. No name is supplied; the English says "her sister" and stops where Migne stops |
| 0540C | quia non habebat filios quibus amplius daret | "he" = Elcana | ⭐ the site where the class would have bitten. dedit two words earlier is Elcana's and daret must be his (he is the giver), and nothing signals a change of subject. The rival — "she had no sons", Anna — lost because it requires two subject shifts inside one relative chain, and because 1Rg 1:2's Annae autem non erant liberi is a different construction Migne did not use |
| 0540C–D | tandem redemptor … consolationem adhibuit | none needed; the subject is printed | recorded because the surrounding participles (flenti et non capienti cibum) attach to Annae, and the English keeps them there |
| 0542C | Quia aufert se a cogitationibus | "he" = God | the lemma above is Deus scientiarum, and 1Rg 2:3's ipsi praeparantur cogitationes has God as its ipse |
| 0545D | donec funditus cum prole sua exstingueretur | "he" = Heli | quod Heli in caecitate permaneret is the same clause's subject |
| 0546C | tunc quia fideles immundi Spiritus … conspiciunt | "the unclean spirits" as SUBJECT, "the faithful" as OBJECT | immundi Spiritus is nominative plural and fideles accusative; the resumption immundi spiritus facile prosternunt five lines later has the same subject, spelled out |
| 0554A | Jonathas non nisi provocatus ad hostes transit | none supplied | printed |
| 0554B | Vel comedit cum sanguine, quia ante sacrificium vespertinum comederunt | "he" then "they" | Migne changes number himself, singular then plural, and the English changes with him rather than harmonizing |
| 0541D | si autem dimiserit, vincit Amalec | "he" = Moses; "[them]" bracketed | Ex 17:11's own scene, which the resumption tail is quoting; no verb was supplied to complete the abridgment (Tobit §1b) and the object is bracketed, not asserted |
| 0562B | jam mortuus mortem vivo praenuntiavit | "he" = Samuel | the lemma is Quem suscitabo and the whole Augustine extract is about Samuel's shade |
| 0561C | quia custodivit eum ne tangeret eam | "he" = Phalti | the paragraph names him five times |
§7 · PLATE READS — ALL FOUR RUN AND SETTLED (orchestrator, 2026-08-24)
⭐⭐⭐ Four for four, and two of them are new classes for the register. PL 113 patrologiaecurs04migngoog on disk, page map calibrated against the printed corner numbers before any crop, each crop wider than its target, whole line read.
| site | our file | the plate at 450 dpi | verdict |
|---|---|---|---|
| 0539C | quia fuit a de Ramatha | quia fuit de Rama- / tha | OURS — INTERPOLATED WORD |
| 0542A | dilatavum | dilatatum | OURS |
| 0546A | Judic um | Judicium | OURS |
| 0561A | …et semis Polenta est… | …et semis. Polenta est… | OURS — LOST FULL STOP |
All four are patched in data/tei-patches/8952.json and the work is re-chunked; verify-english passes, no boundary moved.
⭐ 0539C — the digitization ADDED a word
The plate prints one preposition: quia fuit de Ramatha civitate Ephraim. Our file carries a spurious a in front of the de, which is exactly what produced the crux logged above about two prepositions governing one object. This is a new class: not a corruption of a word Migne printed, but the insertion of one he did not. It reads like a scribal doublet rather than like damage, which is what makes it dangerous — a reader meets it as Migne being clumsy.
⚑ The English needed no change. It had rendered "because he was from Ramatha", declining to invent a construction for a word that would not construe — which is what 7a asks for, and which is why the repair could be made in the Latin alone.
⭐ 0561A — the digitization DROPPED A FULL STOP MIGNE PRINTS
The plate reads continens modium et semis. Polenta est accuratissima farina. The stop is there; our file lost it and fused the measure-gloss into the meal-gloss.
⚑ This is the case that justifies spending a plate read on punctuation. Pattern 8 forbids the translator to supply a mark Migne does not print, so running the two sentences together in English and flagging the site was the only correct move available without the plate — and it was made. The plate licenses the repair, the repair belongs in the Latin, and the English is corrected in the same commit to carry the sentence break the plate shows. ⛔ Note what the alternative would have been: supplying the stop in English on the strength of the sense would have been right by accident, and indistinguishable from the many cases where it is wrong.
The sites as the translator filed them
Listed in the order I would spend the reads, with the addendum's classification, because that is what decides whether the answer will be worth anything.
- 0542A · dilatavum — NON-WORD. Highest value: an internal control (Dilatatum est. in the same chunk) already predicts the answer, so a plate read either confirms a
tei-patchor turns up something genuinely unexpected. - 0546A · Judic um — NON-WORD, split across a space. Same class as the six that were all ours in the 2026-08-23 batch.
- 0539C · a de Ramatha — a stray whole word. Not strictly a non-word, but two prepositions with one object is not Latin, and it is the cheapest of the four to settle.
- 0561A · the missing point between semis and Polenta — a missing MARK, not a word. Worth a look only if a reader is on that page anyway; nothing in the apparatus can carry it either way.
⭐ And three sites where the suspect is a REAL, well-formed word, listed so that a later reader knows they were seen and deliberately not raised as plate candidates: 0561D deprecationes (§2v), 0562C praecipiunt (§2u), 0541D agniculos (§4). The addendum's finding — that every real-word suspicion tested against a plate has been refuted, four for four — is the reason, and it is a reason about the strength of the argument, not about whether raising it was wrong.
§8 · THE BLIND POLARITY READ (2026-08-24) — three repairs, and one logged resolution OVERTURNED
A reader fenced from this file read all nine chunk pairs and reported before opening it.
8.1 · ⛔⛔ 0561C — the one site where the reader overturned a decision this file had already made
Phaltiel, id est evadens a Deo dicitur: quia custodivit eum ne tangeret eam, ne fieret transgressor legis.
§6's table resolved the subject as Phalti — "he kept himself" — on the ground that "the paragraph names him five times." The reader disputed it, and the reader is right. Migne prints eum, an accusative object. A Latin sentence whose subject and object are the same man requires se; eum forecloses that reading outright. The unexpressed subject is God, named two words earlier in the very etymology the clause exists to explain — Phaltiel = evadens a Deo — and under the old rendering the -el did no work at all.
Now "because God kept him from touching her." The supply of the name is recorded here as a supply.
⭐ The reason this one matters beyond itself: the losing argument was FREQUENCY, and the winning one was STRUCTURE. That is the same split the batch's plate reads produced — corpus frequency decisive for non-words, worthless against a real form — and the same trap test 2a names: the test validates the mapping you propose, not the mapping the Latin makes. Counting how often a man is named tells you nothing about which case ending stands in front of you. ✅ RULED BY WILSON, 2026-08-24: the overturn stands — God kept him. Structure beats frequency; a count of how often a man is named is not evidence about a case ending. Both readings stay stated here so the rival can be seen, but the text is settled and is not to be reopened.
8.2 · 0543B — a verb annexed from the verse, leaving "the same" with no referent (repaired)
Lemma 1 Rg 2:7, Dominus pauperem facit et ditat. The gloss substitutes Jesum for the verse's pauperem and then resumes the verse's own second verb: Jesum, qui propter nos pauper factus est, et ditat eumdem gloria resurrectionis. The subject of ditat is the verse's Dominus; eumdem is Jesus, its object.
The English had attached ditat to the relative qui — "Jesus, who … was made poor, and enriches the same" — which makes Jesus the enricher and leaves eumdem pointing at nothing. ⚑ The tell is that the finished English does not construe, and it shipped anyway: containment checks cannot see a dangling referent. Now "…and the Lord enriches the same with the glory of the resurrection", which restores the verse's own shape — the Lord makes poor (the incarnation) and enriches (the resurrection). The supplied subject is recorded as a supply.
8.3 · 0549D — a word of the Latin unrendered (repaired)
Agnus oblatus pro populo Dei victoriam praestat… was Englished "The lamb offered for the people affords victory". Dei appeared on neither construal. Now "for the people of God" — populus Dei being this gloss's standing phrase, which recurs unmodified six lines later (coram Israel, id est populo Dei). ⚠ The rival construal, Dei victoriam, "God's victory", is possible and is recorded here; it loses to the work's own usage, and either way the word had to reappear.
8.4 · 0553D — one of Migne's own, previously unlogged: non for noti (no change)
…sic jam non eramus ut in ipso essemus electi ante mundi constitutionem. Augustine's clause (the Quaesivit sibi extract) reads sic iam noti eramus, ut in illo essemus electi — "we were already so known that we were chosen in him". Migne's non for noti inverts the sentence against its own sed ipsi quoque Unigenito, qui venit quaerere quod perierat.
The English renders him literally, which is correct and stays. It is recorded here only so the ledger carries it: a defect of Migne's, reproduced deliberately, not one of ours.
8.5 · Rejections that agreed with this file
The reader independently reached §2's conclusions at praecipiunt (0562C), deprecationes (0561D), the ille/iste pair at 0554C, the anteponit suspicion at 0555C, and humero (0552C); and it confirmed the negation hosts at 0544A and 0542B. Its list of clean chunks names 0000, 0002, 0004, 0005, 0006 and 0008 individually.