MIGNE.APP

Cruces

Liber Ecclesiastes

Anselm of Laon and his school — read the work →

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 2026-08-19, one translator, 4 chunks (3,788 Latin words), on Opus. verify-english.mjs clean: frontmatter verbatim, columns, notes and sections aligned, no duplicate paragraphs, question-mark parity exact.

Apparatus: 15 [var:] · 1 [sic:] · 0 [cj:] · 0 [d:] · 0 [ed:] · 0 [nt:] · 86 [n:] (24 / 29 / 19 / 14, exact parity and order with the Latin twins). Structure: PROLOGUS + CAPUT PRIMUM → CAPUT XII, thirteen ## heads, 1:1. Guillemets: ZERO in either language — this book sets its lemmata in italics, like Ruth, not in « » like Tobit/Judith/Esther. Question marks: 0 / 1 / 3 / 0 in the Latin, the same in the English. None supplied, none deleted. Exclamation marks: 1 in the Latin (1117B quanto magis caetera!), 1 in the English. Column anchors: 11 / 11 / 11 / 9, verbatim and in order. Em-dashes: zero.

§1 · CONVENTIONS — INHERITED, NOT RE-DERIVED

Ruth (8968) §1 governs this book unchanged, and nothing here re-opened it: VERS. n.-- is an address and passes through verbatim; the lemma is Englished; a fragmentary lemma is rendered as the phrase it stands for; the attribution sigla ((HIER.), (ID.), (HIERON.), (ALBINUS.)) ride as ordinary [n: …] notes, unanglicized.

Two inherited refinements that actually decided calls here:

§2 · THE LEMMA COLLATION — ALL 191 SPANS, CHECKED AGAINST THE CLEMENTINE

Collated against sources/vulgate/clementine-flat.txt (Tweedale "Quasimodo", 2005), not from memory. Every lemma in the launch list was looked up. This section names the divergences AND the verified agreements, because a findings-only list cannot be told from a report by an agent that never looked.

2.1 The fifteen [var:] — lexeme-level divergence from the Clementine

The rule I applied, stated so it can be checked: [var:] where Migne's citation differs from the Clementine by a LEXEME — a different word, or a substantive word added or dropped; cruces-only where the difference is word order, orthography, an enclitic particle (enim, autem), a tense/number inflection, or a truncation. One deliberate extension: an order reversal in a formula whose received order is fixed and famous (Dies irae dies illa) takes a marker too, because a reader meeting our English otherwise reads OUR error.

colMigne printsClementinenote
1117CVadit ad austrumEccl 1:6 gyrat per meridiemwholly different verb and noun for the same sense; a Hexaplaric/Old-Latin reading, exactly the class the brief warns of
1117DTerra es, et in terram ibisGen 3:19 pulvis es et in pulverem reverterisearth for dust throughout; and the gloss turns on terra (in terram, unde sumptum est redit), so conforming to Douay would have removed the argument's own keyword
1118BRequievit Deus die septimo ab omni opere suoGen 2:2 requievit die septimo ab universo opere quod patraratDeus supplied, quod patrarat dropped
1118CNon est memoria sapienti cum stulto in aeternumEccl 2:16 Non enim erit memoria sapientis similiter ut stulti in perpetuuma different rendering of the verse, not a truncation; dative sapienti, cum stulto, in aeternum
1118DErroresque et stultiEccl 1:17 erroresque et stultitiamsee §3.1 — the one place I weighed [cj:] and rejected it
1119ADixi ergoEccl 2:1 Dixi ego in corde meoergo for ego; see §3.4
1121CMelius illi erat si natus non fuisset homo illeMatt 26:24 bonum erat ei, si natus non fuisset homo illemelius illi for bonum ei; the negative is identical in both and did not move
1121DCum votum voverisEccl 5:3 Si quid vovisti Deo, ne moreris redderea different clause of the same verse, differently worded
1122AOmnia quaecunque praecepit nobis Deus, faciemusExod 24:3 Omnia verba Domini, quae locutus est, faciemus (cf. 24:7, 19:8)praecepit nobis Deus for verba Domini quae locutus est. Migne's own note reads (Exod. XXIV, XXXII); the Golden Calf is Exod 32, so his double reference is to the promise and the breach, and it is right as printed
1123AMemoria justorum in laude, nomen impiorum putrescetProv 10:7 Memoria justi cum laudibus, et nomen impiorum putrescetplural justorum, in laude for cum laudibus, et dropped
1123CDies illa dies iraeSoph 1:15 Dies irae dies illaorder reversed in the most recognizable clause in the work
1123CImpii non dimidiabunt dies suosPs 54:24 Viri sanguinum et dolosi non dimidiabunt dies suossubject wholly replaced; the negative sits on dimidiabunt in both and did not move
1124BNon est in hominis ditioneEccl 8:8 Non est in hominis potestateditione for potestate. Checked corpus-wide: this is the ONLY occurrence of hominis ditione in all of src/latin/
1125BIn risu faciunt panem et vinum, ut epulentur bibentesEccl 10:19 In risum faciunt panem et vinum ut epulentur viventesbibentes for viventes, and In risu for In risum. The one-letter-class divergence that matters most here: the gloss is about preachers who allure rather than profit, and "the drinkers" is what Migne's plate feasts
1126AQui enim perseverat, salvabiturMatt 24:13 qui autem perseveraverit usque in finem, hic salvus eritpresent for future perfect, salvabitur for salvus erit, usque in finem dropped. ⚑ The same verse is quoted at 1123B in a DIFFERENT form — Qui perseveraverit salvus erit — which agrees with the Vulgate except for the dropped usque in finem and hic, and therefore took no marker. Two forms of one proof-text three columns apart is a fact about this recension worth having

2.2 Divergences recorded WITHOUT a marker, and why

Order, orthography, particle, inflection or truncation only — no lexeme moved, so Pattern 14's "changes the sense" test is not met. Listed because the divergence list is the deliverable.

2.3 ⭐ THE VERIFIED ZEROS — lemmata collated and found to agree EXACTLY

These are the ones the scripture-conformation hunt (§3) was run against and cleared. Naming them is the point: it is the only way to tell this collation from a list of hunches.

Exact against the Clementine, word for word (allowing only Migne's truncation of a verse he is glossing, which is his lemma practice and not a divergence):

Not in the Clementine at all, and correctly so — these are Jerome's own words inside the abridged extracts, not scripture lemmata: Estne verbum (1118B, his rendering of Eccl 1:10), Et quidem possessio armenti (1119B), Omnia enim tempore suo requirentur (1121B), vapulabit multis plagis (1121D, Luke 12:47 with an addition), anian / distensio / Routh (1118C–D, transliterated Hebrew and its Latin gloss), and the Greek περισπασμὸν (1118C), which is carried verbatim.

§3 · THE SCRIPTURE-CONFORMATION HUNT — this book's characteristic risk

The brief named this book the highest-risk in the batch because its verses are the most familiar in English of any Glossa book yet done. Four sites were live and all four were refused. Each is named with what the familiar English would have given.

  1. 1117C Vadit ad austrum. AV/Douay: "He goeth toward the south … turneth about unto the north". The Clementine reads gyrat per meridiem. Migne prints neither, and his verb is vadit. Rendered "It goeth to the south", marked [var:]. ⚑ On the pronoun, see §4.1.
  2. 1117D Terra es, et in terram ibis. Every English Bible has "dust". Rendered "Earth thou art, and into earth thou shalt go", marked [var:] — and the gloss's own argument (humanum genus … quod in terram, unde sumptum est redit) is built on terra, so the conformation would have taken the keyword out of its own proof-text. This is 7914 @1079C exactly.
  3. 1123C Dies illa dies irae. The received order is fixed by the sequence and by every English rendering. Migne prints it reversed. Rendered "That day is a day of wrath", marked.
  4. 1125B ut epulentur bibentes. The Vulgate and every English have "that the living may feast". Migne prints bibentes. Rendered "that they who drink may feast", marked.

Two further sites where the sense rather than a word was at risk:

  1. 1122B Est et alia infirmitas. Douay gives Eccl 5:12 as "another grievous evil", which is the phrase that arrives first. Migne's noun is infirmitas. Rendered "another infirmity", which also keeps it distinguishable from Est et aliud malum at 1122C, where the noun really is malum and the English really is "evil". The two lemmata are five lines apart and the familiar English would have collapsed them into one.
  2. 1124D VERS. 10 Erunt. ⚠ The polarity trap of this book. Eccl 9:10 in the Vulgate is quia nec opus, nec ratio, nec sapientia, nec scientia erunt apud inferos — four negatives. Migne lemmatizes the bare affirmative Erunt and glosses it affirmatively: Id est, valebunt tibi post mortem. Rendered affirmatively, exactly as printed. A reader who remembers the verse will want to restore a negative here; there is none on the plate, and supplying one would be the 11536 defect precisely.

3.1 · 1118D Erroresque et stulti — [var:] TAKEN, [cj:] WEIGHED AND REJECTED

Migne's lemma for Eccl 1:17 reads Erroresque et stulti. The Clementine reads erroresque et stultitiam, and the gloss immediately below is about stultitia (Prima sapientia est stultitia caruisse), which is structural evidence for the Vulgate noun.

Both markers were available and the choice matters, so it is stated. [cj: stulti; read stultitiam] was drafted and rejected: [cj:] asserts that the plate is in error and that our conjecture supplies what he meant, whereas stulti is a perfectly good Latin form and a lemma differing from the Clementine by an inflection is, in a recension full of such divergences (§2.1 lists fourteen more), at least as likely a variant as a compositor's slip. [var:] states both readings and takes no position on cause, which is what Pattern 14 asks and what Pattern 18 does not. The English renders stulti — "And errors and the foolish" — and the Vulgate noun stands beside it in the marker.

⚠ What I did NOT do: treat stulti as a mid-word truncation of stultitiam. Migne truncates lemmata constantly in this work (Nec valet, Et omnia quae, Sed et cunctis) and always at a word boundary; no other lemma in these 191 spans breaks inside a word. That is the evidence against the truncation reading.

3.2 · 1122A Ne temere, Deus — Migne's lemma joins two SEPARATED words of one verse

Eccl 5:1 is Ne temere quid loquaris, neque cor tuum sit velox … Deus enim in caelo, et tu super terram. Migne's lemma is Ne temere, Deus. — the opening two words and a word from the far end of the verse, joined by his comma. His gloss addresses both halves (Quia omnia audit et videt … quamvis ipse in coelo, et tu in terra), which is what shows the lemma is deliberate and not a broken run.

Rendered as printed: "Be not rash, God." No ellipsis was inserted and neither half was completed. Completing Ne temere would have required supplying quid loquaris, and the Esther rule forbids a completion that moves a negative: ne governs loquaris, so any completion relocates it. The stump wins, exactly as at Esther 0741C.

3.3 · Where a fragment WAS completed, and the evidence for each

Completed, because the supply moves no negative and the gloss below names the supplied word:

Left as a stump, deliberately: 1119C Et vidi quia tantum → "And I saw that so far." Eccl 2:13 continues praecederet sapientia stultitiam. Completing it would put three Vulgate words into an italic lemma that Migne does not print — the 7a″ conformation, in its purest form — and the gloss below turns on none of them (it is about the wise man as light and the fool as darkness, and about their common death). The stump is awkward English and it is the honest reading. The Latin stands opposite.

3.4 · 1119A Dixi ergo — [var:], and why not [cj:]

Eccl 2:1 is Dixi ego in corde meo. Migne prints Dixi ergo, one letter from ego, and ergo is a perfectly good word that construes ("I said therefore"). This is Pattern 18's exact shape and [cj:] was considered. Rejected, and on the ground that the faithful English does NOT mislead here: the gloss that follows (Postquam in multitudine sapientiae … tunc non vanitas erit, sed veritas) is a postquam clause, so "I said therefore" is precisely the connective the gloss wants, and nothing in the exposition rests on in corde meo. [cj:] fires where the English asserts something the author did not; here it does not. [var:] records the divergence and leaves the reader both readings.

§4 · THE UNEXPRESSED SUBJECT — every supply, with its evidence

The dominant defect class. Every "he"/"she"/"it"/"they" this English prints that the Latin does not is listed, with what decided it. Where nothing decided it, that is said.

4.1 · Supplies with evidence

4.2 · Supplies REFUSED — the tail is cut and the antecedent is off the page

The usque ad formula removes words, and where it removed the subject I did not invent one. In each case the English carries the verb without a subject pronoun, as a fragment, because the Latin is a fragment.

4.3 · 1119D quia per omnia desideria rapiebat — a supplied object, flagged

"because it hurried [me] through all desires." The clause sits in first-person narration (putabamque hunc esse fructum laborum meorum), the verb is third-person singular, and there is no printed subject and no printed object. I take studio (from the quanto studio ista crescebant two clauses back) as the subject and bracket the object. The rival is that Migne's plate has rapiebat for rapiebar — first-person passive, "I was hurried through all desires" — which construes the sentence effortlessly and is the reading the argument wants. ⛔ I did NOT adopt it, and did not mark it: Pattern 7 renders the printed word, and a one-letter conjecture on a verb whose printed form is grammatical is not a [cj:] case (the faithful English does not assert anything false — it just leaves the agent vague). Named here so a plate reader can settle it in one glance.

§5 · THE BOOK'S OWN CRUCES

5.1 · 1116D–1117A Salomomonis — [sic: Salomomonis], and a PLATE READ IS WANTED

Tradunt Hebraei hunc librum Salomomonis esse agentis poenitentiam. A doubled syllable; not a word.

Corpus-frequency check, which is the right way to raise this (per the brief's test 4): across all of src/latin/ there are 311 Salomonis, 312 Salomon, 90 Salomonem, 17 Salomoni, 17 Salomone — and exactly ONE Salomomonis, here. So the form is unique to this plate or to this transcription of it.

⚠ Which of the two it is cannot be told from the Latin, and that is the whole point of test 4. Migne may have set Salomomonis, or Corpus Corporum may have doubled a syllable the way it metathesized Nephthali on 8970 @0725C. One look at PL 113 col. 1116D settles it, and the answer changes what we do: if the plate reads Salomonis, this is a data/tei-patches/8946.json entry and the [sic:] comes out; if it reads Salomomonis, the marker is right and stays. Filed as the single thing in this work I most want a plate reader to open.

⚑ The [sic:] does not swallow the word. The English reads "this book is Solomon's [sic: Salomomonis], as of one doing penance" — strip the marker and the sentence still names Solomon, still has its verb, and still has its object. Test 1 run on all four chunks: one marker, no stranded article or preposition anywhere, no doubled punctuation.

5.2 · 1119A · the column anchor falls INSIDE a word: Prima sa [1119A] pientia

Migne breaks sapientia across the column, and our TEI puts the anchor at the break. English cannot split "wisdom" at the same point and must not pretend to, so the anchor sits at the nearest true word boundary inside the same construction: "The first [1119A] wisdom is to have been without folly." Precedent: 7871 @0992C, where Cultura assidua ad [0992C] hibita est became "Constant cultivation was [0992C] applied" — the anchor moves to the English word boundary, the word is not broken. Not a defect; recorded so nobody reads the English anchor as displaced.

Every other anchor was checked against the word it stands beside, not merely for presence (test 2b). Three sit mid-construction in the Latin and were split to match: ostendens vanitati [1117B] subjectum → "subject to vanity, [1117B] that is"; Conditor [1117D] aquarum → "the Maker [1117D] of those waters"; nihil [1118B] novum → "nothing [1118B] new"; seminandum [1125D] est → "one must [1125D] sow"; Quasi, [1124C] non debes → "As if, [1124C] thou oughtest not". The anchor inside an italic span at 1121D (usque ad [1121D] finem cap.) is reproduced inside the italic span.

5.3 · 1124D Bellum. Victorum vel etiam imitatores belli — will not construe. RENDERED AS PRINTED.

The gloss on Eccl 9:11's nec fortium bellum reads Victorum vel etiam imitatores belli: a genitive plural (Victorum) coordinated by vel etiam with a nominative/accusative plural (imitatores). The two cannot stand in apposition to each other, and neither agrees with the lemma.

Rendered literally: "Of the victors, or also the imitators of war." No case was silently harmonized. ⛔ No marker, and the reason is the one Ruth stated at 0534A: nothing here is broken type (both are real words in real forms), nothing is missing from the digitization, and no marker in this corpus means "Migne prints words that will not construe together". The obvious suspect is Victores for Victorum, which would give "the victors, or even the imitators of war" — deliberately not adopted, because it is a conjecture and the apparatus is not where the text gets repaired.

5.4 · 1123B Verba adulantis magis insensata ad spinas — rendered as printed

A second italic lemma inside the VERS. 6 gloss, and it is not in the Clementine at all: Eccl 7:7 reads quia sicut sonitus spinarum ardentium sub olla, sic risus stulti. Migne's phrase alludes to the thorns of that verse but is not a quotation of it, and it has no finite verb — moventia (neuter plural, agreeing with verba) carries the predication.

Rendered "The words of a flatterer, more senseless, unto thorns, that is, stirring the cares of the world, which lead to the burnings of Gehenna." The ad spinas is left as the bare prepositional phrase the plate prints; no verb was supplied to govern it. Not marked as a divergence, because it is not offered as a citation.

5.5 · 1120A Nonne melius … quantum sufficit? — Migne's question mark, kept

The only ? in chunk 1 closes a resumption tail (usque ad quantum sufficit?), not a sentence of Migne's own construction. Kept exactly where he prints it. Pattern 8b was NOT invoked: this is a direct question in the tail, not an indirect one, so English takes the mark without recasting. No ? was supplied anywhere in this work and none was deleted — 0/1/3/0 in both columns.

5.6 · The three heads that are not CAPUT n.

B. HIERONYMI IN LIBRUM ECCLESIASTEN PROLOGUS. → THE PROLOGUE OF BLESSED JEROME ON THE BOOK OF ECCLESIASTES., matching Esther's shipped THE PROLOGUE OF BLESSED JEROME ON THE BOOK OF ESTHER. CAPUT PRIMUM. → CHAPTER ONE. (Ruth's form); CAPUT II.–CAPUT XII. keep the Roman numeral. The prologue head has no text under it at all — only the note [n: (Vide inter Opera Hieronymi, tom. III.)], which is Migne's, and which rides verbatim. Not a hole in the digitization; the prologue is represented in this recension by its reference alone, and the first VERS. gloss carries the same reference again at 1115D.

5.7 · Notes: all 86 are CITATIONS, no [nt:]

Judged by content, not length, per Pattern 15. Every note in this work is an attribution siglum or a scripture locator — (HIER.), (ID.), (HIERON.), (ALBINUS.), (Psal. XVIII), (Exod. XXIV, XXXII), (Vide inter Opera Hieronymi, tom. III.). The longest, (HIERON. in Eccle., t. III.), is still a locator, and the (Vide inter …) note is a cross-reference, not a statement. Zero prose notes; recorded as a checked zero, not a skip.

⚑ Two notes print with a following period of Migne's own — etc. [n: (ID.)]. at 1117C and etc. [n: (Matth. XIV)]. at 1121C. Reproduced. They are not our doubled punctuation.

⚑ Migne's citation (Matth. XIV) at 1121C is wrong — Melius illi erat si natus non fuisset homo ille is Matt 26:24 (or Mark 14:21, which may be what the numeral garbles). The note rides verbatim under the marker rule; the [var:] beside it names Matt 26:24 as the comparand, so a reader is not left with the false locator alone.

5.8 · Plate apparatus: NOT READ, and that is stated rather than assumed

I did not read the plate for cols 1115–1126. PL 113's "no foot-of-page conjecture apparatus" finding now rests on fourteen pages across Ruth, Judith and Esther (8968 §1.5, 8960 §1), and those three books make it likely that this one is the same. ⛔ It is a likelihood, not a finding, and importing it would be the "a rendering found in another BOOK is a hint, never an authority" failure that 8960 §2.2 named. data/plate-notes/coverage.json should not record this work as checked on my say-so. Two pages would close it, and §5.1's Salomomonis wants a reader on the same visit.

§6 · THE FOUR TESTS, RUN ON MY OWN FILES

  1. [sic:] does not swallow its word. One marker; markers stripped mechanically from all four chunks and the residue scanned for a stranded article or preposition sitting on punctuation and for doubled marks. Clean. The only two hits were Migne's own etc. [n: …]. periods (§5.7) and a legitimate "and, as though they had not been" — both verified against the Latin.
  2. Negation attachment, host by host. Every printed negative was traced to the single Latin word it attaches to, and the host was proved from the sentence's structure, not from what reads well. The 2a trigger population (a negative resting on a verb with a quantifier, correlative or pivot nearby) was isolated and run separately. The three that could have gone wrong: - 1120C Omnes artes absque doctore non discimus, sola haec … The sola haec pivot fixes the scope as universal, so the English reads "We do not learn any of the arts without a teacher" — not "we do not learn all arts", which is the reading that would let one art through and is exactly what the sola clause is about to concede. - 1121C Melius est itaque omnino non esse quam male vivere. omnino stands immediately before non esse and the comparison is between non-being and evil life, so the English is "better not to be at all than to live ill" — not "it is altogether better not to be", which would put the adverb on the comparison. This is the 8990 @0614B shape. - 1122D et non ex praescientia Dei tollitur liberum arbitrium, sed causae ante praecedunt. The sed clause names the causes, so the negative governs the agent phrase: "free will is not taken away by the foreknowledge of God, but causes go before." This is the ratified 8960 @0740B model, not the 8960 @0734B failure. Also checked and clean: five non … sed correlatives (1117B, 1118A, 1118B, 1119C, 1125A ×2), two nec … sed (1125A), nulla argutione / nulla calumnia (1123B), nullam memoriam (1122C), nisi solus Christus (1124A), nec … quidem dignum (1120A). ⚑ One rendering was checked twice on purpose: 1125B non vult confiteri → "is unwilling to confess." That is the 8987 @0624B shape in reverse, and it is safe here because non is printed — the negative is not being pressed into double duty from elsewhere.
  3. Punctuation by column band, never by total. The eleven/eleven/eleven/nine anchors partition both twins identically; colons, semicolons, question marks, exclamation marks and guillemets were compared band by band. First pass returned thirteen mismatched bands, all ours, all supplied marks — twelve were the habit of rendering Migne's Quasi, as "As if:" with a colon he does not print, plus four semicolons for his commas. All thirteen repaired; the check now returns ZERO mismatched bands in all four chunks. ⚠ Where Migne DOES print the colon (Ac si dicat:, quasi diceret:, Quasi dicat:, dicatur:, qui ait:) the English keeps it. No comma→semicolon sweep was attempted: Migne prints nineteen semicolons of his own in this work, so Tobit's narrow exception does not apply here.
  4. The assertion read. Each finished paragraph read back against the Latin asking only does it assert what he asserts. It produced §3 item 5 (Est et alia infirmitas — a smooth Douay phrase that had collapsed two distinct lemmata), §3 item 6 (Erunt, where the remembered verse is negative and the plate is not), and §4.3 (a clause that read smoothly in the passive and is printed active). All three read well before they were checked, which is the point of the test.

§7 · WHAT A BLIND READER SHOULD OPEN FIRST

  1. 1117C alia frigus consumeret (§4.2) — a formally reversible clause where I chose, and said so.
  2. 1119D rapiebat (§4.3) — a supplied object and a subject taken from two clauses back; the rival (rapiebar) is named and was refused.
  3. 1124D Victorum vel etiam imitatores belli (§5.3) — will not construe; rendered as printed with no marker, and the reason is given.
  4. 1116D Salomomonis (§5.1) — the one place a plate read changes what ships.
  5. 1118D Erroresque et stulti (§3.1) — where [cj:] was drafted and rejected.
  6. 1122A Ne temere, Deus (§3.2) — a lemma joining two separated words, rendered as a stump.

§Adjudication — blind polarity read (2026-08-19)

Blind reader, fenced from this file, all four chunks read in full: 9 sites, 3 repaired. ⭐⭐ The two zeros are the headline and they are measured, not assumed: NO polarity defect in either direction, and NO silent scripture-repair in the running gloss — in the book that was flagged before launch as the batch's highest conformation risk. The reader checked every printed negative individually (twenty named in its report) and confirmed the two sharpest plate-vs-Vulgate divergences were resolved for Migne: ut epulentur bibentes (1125C) rendered "that they who drink may feast" with the Vulgate's viventes in [var:], and Impii non dimidiabunt dies suos (1123C) rendered "The wicked" with Viri sanguinum et dolosi in [var:]. Ruth's four conformations were on exactly this ground; this book has none.

  1. REPAIRED · 1124B · Nunc quae bona eis dederit … enumerat — now "he enumerates what good things [God] gave them." dederit is subjunctive with an unexpressed subject, and that subject is God, established by cum a Deo homines sint boni facti in the previous clause, while enumerat's subject is Solomon. Giving both verbs one bare "he" made Solomon the giver of wisdom and reason. Nothing was missing and the sentence parsed — the invisible class, and the dominant one corpus-wide.
  2. REPAIRED · 1123B and 1123C · two [var:] added — Qui perseveraverit salvus erit against Matt. 24:13, and Jacob dilexi, Esau odio habui against Mal. 1:2–3 (word order reversed, autem absent). Both were unmarked while the identical Matt. 24:13 divergence carries a [var:] three columns later at 1126A. ⚑ The brief's own warning, confirmed a second time: using the apparatus correctly two columns away is not evidence it was used where it was needed.
  3. DECLINED · the six-site lemma-completion cluster (1118C Non est priorum. · 1118B Nec valet, · 1121D Multas. · 1122A Multoque. · 1122A Ubi multa sunt. · 1123A Melius est a sapiente.). Convention 3 requires a fragmentary lemma to be rendered as the phrase it stands for, and the settled boundary is that expansion is never the defect — conformation is. See the batch ruling in 8998's cruces §Adjudication. ⚑ The reader's control set does not control: it contrasted these with Sepultura. → "Burial.", Virum. → "A man.", Bellum. → "War.". Those are complete single-word lemmas with nothing elided; Multas. is an adjective whose noun is missing. Two different classes, so the "inconsistency" is an artefact of the comparison.

⚑ Reader's flagged near-miss, recorded and NOT acted on: 1120C Saulis filiae displicens — displicens is common-gender, so the subject could be saltatio ("that dance … pleased God the more") rather than David. The English chose David, which the preceding nominative and the 2 Sam. 6 sense support. Genuinely ambiguous; named here as the one place a second reader should look.

⬜ Unchanged and still the batch's highest-value plate request: col. 1116D, Salomomonis — 311 Salomonis in src/latin/ against this single form. Indistinguishable from a Corpus Corporum doubling without the plate, and if the plate reads Salomonis the [sic:] comes out and a data/tei-patches/8946.json entry goes in. PL 113's "no foot-of-page apparatus" finding rests on a 19-page sample across four other books and was correctly NOT imported onto these columns.

§Plate read, 2026-08-19 — the [sic:] is WITHDRAWN

Col. 1116D, Salomomonis — settled at the plate, and it was OURS, not Migne's. p. 563 of patrologiaecurs04migngoog (cols 1115/1116) read at 300 dpi and again on a 3× crop: Migne prints the ordinary Salomonis, hyphenated across the last two lines of col. 1116 — (ALBINUS.) Tradunt Hebraei hunc librum Salomo- / nis esse agentis poenitentiam. Corpus Corporum rejoined the broken word and duplicated the syllable in front of the hyphen. ⚑ Corroborated from the source side as well: the TEI keeps a newline immediately AFTER the word, exactly where Migne's line break falls — the transcription preserved the break and doubled the mo.

Actions taken: data/tei-patches/8946.json restores Salomonis; the work was re-chunked; the [sic: Salomomonis] marker has been removed from the English, which now reads simply "this book is Solomon's". The page has been rebuilt and re-indexed.

⭐ This is a new sub-class of digitization corruption for the register: the LINE-BREAK REJOIN. It is not the normalization class (8948 convivum, 8990 concilio), where the digitization silently corrected a form Migne printed. Here the digitization manufactured a word that exists nowhere. ⚑ The detector that found it was corpus frequency — 311 Salomonis against a single doubled form — and the translator was right to raise it as a candidate and not a finding, and right that only the plate could decide. Had it shipped unexamined, a live page would have asserted a defect in Migne's plate that Migne never committed. ⚑ And the general rule it confirms: where our text has a form that is a corpus-wide hapax, suspect OUR files first. Same shape as 8970's Nephtlali and 11535's angüstiae.

Plate apparatus on that page: checked zero — no numbered Forte notes at the foot or mid-page at the Proverbia→Ecclesiastes division. Recorded coverage: "spot".