Homiliae in Ecclesiasten

Merged 2026-07-30 from six agent batches (0000–0008 · 0009–0018 · 0019–0027 · 0028–0036 · 0037–0044 · 0045–0052). 53 chunks, 62,318 Latin words — the largest single work in the corpus. verify-english OK on all 53.

Work-wide conventions (reconciled at merge)

Pre-set before launch (the six agents could not see each other, so these were fixed in every prompt, and all six honoured them): vanitas → "vanity" · Ecclesiastes kept as the name, never "the Preacher" · occupatio pessima → "the worst occupation" · distendi → "distended" · sententia → "sentiment" · narratio → "narrative" · littera → "letter" · the sense-vocabulary (literal/moral/mystical/tropology/allegory) · contemplatio → "contemplation".

Reconciled AT MERGE — two genuine drifts, both fixed:

1. Stultus in tenebris ambulat → "The fool walketh in darkness", unified across all ten occurrences. Five chunks had "walketh" (0000's argument list, 0023–0026's Homily X head) and six had "walks" (0026 body, 0030–0034, the Homily XII head and its exposition). The Latin is the same three words at every site. "walketh" chosen: it is the traditional English of Eccl. 2:14, which the register rules permit echoing where the printed Latin matches, and it was reached independently by two of the three agents who met the lemma. This one mattered more than a verb ending normally would: the argument list at 0000 forward-references Homily XII, and Homily XII is built on setting this reading against Oculi stultorum in finibus terrae. The table of contents was promising the reader one wording and the homily delivering another.
2. arguere → "reprove"; convincere → "convict" — two different Latin verbs, two English words, kept distinct. Four sites had arguere rendered "convict" (0019 arguat, 0021 arguat, 0035 arguens, 0039 arguit/arguendam) and were changed. Checked per instance against the Latin, never swept: chunks 0006, 0045 and 0051 render genuine convincere as "convict" and were left alone, as was the convinceris/convinci rhetorical chain at 0021 ("having denied, thou art convicted; convicted, thou art condemned").

Examined and deliberately NOT changed — these look like drift and are not:

• « » guillemets appear on this lemma in some chunks and not others, and that is FAITHFUL. Migne prints « Stultus in tenebris ambulat. » in the argument list (0000) and the Homily X head (0023–0026), but sets it bare in the Homily XII head (0031): ubi nostra habet: Stultus in tenebris ambulat. Pattern 6 — « » is Migne's, " " is ours. Normalizing would invent punctuation he did not set.
• universa vanitas → "universal vanity" vs omnia vanitas → "all is vanity." Not drift: chunk 0003 alone prints omnia vanitas ×8 and universa vanitas ×2, and the two Englishes track them exactly. Hugh's gloss Universa quippe vanitas homo est collapses if the second is read as "all things are vanity."
• Migne's own repeated lemmata printed differently are preserved, never harmonized: Eccl. 1:13 sapienter everywhere but prudenter at the Homily V head (0154B) · Eccl. 1:8 non saturatur but non satiatur at 0144A ("sated" vs "satisfied") · Eccl. 3:6 congregandi at 0212C but acquirendi at 0214B/0222D/0231C ("gather" vs "acquire") · the tail of Eccl. 3:19 printed three ways — aeque / aequa / similis utriusque conditio — all three rendered distinctly, and the third is the one Hugh's gloss leans on.
• Sense-vocabulary, thou/you, Wisdom feminine — the 0019–0027 and 0028–0036 agents reached the same Sapientia/sapientia capitalization and feminine-pronoun rule independently; thou is reserved for scripture and apostrophe (the sustained Adam-apostrophe of Homilies VIII–IX), you for reader address.
• sapere → "savor" (forced by the sapere/sapor/sapientia pun at 0195B–C) · caduca → "falling" · reprobare → "reject" · consumo/consummo and fornax/caminus kept lexically distinct · disputatio/concertatio, calumnia/oppressio, laetari/gaudere distinct.

⚠ The version hazard — what this work is, and what it is not

Hugh repeatedly quotes a verse in one Latin version, then a second, and builds the exposition on the difference (aliam translationem ×11, nostra habet ×11, aliam interpretationem ×5, Hebraeus ×12, LXX ×3). Homily XII is the head-on case. Every such passage keeps the two wordings verbally distinct.

Two traps recorded so a later sweep does not misread them:

• per aliam interpretationem in the Homily XV head is NOT a competing version. It is Hugh re-expounding the same Eccl. 3 lemmata under a different sense (XIII–XIV literal, XV moral, XVI spiritual).
• Hebraeus at 0174A–B is an ETYMOLOGY, not a version-siglum. A merge or sweep counting occurrences must not treat it as a collation point.

Per-batch logs

The six batch logs follow verbatim, in chunk order.

Batch: Preface + Homily I (cols 0113C–0133B) + Homily II, part 1 (cols 0133C–0138D). Prepared per translation-style.md (Patterns 7, 7a, 7a′, 8, 9, 10, 12, 15).

Work-wide conventions adopted in this batch (for merge reconciliation)

Beyond the pre-set vocabulary (vanitas → "vanity"; Ecclesiastes never "the Preacher"; occupatio pessima → "the worst occupation"; distendi → "to be distended"; sententia → "sentiment"; narratio → "narrative"; littera → "letter"; etc.):

Version-divergence watch (the work's defining hazard)

No aliam translationem / nostra habet / Hebraeus / LXX passage falls inside chunks 0000–0008. Two forward references appear in the argument list at 0000, and both are rendered so that the two readings differ in English exactly as they differ in Latin — later translators of Homilies X and XII must match these:

• « Stultus in tenebris ambulat. » → « The fool walketh in darkness. » (Hom. X and Hom. XII arguments — the nostra habet reading, i.e. the Vulgate.)
• « Oculi stultorum in finibus terrae, » → « The eyes of fools are in the ends of the earth, » (Hom. XII argument — the alia translatio.)
• ubi nostra habet: → "where our copy has:" — literal, as instructed; this is the phrase that flags a Vulgate reading set against another version.

Guillemets: 3 in the Latin of 0000, 3 in the English, 1:1 (Pattern 6). No other chunk in this batch carries guillemets.

Carried type — [sic: …] (Patterns 7, 10, 12)

Six runs. Each is carried untranslated, in italics (except where inside an already-italic scripture lemma, per Pattern 10), and each is verified as a verbatim substring of the Latin twin.

1. 0121C (chunk 0002) — copiosa disputat one extendit. Split type for disputatione (Pattern 10, split-apart class). Carried whole, spacing included: he extends by copious [sic: disputat one]. Conjecture: disputatione, certain from copiosa (abl. sg. fem.) and the sense.
2. 0125A (chunk 0003) — omnium mutabilium rerum … inconstantinam et fluctuationem intuens. Non-word for inconstantiam. Carried: beholding the [sic: inconstantinam] and the fluctuation. Conjecture: inconstantiam, near-certain (coordinate with fluctuationem, acc. sg.).
3. 0128A (chunk 0005) — Ps 103:23 printed Exibit homo od opus suum, et ad operationem suam. od is not a Latin word; conjecture ad, certain from the parallel et ad operationem. Because the run sits inside an italic scripture lemma it takes no second italic marker (Pattern 10): [sic: od].
4. 0128C (chunk 0005) — aliquid in nuere voluit. Split type for innuere. Carried: willed to [sic: in nuere] something to us.
5. 0135C (chunk 0007) — secundum convexitatem circum actionis suae. Split type for circumactionis. Carried: according to the convexity of its [sic: circum actionis]. Conjecture circumactionis, "of its wheeling round" — certain from the astronomical context.
6. 0136C (chunk 0008) — spiritum nunc non incovenienter accipere possumus. Non-word for inconvenienter (dropped n). Not an attested medieval orthography, so Pattern 9's exemption does not apply; carried: we can now not [sic: incovenienter] take the spirit.

Negation and polarity (Patterns 7a, 7a′) — every printed negative carried

⚠ 0132C (chunk 0006) — the one that matters. Migne prints:

extra terminos naturae suae transire non habet de eo quod est in id, quod non est: et idcirco, terra in aeternum stat in eo quod non est.

Hugh's whole argument in this paragraph is that the earth does stand for ever in what it is — the immediately preceding clauses say Perseverat in eo quod facta est; servat naturam suam … nunquam aliud est quam quod est, and the sentence's own first half denies that earth can pass in id, quod non est. The printed non in the final clause therefore makes the conclusion assert the opposite of its own premises.

Rendered as printed: "and for that reason, the earth standeth for ever in that which it is not." Conjecture: in eo quod est (drop the non), near-certain. The negative was NOT silently dropped, and no [sic:] was used to quarantine it — the run is ordinary Latin words, and hiding the contradiction inside a marker while translating the clause affirmatively is precisely the 7a′ failure. The plate's self-contradiction is the evidence.

Other negation-bearing constructions checked and carried intact (no repairs, no smoothing): 0113C non tam … quam, Neque hoc tamen nego; 0117D nec … nec; 0119A non solum sed etiam; 0120C nec vera mala, nec vera bona; 0123A non est vanitas, sed veritas; 0123D Non omnis homo vita, sed omnis homo vivens; 0124A non per vanitatem sed per veritatem; 0129D Non ergo tam laborem quam laborantem; 0130A non laborat sub sole, sed supra solem; 0133A nihil dixisse, quod dicendum non erat; 0138C nunquam indigentia accipiendi posse carere; 0138D mare non redundat.

Real words that are wrong — rendered literally, no marker (Pattern 12 boundary)

These can be rendered, so the English word on the page is ours and takes no [sic:]. Each is a conjecture recorded, not an emendation made.

Prose notes converted to [nt: …] (Pattern 15)

Two notes in this batch are editorial prose (a supplied gloss on the elliptical question of Eccl. 1:3), not citations, and are therefore translated:

• 0126C (chunk 0004) — [n: (Subauditur praeter vanitatem.)] → [nt: It is understood: apart from vanity.]
• 0129C (chunk 0005) — [n: (subauditur praeter vanitatem.)] → [nt: it is understood: apart from vanity.]

The plate's own capitalization difference between the two occurrences is reproduced. All other [n: …] in the batch are locators (Scripture references plus Aeneid. VI, 688) and stay verbatim Latin.

Other notes

• 0136D–0137A (chunk 0008), the Virgil quotation. Migne sets Aeneid VI. 724–727 as four italic verse lines with the [n: (Aeneid. VI, 688.)] locator following the block. The four-line structure and the locator's position are reproduced 1:1; the note content (including Migne's line number, which does not match the modern numbering of the passage) is untranslated and unaltered, per the verbatim rule.
• 0136A (chunk 0008) — the column anchor is printed hard against the opening word ([0136A]Lustrans) with no intervening space, and the lemma is not italicized here though it is italicized when first quoted at 0134B. Both features reproduced as printed.
• Column-band sequence. 0113C → 0114C → 0115A skips the intervening bands; 0122B closes chunk 0002 and 0122C opens chunk 0003. Per the standing ruling these are positional quarter-guides, not a four-per-column grid: no defect, no patch, no plate check.
• (cont.) heads on chunks 0002–0006 and 0008 reproduce our own chunk division and are left untranslated in the marker itself; the head text is translated. # Cruces — Hugh of St Victor, Homiliae in Ecclesiasten (PL 175), chunks 0009–0018

Batch: Homily II, parts 2–5 (cols 0139A–0149D) + Homilies III–VII (cols 0150A–0162D). Prepared per translation-style.md (Patterns 7, 7a, 7a′, 8, 9, 10, 11, 12).

Work-wide conventions adopted in this batch (for merge reconciliation)

Pre-set vocabulary followed throughout (vanitas → "vanity"; Ecclesiastes never "the Preacher"; occupatio pessima → "the worst occupation"; distendi/distendantur → "to be distended"; sententia → "sentiment"; contemplatio → "contemplation"). Conventions set in cruces-0000.md (chunks 0000–0008) were checked and adopted where they recur in this range: sub sole → "under the sun", sub tempore → "under time"; curiositas → "curiosity"; caduca → "falling"; arguere → "to reprove"; homily heads keep their printed abbreviation (HOMILIA II → "HOMILY II", HOMILIA VII → "HOMILY VII").

New decisions made in this batch:

Version-divergence watch (the work's defining hazard)

No aliam translationem, aliam interpretationem, nostra habet, Hebraeus or LXX passage falls inside chunks 0009–0018. Homily XII (the Oculi stultorum in finibus terrae / nostra habet: Stultus in tenebris ambulat homily) lies beyond this range.

Two internal divergences in Migne's own repeated lemmata were nevertheless found and are preserved rather than harmonized:

• 0150A vs 0154B — Eccl. 1:13, sapienter vs prudenter. The Homily III lemma (0150A) and every later quotation print quaerere et investigare sapienter; the Homily V lemma at 0154B prints quaerere, et investigare prudenter, then reverts to sapienter three lines later. Rendered "wisely" and "prudently" respectively. Hugh's whole dialogue at 0154C–0155A turns on the adverb ("Tolle, ait, sapienter"), so flattening the one prudenter would hide a real variation in the printed lemma.
• 0155D — Eccl. 1:13, the bracketed doublet. Migne prints ut [distendantur] occuparentur in ea. The bracketed word is the reading Hugh expounds for the next two whole homilies (VI is titled with it), the unbracketed one is the Vulgate's. Rendered "that [they may be distended] they might be occupied therein" — the brackets kept in place, both readings Englished, because the bracket is not an apparatus siglum here but a second reading of the lemma that the exposition depends on. This is the one place in the batch where matter inside Migne's own square brackets is translated rather than passed through; flagged for the merge.
• 0159C — Isa. 5:20 is printed with the members inverted relative to the Vulgate (ponentes tenebras lucem, et lucem tenebras, i.e. "putting darkness light, and light darkness", against Vulg. lucem tenebras et tenebras lucem). Rendered as printed, in the printed order; the scattered italic runs of the plate are reproduced 1:1.
• 0143C — Jer. 2:13 prints Duo peccata fecit populus meus where the Vulgate has duo enim mala fecit populus meus. Rendered "two sins", as printed.
• 0147A — John 1:3–4 prints omnia in ipso vita erant, et vita erat lux hominum (Vulg. quod factum est in ipso vita erat). Rendered as printed, plural verb included.
• 0158A cites Quid enim prodest homini si universum mundum lucretur… under the note (Matth. VI.); the verse is Matt. 16:26. The printed reference stands (refDisplay rule); no correction proposed.

Text-critical cruces

0139B — Sap. V, Ibunt directe emissiones. Migne prints the clause without fulgurum, so emissiones is left without its genitive. Rendered literally, "They shall go, he says, straight, the things sent forth", rather than supplying "of lightnings".

0142B — dittography, [d: …]. The plate sets sic oculus instrumentum est ad percipiendum verbum hominis, sic oculus instrumentum est ad percipiendum verbum Creatoris. The run sic oculus … verbum is printed twice, the first time closing with hominis (repeated from the preceding auris … verbum hominis), the second correctly with Creatoris — a compositor's eye-skip, then a reset. Both occurrences are rendered per Pattern 11 and the second is wrapped [d: …] (the marker takes the repeated run only; of the Creator falls outside it). No word supplied, none dropped.

0143B — carried type, [sic: cultatem]. Across the column break the plate prints rerum difficultatem | cultatem sermone explicare non potest: the tail of difficultatem is set a second time as a catch-fragment. cultatem is not a word, so it is carried untranslated in italics inside a [sic: …] marker at its printed position, immediately after the column anchor, and difficultatem itself is translated normally.

0144A — vel vivendo vel audiendo. The argument requires videndo (the whole paragraph opposes eye to ear: non satiatur oculus visu, nec auris impletur auditu). Migne prints vivendo, a real word, so it is rendered as ordinary English — "whether by living or by hearing" — per Pattern 12's boundary. Conjecture: videndo, near-certain. No marker; the crux carries the conjecture.

0144C — quam producet supervenerint in futuris. Unconstruable as printed: supervenerint (future perfect / perfect subjunctive, 3 pl.) sits inside a relative clause whose verb is already producet. Almost certainly displaced from the parallel sentence eleven lines above (ex futurorum comparatione cum supervenerint, 0147D). Rendered as closely as the printed words allow — "…the same likeness which it will bring forth they shall have come upon in things to come" — with no word supplied, deleted, or re-cased. The English is as broken as the Latin, deliberately.

0145A — fuit quando nullum tempus fuit. Impersonal fuit quando; "there was [a time] when there was no time", with the supplied noun in square brackets per the corpus's supplied-word convention. The bracketed word is ours, not Migne's.

0149C — Et non est jam vel quod factum est. The vel answers nothing: the parallel sentence just above has quod fuit … et quod factum est, so a first member (vel quod fuit) appears to have dropped. Rendered with the vel carried — "And there is not now even that which hath been done" — and no member supplied.

0150A — ecce ille qui dedit Deus sapientiam et intelligentiam. qui cannot be nominative here; the sense demands cui ("he to whom God gave"). Per Pattern 7a mechanism 5 (reference mis-bound) the printed pronoun is not silently re-bound: rendered "behold, he who God gave wisdom and understanding to lay it open", which reproduces the printed case. Conjecture: cui.

0150A — Ergo non meditabimur … et requiescere poterit cor nostrum donec inveniat illam. The printed sentence says the opposite of Hugh's argument twice over: the context is an exhortation to meditate and to be restless until wisdom is found. Both printed readings are carried — the non of non meditabimur, and the affirmative requiescere poterit — so the English reads "Therefore we shall not meditate on his commandments, nor shall we search out her paths, and our heart will be able to rest until it find her." Conjecture: the clause was nec requiescere poterit, and the opening was a question. No negation supplied and none removed (Pattern 7a). This is the sharpest polarity crux in the batch; a later reader should not "fix" it without the plate.

0150B — si inventam non possumus, tamen quaesitam doceamus. The verb of the first member is not printed; supplied in brackets as ours — "if we cannot [teach] her as found".

0152A — Et quod vult aut non esse quod est, aut esse quod non est: hoc utroque pejus est. The preceding sentence has just said the mind nec … velle possit either of these two things; the printed vult therefore asserts what the argument has just denied. Rendered as printed ("And that it wills either that what is should not be, or that what is not should be"), no negation supplied. Conjecture: non vult.

0153A — novissimae ad appetitum voluptatis inflammatur. novissimae for the adverb novissime. A Latin case-ending has no English exponent (Pattern 9), so the adverb is rendered "lastly" and the printed form is recorded here rather than carried into the text.

0154B — ibi deficiendo veraciter directionem mei agnovi. directionem ("the straightening / direction") stands where the sense wants dejectionem ("my own casting down") — the clause is the antithesis of supra me tumui. Real word; rendered literally, "there by failing I truly recognized the direction of myself." Conjecture: dejectionem.

0155D — omnes non attentos et exspectantes jam fecerat. The grand promise is said to have made everyone not attentive, in a sentence whose next line asks who could have expected so little from so great a promise. The printed non is carried: "had already made all men not attentive and expectant." Conjecture: the non is intrusive. Pattern 7a, mechanism 2 — recorded rather than repaired.

0156C — quasi per rivulos quosdam visibilium … saltem fugere conatur. The image is of drawing water through channels, so sugere ("to suck") is expected; fugere is printed. This is the exact class Pattern 12 names as not taking a marker (a real word merely wrong), so it is rendered as ordinary English — "it attempts at least to flee, with parched inward parts" — and the conjecture lives here.

0157C — quoniam si poenam habent a beatitudine alienam, culpam non habet…. Number disagreement between habent and habet within one sentence, both governing the same subject (haec occupatio). No English exponent for the Latin ending on the second verb; rendered as printed ("if they have a punishment … it has not a fault"), left alone per Pattern 9.

0158B — Preposui, ait, in animo meo…. Every other occurrence of this lemma in the work prints Proposui. Preposui (= praeposui, "I set before") is a real word, so it is rendered literally — "I set before, he says, in my mind…" — rather than read as the expected Proposui. Conjecture: Proposui, near-certain.

0159B — in omnibus quae fiunt sole. The preposition sub is absent, in a phrase the same sentence prints correctly twice. sole alone cannot be rendered in the slot, so the bare printed word is carried untranslated in italics inside [sic: sole]; the rest of the clause is translated. Conjecture: sub sole.

0160A — Nunc ergo perversi et stultitam sunt ad correctionem perversitate difficiles, quam multitudine contra veritatem audaces. stultitam is a non-word run: it conflates stulti with the correlative tam that quam later answers. Carried whole and untranslated inside [sic: stultitam] per Patterns 10 and 12. Per Pattern 7a′, the correlative force quarantined inside the marker is still carried by the translated text: the English reads "are as hard to correction by their perversity as they are bold against truth by their multitude." The stulti inside the marked run is deliberately not translated. Conjecture: stulti tam.

0160A (same sentence, punctuation) — the plate prints …infinita est stultorum multitudo Nunc ergo… with no stop. Per Pattern 8 no stop is supplied; the English runs on identically.

Non-defects checked and deliberately left alone

• 0157D, 0159C — column anchors printed inline mid-sentence and inside an italic citation run ((Job [0159C] XXX)). Reproduced in place; not moved to a paragraph break.
• Inline citations in italic parentheses ((Job XXX) at 0159C) are Migne's running references, not Pattern 4 [f: …] tails: this is homiletic exposition, not a florilegium. Left as printed, untranslated, untagged.
• 0149D — Vanitas vanitatum, vanitas vanitatum, et omnia vanitas. The doubling is Eccl. 1:2's own, not a compositor's; no [d: …].
• 0143B — persemetipsum set as one word; attested medieval agglutination (Pattern 9), translated normally.
• 0152A — adversionem mentis for aversionem; attested spelling, translated normally, no marker.
• 0160C — involumentum; a formed word, translated "entanglement", no marker.
• Column-band sequences in this range (0139A–0139D, 0150A–0150D, a D-band followed by an A-band at 0151C→0151D, 0157C→0157D) are positional quarter-guides and are transcribed where they fall. No text breaks mid-word or mid-construction across any anchor in the batch. # Cruces — 11065 (Hugh of St Victor, Homiliae in Ecclesiasten), chunks 0019–0027

Columns 0163A–0186A. Homilies VII (cont.), VIII, IX, X, XI.

Work-wide conventions adopted in this batch

Recorded so the merge can reconcile with the other five agents' batches. Only decisions not pre-set in the batch prompt are listed.

1. transire ad contemplandam sapientiam → "to pass over to contemplate wisdom", uniformly, including the Eccl. 2:12 lemma Transivi ad contemplandam sapientiam → "I passed over to contemplate wisdom." The whole of Homily X turns on the etymology Hugh gives at 0174A (Hebraeus transiens interpretatur), so transire must read as the same verb in the lemma, in the etymology, and in the twenty-odd prose repetitions. "Cross over," "go over," and "migrate" are used nowhere for transire; migrat @0179C is rendered "removes," which keeps it distinct.
2. Sapientia / sapientia → "Wisdom" / "wisdom", following Migne's capitalization 1:1, and feminine ("she / her") throughout — agreed with the 0028–0036 batch, reached independently. Hugh's Moses/Shunamite figure at 0175B–0176C depends on wisdom being a woman set against the Ethiopian wife.
3. thou vs you. thou is used for (a) scripture as quoted, and (b) the sustained apostrophe to a named figure — Adam and the cor humanum through almost the whole of Homily VIII (0166A–0169D) and Homily IX (0170A–0172B), where Hugh is re-staging Gen. 3:9 and the second person is dramatic, not didactic. Generic second-person address to the reader is you (Si habueris oculos @0182A; Si ipse praecedit, et ut sequeris @0180B; Nam secundum aliquid omnia hoc tibi esse incipiunt @0173D). This extends the 0028–0036 batch's rule rather than contradicting it: that batch reserved thou for apostrophe to God or Wisdom and did not meet an Adam-apostrophe.
4. species → "appearance" (never "species," "form," or "beauty"), because Homily IX @0171C–D turns species rerum visibilium against veritas, and Homily X @0173A repeats vidit speciem, non praevidit corruptionem. pulchritudo → "beauty," decor → "comeliness," kept lexically distinct in 0185A–D where Migne pairs them.
5. folia → "leaves", umbra → "shadow" (or "shade" where the sense is the cast shelter rather than the dark image). Homily IX's whole figure is the fig-leaves of Gen. 3:7 read as the species rerum visibilium.
6. elatio → "elation" (not "pride"), kept distinct from superbia → "pride" and tumor → "swelling"; all three occur in one sentence at 0163B.
7. occupatio → "occupation" (pre-set for occupatio pessima) is used for the bare noun too — curiosus occupationem invenit @0164A, ad occupationem vanam converteris @0169D — so the Eccl. 1:13 lemma stays audible.
8. finis → "end" throughout Homily XI, never "limit" or "boundary," since 0183B–0184B is an extended play on finis as both terminus and caput. terminus @0183A, where it stands beside finis, is "limit."
9. accidiatus @0164C → "seized with accidie." The medieval technical term is kept rather than flattened to "listless"; acedia is a named vice and Hugh is naming it.
10. Migne's own bracketed matter is reproduced verbatim, untranslated — [urceos] @0171A, on the same footing as [ al., …].
11. Printed parentheses are reproduced — (esse coeperat) @0167B.

Version divergence — Hugh's competing Latin Ecclesiastes texts

This batch contains no aliam translationem / nostra habet / Hebraeus + LXX version-collation. The one place a second witness is invoked is @0182C–D:

• 0182C–D, Aliam itaque Scripturam interrogemus. Hugh has just quoted Eccl. 2:14 as Sapientis oculi in capite ejus; stultus in tenebris ambulat (his lemma, @0182B). He then cites, as another Scripture, Oculi sapientis in capite ejus; oculi autem stultorum in finibus terrae — Prov. 17:24. The two are rendered with their printed differences intact, including the word order (Sapientis oculi in the lemma, Oculi sapientis in the Proverbs citation), because Hugh's whole argument is that the second text supplies what the first leaves unsaid: where the fool's darkness is. Neither was normalized toward the other or toward a conventional English. Migne prints no [n: …] for the Proverbs text.
• The Hebraeus of 0174A–B is not a version-siglum. It is the etymological gloss Hebraeus transiens interpretatur (Jerome's Liber interpretationis), and it is rendered as a proper name plus its interpretation, not as a witness. A merge scan counting Hebraeus occurrences must not treat these as collation points.

Text-critical notes, by column

### 0164C — accidiatus Migne prints attritus et accidiatus. Accidiatus is a well-attested medieval formation from acedia and is not a defect (Pattern 9); no marker. Rendered "seized with accidie."

### 0167A — Quaeris ut tuam ignorantiam arguat; ambulat ut instabilitatem tuam ostendat. Printed Quaeris (2nd sing.) with a 3rd-sing. purpose clause whose subject must be veritas, exactly as in the parallel clause ambulat ut … ostendat. The argument requires Quaerit ("she seeks thee, that she may convict thy ignorance"), and the preceding sentence has just made truth the subject (clamat tibi, et dicit). Rendered as printed — "Thou seekest, that she may convict thy ignorance" — under Pattern 7. Conjecture: Quaerit.

### 0167B — neque abiit indirectum elongans Printed indirectum as one word, where the same page prints in directum non vadens as two, sixteen lines earlier, in the same sense. This is a run-together of Pattern 10 shape, but indirectum is also a real Latin word, so under the Pattern 12 boundary it is rendered as ordinary English rather than carried: "nor did it go off straight on, drawing far away." Conjecture: in directum. No [sic: …]. Flagged so a later sweep does not read this as a considered choice about in-directum ("crookedly"), which would reverse the sense.

### 0168C — missing terminal stop after effundit Migne prints … ad avaritiam se effundit Etenim prius fructum ad esum expetit, with no point and a capital E. Preserved under Pattern 8: the English runs "…pours himself out toward avarice For indeed he first sought after the fruit…" No stop was supplied.

### 0171B — Cave ne in umbra filiorum decipiaris Printed filiorum ("of the sons") where the argument, and the identical phrase five lines later (Hanc ergo umbram foliorum suspectam habe) and again at 0171C, 0171D, 0172A and 0172B, all read foliorum ("of the leaves"). The clause is about Adam hiding under fig-leaves. Rendered as printed — "in the shadow of the sons" — under Pattern 7. Conjecture: foliorum, near-certain (single-letter minim confusion). This is the single most likely candidate in the batch for a silent repair, and it was deliberately not made.

### 0170C — Intra avaritia sub obtentu parcimoniae Printed Intra where the sentence has no other finite verb and avaritia is nominative. Rendered as an adverbial "Within, avarice under the cover of thrift; and while it persuades the mind…", which is as close as the printed words allow. Conjecture: Intrat avaritia, "avarice enters." Logged rather than emended.

### 0172A — stulti, et imprudentes fiunt tibi Printed tibi (2nd sing. dat.) in a sentence whose subjects are the third-person filii hujus saeculi; sibi is what the sense requires ("they become foolish to themselves"). Rendered as printed — "they become foolish and imprudent to thee." Conjecture: sibi.

### 0171D — et vasa pretiosa divitiae et pompae, et gloria No comma after pretiosa in a list that is otherwise fully pointed. Preserved (Pattern 8): "and precious vessels riches and pomps, and glory."

### 0169C — Quocunque perrexerit, fraudaberis 3rd-sing. subjunctive/future-perfect perrexerit governing a 2nd-sing. main verb fraudaberis, in a passage otherwise addressed to Adam in the second person. Rendered as printed: "Wherever he shall have gone, thou wilt be defrauded." Conjecture: perrexeris. No English exponent is lost either way, but the person-shift is real and is not ours.

### 0169C — non ibi requiem invenire poteris, ubi es? quia veritas ibi non est Migne sets a question mark inside the sentence, after ubi es, and continues with a lowercase quia. Preserved (Pattern 8): "…thou wilt not be able to find rest there, where art thou? because the truth is not there where thou art."

### 0175C — Hujus non enim Sunamitis, id est sapientiae, amorem significabat ignis ille The polarity crux of this batch. Migne prints non. The sentence is the hinge of the Moses figure: the burning bush that burns without being consumed has just been set against the Ethiopian wife, and the very next clause is et ideo Moyses, postquam transivit ad videndam visionem … non amplius teneri potuit Aethiopissae uxoris amore — an inference that only follows if the fire did signify the love of the Shunamite, i.e. of wisdom. The non makes Hugh's own et ideo a non sequitur.

The negation is carried (Pattern 7a): "For that fire, by which the bush was burning and was not burned up, did not signify the love of this Shunamite, that is, of wisdom; and therefore Moses…" Conjecture: the non is a plate defect — either intrusive, or displaced from a lost nonne (Hujus enim Sunamitis … amorem nonne significabat ignis ille?, which would give the interrogative Hugh uses constantly in this homily). No repair was made in the text, and no [sic: …], because non is a real word correctly set: only the reading is wrong.

### 0176B — Illa propter se appetenda est, ista propter se fugienda propter nos sustinenda. illa = sapientia, ista = vita carnis, on the medieval remote/nobler vs near/baser use of ille / iste, not on the classical former/latter rule (which would invert the sense). Rendered "That one is to be sought after for her own sake; this one is to be fled for its own sake to be endured for our sake." The missing comma before propter nos is preserved. Not a defect; recorded so a sweep does not reverse the demonstratives.

### 0179B — Quid est sapientia? cognocere et amare bonum. cognocere is not a word. Carried into the English untranslated, in italics, inside [sic: cognocere] per Patterns 7 and 12. Conjecture: cognoscere, certain — the same verb is spelled correctly three times in the surrounding fifteen lines (cognoscere et odisse, cognovisse malum, cognoscere et amare bonum). This is the only carried type in the batch.

### 0179C — cum sapientia videtur et quo ipsa est ut diligatur Printed quo where the parallel clause four words later has quod ipsa non est. Rendered "and what she is, that it may be loved," i.e. as though quod. Conjecture: quod. Flagged because the rendering silently regularizes a one-letter form that has no English exponent; nothing else in the clause was touched.

### 0180D — Quod longe enim putas homo est a Deo Printed Quod for the expected Quam. Rendered "For how far dost thou think man is from God" — the interrogative adverb is the only construal the rest of the sentence permits. Conjecture: Quam longe.

### 0178D — Et sit stultus nihil videt, ergo stultitia excaecat. Printed sit. The preceding sentence is the exact structural twin — Si enim sapiens oculos illuminatos habet, ergo sapientia illuminat — so the protasis marker must be si. As printed the clause will not construe at all: sit is a subjunctive with no place in the sentence, and there is nothing to render. Carried untranslated inside [sic: sit] (Pattern 7's "if you cannot make the printed reading say anything at all"): "And [sic: sit] the fool sees nothing, then folly blinds." Conjecture: si.

### 0182A / 0182B — Si habueris oculos quo illa attingit; A protasis with no apodosis; Migne closes it with a semicolon and starts a new conditional. Preserved as a fragment — "If you have eyes where she reaches; but if you have eyes where the light is not…" — rather than joined to the preceding sentence. Pattern 8.

### 0185D — quando magis priora faciendo Printed quando in the second limb of a tanto … quanto correlative (tanto minus semper posterioribus … contulit, quando magis priora faciendo … consumpsit). Rendered as the correlative requires ("so much the less … as she the more"). Conjecture: quanto. Recorded because the correlative is what carries the argument's comparative force, which Pattern 7a mechanism 4 warns is the easiest thing to reverse.

### 0185D–0186A — the "wearied Wisdom" passage Fortassis fatigata erat priora illa faciendo … and what follows attribute exhaustion and diminishing power to Wisdom herself. This is Hugh stating a position he will demolish (the refutation begins in the next chunk, 0186A ff.), not his own doctrine. Rendered straight, with no softening and no editorial signal, since Migne prints none. Flagged only so that a reader of this chunk in isolation — or a later sweep hunting doctrinal anomalies — does not read it as a defect.

### Column bands [0180A] is followed later in the same chunk by [0181A], and the standalone anchors [0172C] and [0183A] sit on their own line after a ## head. Both are normal transcription behavior (known false positive 1); no text is broken across any anchor in this batch. Anchor sequences were checked 1:1 against the Latin twins for all nine chunks. # Cruces — 11065 (Hugh of St Victor, Homiliae in Ecclesiasten), chunks 0028–0036

Columns 0186B–0209D. Homilies XI (cont.), XII, XIII.

Work-wide conventions adopted in this batch

Recorded here so the merge can reconcile with the other five agents' batches. Only decisions not pre-set in the batch prompt are listed.

1. Sapientia / sapientia → "Wisdom" / "wisdom". Migne's capitalization is followed 1:1: capitalized where the plate capitalizes, lowercase where it does not. Hugh alternates within a single sentence (0188A–B) and the alternation is data, not noise. Wisdom is feminine throughout ("she / her"), following the Latin gender, which Hugh's imagery (right hand, left hand, bride) relies on.
2. thou vs you. thou is reserved for (a) scripture as quoted, and (b) direct apostrophe to God or to Wisdom (O Sapientia, quam late diffunderis @0186C; Quid gloriaris in malitia @0202C and the sentences Hugh builds around it). Second-person address to the reader — far commoner — is you (Si enim ascendis in coelum justitiae @0187B; Vis scire quod Christus caput est @0194D; Quare ergo tu hujus mundi… metuas @0206A).
3. sapere → "savor", not "mind" or "be wise." Forced at 0195B–C, where the Colossians 3 lemma (quae sursum sunt sapite) is immediately glossed Quid est sapere? … Amor enim ipse sapor est — the whole gloss is a pun on sapere / sapor / sapientia. "Mind the things above" (the usual English) destroys it. Applied to every occurrence of sapere in this batch, including qui terrena sapiunt @0195D ("who savor earthly things").
4. sententia → "sentiment" throughout (pre-set), including the many places where Hugh means the printed clause of Ecclesiastes he is about to expound (Et nota quomodo se consequitur sententia veritatis @0196D). Rendered "sentence" nowhere in this batch.
5. distendi family → "distend" (pre-set), extended to the noun: distensio laboris @0200C → "the distension of labor"; distendit se ac replet @0201B → "distends himself and fills himself".
6. consumo vs consummo kept lexically distinct — "consume" vs "consummate". Migne sets them adjacently and antithetically at 0192C–D (morte consumuntur … damnatione consummatur), so levelling them would destroy the argument.
7. fornax / caminus kept distinct ("kiln" / "furnace") in the Eccli. XXVII citation @0202D, because Migne's line pairs them.
8. Inline citation parentheses that are NOT [n: …] markers ((Psal. XXVI) @0202C, (Eccle. VII) @0192D–0193A) are kept verbatim Latin, on the same footing as [n: …] contents. Both happen to have a column anchor set inside them; the anchor is reproduced in place.
9. The [qua] bracket @0199B is Migne's own bracketed variant and is reproduced verbatim, untranslated, like [ al., …]. Same for [pessundatus] @0202B.

The two-version hazard in this batch

Homilies XII and XIII sit squarely on the competing-version problem, and the two Latins are kept verbally distinct in the English wherever Hugh sets them side by side:

• Head of Homily XII (0190C, repeated as (cont.) on 0031/0032/0033/0034): secundum aliam translationem: Oculi stultorum in finibus terrae; ubi nostra habet: Stultus in tenebris ambulat. Rendered "according to another translation: The eyes of fools are in the ends of the earth; where our copy has: The fool walks in darkness." The whole of chunk 0032's opening (0196B–0196C) is an argument that the two are not contradictory — "both seem to be affirmed, that they see and that they do not see" — so the two wordings must never be levelled toward a single conventional English. They are not.
• Eccl. 2:14 quoted a third way @0197A, from the Vulgate as Hugh's running text: Sapientis oculi in capite ejus stultus autem in tenebris ambulat → "The eyes of a wise man are in his head, but the fool walks in darkness." This third form is the nostra habet wording, and is rendered identically to the head's nostra habet half — deliberately, since it is the same version.
• Eccl. 7:5 (Eccle. VII) is quoted twice in this batch, in two different Latins, and the two Englishes differ exactly as the Latins do:
• @0192D–0193A: Cor stultorum ubi laetitia est et cor sapientium ubi tristitia → "The heart of fools is where mirth is, and the heart of the wise where sadness is."
• @0203C: Cor stulterum ubi laetitia, et cor sapientium ubi tristitia est → "The heart of [sic: stulterum] is where mirth is, and the heart of the wise where sadness is." (stulterum is broken type; and note est has moved clauses between the two quotations.) Both invert the Vulgate's clause order (Cor sapientium ubi tristitia est, et cor stultorum ubi laetitia). Rendered as printed in both places, and the inversion is not corrected.
• Eccl. 2:24–26 is quoted twice in chunk 0034 with a spelling divergence Migne himself does not flag: curam surperfluam @0202C (broken) and curam superfluam @0203D (correct), and Ut addat et congreget @0204A against the first quotation's ut laboret, ut distendatur, et congreget. Both forms stand.

Chunk 0028 — cols 0186B–0188C

• @0186A–B quia summa esset pulchritudo in singulis quaelibet excellens illa foret, et caeteris omnibus eminentior, non esset omnis pulchritudo in universis. Unconstruable as printed: the protasis has no si and the apodosis has no connective. Rendered as closely as the printed words allow ("because the highest beauty would be in individual things — whatever one there were would be surpassing that one, and more eminent than all the rest — not every beauty would be in all things together"). Conjecture: quia si summa esset pulchritudo in singulis, quaelibet excellens illa foret …; non esset …. Not emended; the printed non is carried.
• @0186C singula quaeque ad terminum suum promonens. promonens is not a word. Carried untranslated per Patterns 7/12: [sic: promonens]. Conjecture: promovens ("advancing"), near-certain from ad terminum suum.
• @0187A et anima eorum in malis tabescat (Psal. CVI [107]:26). Vulgate has tabescebat (imperfect indicative); Migne prints the present subjunctive, and the italics stop before it, so it is grammatically Hugh's clause rather than the quotation's. Rendered "wastes away." No [var:] — the divergence is of mood and tense only and does not turn the sense.
• @0187B et descendunt usque ab abyssos. The same verse is quoted twice in four lines; the first time Migne prints usque ad abyssos, the second usque ab abyssos — a preposition that cannot govern the accusative that follows it. Rendered literally as printed ("they descend even from the depths"), which contradicts the first quotation four lines above. The contradiction is the plate's and is left standing. Conjecture: ad, certain from the first quotation and from the Vulgate.
• @0188A quia non esset ubi esse posset malitiam. malitiam accusative where the construction wants a nominative subject of esse posset. Rendered "because there would be nowhere that malice could be." Gender/case has no English exponent here (Pattern 9); no marker, no emendation.

Chunk 0029 — cols 0188D–0190B

• @0188C–D Non enim malitia Sapientiam praecedere potest, ut vel ante illius exortum prior inveniatur, vel post illius defectum, posterior quia et prius corruptione invenitur … The comma after posterior and the absent punctuation before quia leave the sentence hanging. Rendered with the printed word order and the printed Non; no punctuation supplied (Pattern 8).
• @0189C Sed et si quando exterius violentiam pravis voluntatibus adhibendis, sive plectendis adhibet. adhibendis and adhibet in one clause: the gerundive duplicates the finite verb, and the sentence does not construe as it stands. Rendered as closely as the words allow ("applies violence to crooked wills to be applied to, or to be chastised"). Possible intended reading: coercendis sive plectendis — but the plate gives no warrant for choosing one conjecture, so none is adopted.
• @0189D Restat nunc maximus ille et supremus Sapientiae excursus, a fine usque ad finem: qui nullum habent finem. qui … habent — masculine singular relative with a plural verb, and the antecedent is the two fines. Rendered "which ends have no end." Log only; Pattern 9 (no English exponent for the Latin agreement).
• @0189D sive promovende, sive deserendo. promovende is not a form. Carried: [sic: promovende]. Conjecture: promovendo, certain from the parallel deserendo.
• @0189D ut per justitiam eatur ab beatitudinem. ab + accusative: does not construe, and rendering the preposition literally ("from blessedness") would assert the opposite of the clause's evident direction while looking like a considered reading — the exact trap Pattern 7a′ describes. The run is therefore carried whole and untranslated, [sic: ab beatitudinem], so the reader sees Migne's type and nothing is silently supplied. Conjecture: ad beatitudinem, certain from the balancing perveniatur ad damnationem in the same sentence. Nothing translated in this sentence depends on the marked run for its polarity — the antithesis justice/iniquity is carried in the surrounding words, which are rendered in full.
• @0190B et ab omnipotenti bona mala esse permissa sunt. Construes as bona … permissa sunt esse mala ab omnipotenti — "good things have been permitted by the Omnipotent to be evils" — and is so rendered. omnipotenti stands alone as a substantive ablative. No emendation.
• @0190B non ut mala dona fiant. dona where the argument wants bona ("not that evils may become goods"). Rendered as printed: "not that evils may become gifts." Conjecture bona, likely but not certain — dona is a real word and the sentence, though odd, is not unconstruable, so Pattern 12 keeps it as ordinary English with no marker.
• @0190B Ergo aliqui faciunt mala propter quod esse permissa sunt: et bonum faciunt ipsa et mala et non ipsa mala faciunt bonum. Two problems in one sentence, both left standing. (i) aliqui (masc. nom. pl., "some men") where the neuter ipsa two clauses later, and the whole preceding argument (si nihil facerent … omnino mala non essent), require aliquid — "evils do something." (ii) The result, rendered literally, is a flat self-contradiction: "both they themselves and evils do good, and evils themselves do not make good." Both the affirmative and the printed non are carried. Conjecture for (i): aliquid, near-certain.

Chunk 0030 — cols 0190C–0193B

• @0190C quia invenimus finem, et finem de quibus Scriptura locuta est. Singular finem twice with plural de quibus. Rendered "we have found an end, and an end, concerning which Scripture spoke" — the doubling is the point (the two ends of Wisd. 8:1), so the repetition is preserved rather than collapsed. Not marked [d:]: this is not a compositor's doubling but Hugh's own figure.
• @0191A Et si dicimus de terra ista, quod in finibus ejus oculi stultorum sunt quomodo hoc stare non poterit. Printed with a full stop, not a question mark, although the clause is interrogative; and non makes the sense the reverse of what the following sentence argues. Both the stop and the non are carried (Patterns 7a, 8): "how this will not be able to stand."
• @0191B qui a terrena bona tota, mentis intentione semper ambiunt. a + accusative; does not construe, and ambiunt is already transitive, so the preposition is intrusive. The run is carried whole, [sic: a terrena bona], rather than rendered with an invented preposition or silently dropped. Conjecture: ad terrena bona, or simply terrena bona with a deleted.
• @0191D et invisibilia supra coelum nostrum, invisibilia sub terra nostra. The second invisibilia should be visibilia: the whole paragraph is built on the antithesis (Invisibilia intus sunt, visibilia foris sunt. Ergo invisibilia supra sunt, visibilia subtus), and the immediately preceding clause states it correctly. Migne's repetition makes the sentence contradict its own premise. Rendered as printed — "and invisible things are above our heaven, invisible things beneath our earth" — because this is precisely the class of defect the edition exists to expose. Conjecture: visibilia sub terra nostra, certain.
• @0193A In omnibus operibus tuis omnis diebus vitae tuae (Eccli. VII, 40). omnis diebus does not construe; carried as broken type inside the italic lemma ([sic: omnis diebus], no nested italic marker, per Pattern 10). Conjecture: omnibus diebus. Note that the Vulgate of Eccli. 7:40 has no "days" clause at all (In omnibus operibus tuis memorare novissima tua), so the phrase is either Hugh's expansion or a conflation; that question is not adjudicated here.
• @0193A et in aeternum non peccabis — the printed non is carried ("and for ever thou shalt not sin").

Chunk 0031 — cols 0193C–0196A

• @0193C Omne enim datum optimum, et omne donum perfectum desursum est descendens a patre luminum (Jac. I, 17). Migne breaks the quotation into two italic runs around enim. English word order cannot put "for" between "every" and "best giving," so the two runs are kept adjacent and For is moved to the head of the sentence. The run boundary is preserved; no words are moved across it.
• @0194C sed in tentationem suam deorsum curvantes. tentationem where the antithesis with intentione (three lines later, mentis intentione ac desiderio) suggests intentionem. Rendered as printed ("bending downward into their own temptation") — a real word, sense intact, no marker (Pattern 12).
• @0194D quae in canticorum dilectum suum commendans. A bare genitive canticorum with no governing noun (in Canticis canticorum expected). Rendered "in Canticles," treating the genitive as the book's title; no word supplied in brackets.
• @0195A–B et caput Christi Deus et unum caput Christus, et caput Christi Deus; sicut unum principium Christus, et caput Christi Deus. et caput Christi Deus is set three times in two lines. Not marked [d:]: Pattern 11 covers an adjacent doubling of one run, and these are separated by other clauses, so the repetition may be Hugh's own incantatory figure rather than a compositor's. All three are rendered.
• @0195B Ego, inquit, [0195B] principium, qui et loquor vobis (Joan. VIII, 25). Note that the column anchor falls between inquit, and the second italic run; it is reproduced at exactly that position. inquit stays inside the quotation per Pattern 5 (there are no guillemets here, but the run boundaries do the same work).
• @0195B quae sursum sunt sapite … quae sursum sunt quaerite (Coloss. III, 1–3). Migne reverses the Vulgate's order of the two imperatives (Vulg.: quaerite first, then sapite). Rendered in Migne's order. No [var:]: no fact changes, and Hugh's gloss immediately following depends on sapite standing where Migne puts it.
• @0195C Et quia quod per concupiscentiam diligitur, deorsum quoque trahat necesse est ipsa concupiscentia. The quia clause has no predicate — the parallel sentence immediately before supplies it (quod per sapientiam diligitur, sursum est; idcirco …). Rendered as printed, with the gap left open. Conjecture: deorsum est; idcirco deorsum …, i.e. haplography of deorsum.
• @0196A ut nec cogitare valcant. valcant is not a form. Carried: [sic: valcant]. Conjecture: valeant, certain. Both the nec and the following neque are carried in the English, and the negative force governs the two translated infinitives ("to think… to look forward"), per Pattern 7a′ — the marker quarantines only the broken verb, not the negation.
• @0196A nisi quia stultitia excaecavit oculos eorum: quam in finibus terrae illos defixerunt. quam (fem. acc. sing., = stultitiam) with illos (masc. acc. pl., = oculos) as the apparent object of defixerunt; the two cannot both be the object. Rendered following the printed cases ("which they have fixed in the ends of the earth"); log only, no emendation.

Chunk 0032 — cols 0196B–0199D

• @0196C et illececebra voluptatum. Not a word. Carried: [sic: illececebra]. Conjecture: illecebra, certain (the correct form appears at 0201B, voluptatum illecebris).
• @0196D–0197A Vides tamen quod corda … tenebrescunt, quasi definitivam boni malique sententiam dedit. Second-person Vides with third-person dedit in the same period. Rendered as printed ("Yet you see that hearts … — he gave, as it were, a definitive sentiment"), with an em-dash standing where the plate has a comma. Conjecture: Videns, which construes perfectly and matches the neighbouring participles (considerans, Attendens).
• @0196D Deinde considerans quod vera sapientia hominis alia non est … qui nos idcirco praecessit. The considerans clause has no main verb; the sentence simply stops and the next begins Attendens quoque. Anacoluthon rendered as printed, no verb supplied.
• @0197B pro interna luce quaerenda labore vanum existimat. labore (ablative) where the object of existimat must be accusative laborem. Rendered as ordinary English with labor as the object ("counts vain the labor for the inner light to be sought") — a real word merely in the wrong case, so no [sic:] (Pattern 12).
• @0197B Nam quia pari sorte similique conditione sapientes patitur et stultos, ab hac luce visibili cernit post mortem carnis subtrahi. patitur (3 sg.) sits inside an accusative-and-infinitive governed by cernit, and its subject is unrecoverable. Rendered by taking patitur … subtrahi as "suffer to be withdrawn," with sapientes et stultos as its subject; the plate's word order is otherwise kept. Flagged as uncertain.
• @0197C cujus fructus an perpetuus esse possit, nondum adhuc ei ratio ulla indubitata manifestat. manifestat transitive with no object. Rendered "no undoubted reason as yet makes anything manifest to it," with "anything" as the minimum English requires; the alternative (leaving the verb objectless) is not English. Flagged.
• @0198C Sicut enim hic eos dispar vivendi conversatio dividit, ita illic quoque cum in memoriam venerint ut ad judicium adducatur. Plural venerint with singular adducatur. Rendered as printed ("when they shall have come into remembrance so that he may be brought to judgment"). No emendation.
• @0198C Propterea sapientis memoria non in perpetuum non erit similiter ut stulti. Two negatives where the argument requires one. Hugh is here restating the lemma Non enim erit memoria sapientis similiter, ut stulti in perpetuum (0198B) affirmatively of his own thesis, and the clause that follows it in the same sentence (et futura tempora non operient cuncta oblivione) carries a single negative. The doubled non therefore makes the restatement assert the opposite of the lemma it restates. Both negatives are carried — "shall not for ever not be in like manner as that of the fool" — per Pattern 7a; the English is self-contradictory because the plate is. Conjecture: one non is a compositor's doubling. Not marked [d:]: the two are not adjacent (non in perpetuum non erit), so Pattern 11 does not reach it, and marking one of them as the repeat would be asserting which one is spurious.
• @0198D quia dignum est ut homo quod transitorium videt etiam prius quam finiatur despiciat? et pro eo adipiscendo … non affligat. A question mark set in the middle of a declarative sentence, before et. Reproduced exactly (Pattern 8); the English keeps the interior ? and continues the sentence in lower case, as Migne does.
• @0199A quae mala videntur hujus vita. Nominative vita for genitive vitae. No English exponent for the case; rendered "of this life," logged, no marker (Pattern 9).
• @0199B Et quae [qua] sub sole studiosissime laboravi. Migne's own bracketed variant [qua] reproduced verbatim and untranslated. Subauditur detestatus sum following is rendered with straight quotes round the supplied English lemma ("I detested"), per Pattern 6 — those marks are ours, and Migne prints none.
• @0199C Si enim vana essent haec omnia … quid erunt ad stultorum voluptatem transmissa? Imperfect subjunctive protasis with future indicative apodosis. Rendered as printed.

Chunk 0033 — cols 0200A–0202C

• @0199D Reliquit enim stultitiam generis sui filium Roboam. Two accusatives (stultitiam, filium) with one verb and no apposition marker. Rendered "he left the folly of his own stock, his son Rehoboam," reading the second as appositive. Flagged; stultitiam may be a displaced gloss.
• @0200D Multo enim melius si dictum fuisset, et ubique verum constaret. Elliptical (esset unexpressed). Rendered "For it would have been much better if it had been said thus, and it would have stood true everywhere," with "thus" supplied by the colon that follows in Migne's sense but not in his type. Flagged as the one place in this batch where a connective word has no exact printed warrant.
• @0200D Vena sollicitudo reprehenditur. Vena ("vein") is a real word but cannot function as an adjective agreeing with sollicitudo, so the run is not renderable as printed. Carried: [sic: Vena]. Conjecture: Vana, certain from the whole surrounding argument about vanitas.
• @0201C Quis est qui obhorreat. obhorreat for abhorreat. Treated as an attested late/medieval assimilation (Pattern 9), rendered normally ("shudders"), no marker. Logged so a later sweep does not churn it.
• @0202A–B Unde tam cito pessimatus [pessundatus] est sensus iste putatis? pessimatus (from late Lat. pessimare, "to make worst") is attested medieval Latin, and Migne has already flagged it himself by printing the bracketed correction [pessundatus]. Per Pattern 12 no [sic:] is stacked on an editor's own note: the verb is rendered ordinarily ("made worst") and Migne's bracket is reproduced verbatim, untranslated, with the column anchor [0202B] in its printed position between them.
• @0202C Si nemo potest quod tu potes, magnum est si bonum potest. Second person potes then third person potest in the apodosis. Rendered as printed ("If no one can do what thou canst do, it is a great thing if he is able for good"). Conjecture: potes for the second verb also. ## Chunk 0034 — cols 0202D–0204D

• @0202C curam surperfluam (Eccl. 2:26, first quotation). Broken type; carried inside the italic lemma as [sic: surperfluam], no nested italic marker. Conjecture: superfluam, certain — Migne prints it correctly when he requotes the same verse at 0203D.
• @0203C Cor stulterum ubi laetitia (Eccle. VII). Broken type; carried as [sic: stulterum]. Conjecture: stultorum, certain from the parallel cor sapientium and from the same verse quoted correctly at 0192D.
• @0203C the [n: (Eccle. VII)] marker is followed in the plate by a comma and then a capital (, Verum quidem hoc est). Punctuation reproduced exactly.
• @0203D Pones eos deorsum, in reliquiis tuis praeparabis vultum eorum (Psal. XX, 13). The Vulgate reads Quoniam pones eos dorsum — "thou shalt make them turn their back." Migne's deorsum ("downward") changes what the verse says, and Hugh's argument here (the gladness of the wicked is not in God's sight) leans on the downward direction. Rendered as Migne prints it, with [var: Vulg. dorsum, "the back"] beside it (Pattern 14). No conjecture as to which witness Migne's copy-text followed.
• @0203D ad malitiam descendendo in occupationem pessimam. Rendered "into the worst occupation," the pre-set rendering of the Eccl. 1:13 lemma occupatio pessima, which Hugh is deliberately recalling here from Homily VI.
• @0204C sed semper a se alterum et sibi adversum atque extra se peregrinum. Three predicates in the neuter agreeing with cor; rendered as printed.
• @0204C Et videmus nunc … et mirabamur quomodo tam cito pessumdetur sensus mortalium. Present videmus with imperfect mirabamur in one sentence. Rendered as printed ("we now see … and we were wondering"). No emendation. Note that pessumdetur here is spelled correctly, which strengthens the case that pessimatus at 0202A–B is the anomaly Migne bracketed.
• @0204C ire patiatur in aliena turpitudinis. aliena (neut. pl.) governing a genitive; unusual but construable. Rendered "suffers itself to go into the alien things of baseness." No emendation.

Chunk 0035 — cols 0205A–0207B

• @0205B–C et fit miro quodam modo, ut qui semper in desiderio futurum fallitur, vix aliquando in experientia praesentium omnino decipiatur. Reads against the argument as it stands — the surrounding paragraph says the mind is undeceived by experience and is deceived by desire, which is what the printed words say when vix … omnino decipiatur is taken as "is scarcely ever wholly deceived." Rendered so, literally. No emendation; flagged only because the double qualification (vix … omnino) is easy to flatten in either direction and must not be.
• @0206C facilitatem tamen conferre non possunt. facilitatem ("ease, facility") where the argument, and the whole homily's vocabulary (veram felicitatem conferre eleven lines earlier at 0205D), require felicitatem ("happiness"). Rendered as printed — "cannot nevertheless confer ease" — under Pattern 7: facilitas is a real word and the sentence construes, so it is not carried type and takes no marker. Conjecture: felicitatem, near-certain. This is exactly the class of one-letter defect that a smooth English would conceal.
• @0206D sed habet suum tempus unumquodque, quando ut sit bonum es et tempus quando bonum est. es for est. Carried: [sic: es]. Conjecture: est, certain from the balancing clause four words later.
• @0207A etiamsi bonum non est illud quod bonum est ut sit … etiamsi bonum est, quod bonum est ut non sit. A deliberate chiastic paradox with three negatives across the period; every one is carried, and the sentence is meant to look self-cancelling. Logged so that no later sweep "repairs" it — this one is Hugh's, not Migne's.
• @0207B Nunc benedic Deum, et morere (Job II, 9). Vulgate: Benedic Deo et morere — dative Deo, and no Nunc. Rendered as Migne prints it ("Now bless God, and die"). No [var:]: the added adverb and the case change do not alter what the verse asserts, and English has no exponent for the dative/accusative difference here.
• @0207A pignora interfecit rendered "he slew the pledges of his house" — pignora in the standard medieval sense of children, kept literal rather than glossed to "children," since Hugh is quoting the Job narrative in its own register.

Chunk 0036 — cols 0207C–0209D

• @0207C Quid causaris sub tempore omnes de legibus temporum? Second-person singular causaris with omnes (nom./acc. pl.). Rendered "Why do you plead, all of you under time, about the laws of times?" — taking omnes as a vocative plural apposition, which is the only reading the printed cases allow. Flagged; omnes causamini would construe but is not what is printed.
• @0207C et non erit ibi tunc, omnia tempus habent. Migne sets the quoted lemma with a bare comma and no marking; rendered with our own straight quotes round it (Pattern 6, marks supplied by us because Migne marks the quotation not at all): there will not then be there, "all things have their time".
• @0208A–B et sunt alia, quae vere non sunt, et ea quae vere sunt: quodammodo aemulantur. The colon falls where the sense wants a full stop, and ea quae vere sunt is the object of aemulantur across it. Punctuation reproduced exactly (Pattern 8); the English keeps the colon and lets the sentence run on.
• @0208D et quod isti tempus fuit, quia non semper fuit; fuit illi aeternitas, quia semper fuit. The semicolon inside the correlative pair is Migne's; reproduced. The whole fuit chain (0208D, fourteen occurrences in one period) is rendered word for word rather than smoothed — the repetition is the argument.
• @0209B–C Tunc longa spatia temporum non placebunt impiis transacta felicitate temporum; et permanentibus culpis quae per tempora contraxerunt. Two ablative absolutes joined to a clause that never receives its second verb. Rendered as printed, the sentence left grammatically open.
• @0209D nusquam sola mala, vel sola bona. hoc enim tempore vitae hujus. Migne sets a lower-case hoc after a full stop. The English capitalizes ("For at this time of this life") — a capital is not a mark that can be "preserved" or "supplied" in the Pattern 8 sense, and the facing Latin column shows the plate's form. Logged for completeness.
• @0209D non est habere bona sine malis, nec mala sine bonis — both negatives carried. # Cruces — Hugh of St Victor, Homiliae in Ecclesiasten (PL 175), chunks 0037–0044

Batch: Homily XIII part 2 (cols 0210A–0215A), Homily XIV (0215B–0219D), Homily XV (0220A–0225D), Homily XVI part 1 (0226A–0233A). Prepared per translation-style.md (Patterns 7, 7a, 7a′, 8, 9, 10, 12, 14).

Work-wide conventions adopted in this batch (for merge reconciliation)

Beyond the pre-set vocabulary (vanitas → "vanity"; Ecclesiastes never "the Preacher"; sententia → "sentiment"; narratio → "narrative"; littera → "letter"; etc.), and consistent with the conventions already recorded in cruces-0000.md:

Version-divergence watch (the work's defining hazard)

No aliam translationem / nostra habet / Hebraeus / LXX passage falls inside chunks 0037–0044. The head of Homily XV (0220A) prints per aliam interpretationem, but this is not a competing Latin version of Ecclesiastes: the homily re-expounds the same Eccl. 3 lemmata (Hom. XIII–XIV had given them literally; XV gives them morally, XVI spiritually). Rendered "according to another interpretation" and flagged here precisely so that a merge does not mistake it for a version note. Hugh's own signal for a version, nostra habet, is absent throughout this batch.

The Eccl. 3 lemmata as Migne prints them here differ from the Vulgate in one place worth recording: at 0212C Hugh prints Tempus congregandi et tempus perdendi where the Vulgate (Eccl. 3:6) has tempus adquirendi et tempus perdendi — and Hugh himself quotes tempus acquirendi everywhere afterwards (0214B, 0222D, 0231C). Rendered as printed in each place ("to gather" at 0212C, "to acquire" elsewhere); no [var:], because Hugh's own later usage supplies the received form and no argument turns on it. Flagged for the merge so nobody normalizes the one occurrence away.

Scripture rendered as Migne prints it

• 0210A — I Cor. XV. Migne: non prius quod spirituale est, sed quod carnale est, prius est. The Vulgate (1 Cor. 15:46) has animale, not carnale, and deinde quod spiritale. Rendered from Migne ("not first that which is spiritual, but that which is carnal, is first"). No [var:]: Hugh's whole argument here is the carnal/spiritual pair of lives, so carnale is doing the work he needs and the substitution is legible in the English without a note. Migne's italic run also closes irregularly (…sed quod carnale est) — the asterisks are reproduced exactly where the plate sets them.
• 0210B — Psal. XXIV (23:10). Universae viae Domini misericordia et veritas. Printed a second time nine lines later without the conjunction — Universae viae Domini misericordia, veritas — and rendered both ways as printed ("mercy and truth" / "mercy, truth"). Not a dittography (Pattern 11): the second is Hugh's own resumption, not a doubled run of type.
• 0219B — I Petr. IV (4:17). Migne: Tempus est, ut incipiat judicium a domo Dei. Vulg. quoniam tempus ut incipiat iudicium de domo Dei. Rendered as printed; no [var:], since a/de does not change the sense.
• 0219C — Thren. IV (Lam. 4:1). Agrees with the Vulgate; rendered with traditional wording.
• 0219C — Isa. LIV. [var:] PLACED. Migne prints Ponam omnes muros tuos in lapides sculptos, omnes filios tuos doctos a Domino. The Vulgate (Isa. 54:12–13) has omnes terminos tuos in lapides desiderabiles … universos filios tuos doctos a Domino. This is material and not cosmetic: Hugh cites the verse as proof that lapides signify the just, and the divergence puts the stones in the walls rather than the borders, which is exactly the figure his argument needs two sentences later ("this refers more properly to the defenses of walls," 0219B). Rendered from Migne, with [var: Vulg. omnes terminos tuos in lapides desiderabiles, "all thy borders of desirable stones"]. Per Pattern 14 no explanation of the cause is offered here.
• 0220A — Maledicta saeviens in necem suam. Introduced as sententiam auctoris but unidentified; it is not scripture in any received form. Rendered literally as printed — "Accursed, raging unto its own slaughter" — with maledicta taken as the feminine nominative it prints (agreeing with the life given over to death). Left unresolved deliberately; do not supply a source.
• 0231B — Cant. I. et non erat dilectus inter ubera — an unmarked allusion to Cant. 1:12 (fasciculus myrrhae dilectus meus mihi, inter ubera mea commorabitur), printed without a [n:]. Rendered literally ("and the beloved was not between the breasts"); the printed negation is carried (see below).

Carried type — [sic: …] (Patterns 7, 10, 12)

Thirteen runs. Each is carried untranslated, in italics, and each is verified as a verbatim substring of its Latin twin.

1. 0210B (chunk 0037) — Nam utraque gustaro oportet. Non-word; no such Latin form exists. Carried: For it behooves him to [sic: gustaro] both. Conjecture: gustare (or gustari), near-certain from utraque and the preceding utra prius accipiat.
2. 0211A (chunk 0037) — Tempus nascendi et tempus moriendi, hoc est devita humana. Run-together type for de vita humana (Pattern 10). The whole run is carried, including humana, which is a correct word: the merge is de+vita, but humana agrees with the swallowed vita and cannot be rendered alone without inventing a noun for it. Carried: this is [sic: devita humana]. Conjecture: de vita humana, "from human life" — certain from the parallel clause four lines later, hoc est de iis quae pertinent ad vitam humanam.
3. 0216A (chunk 0039) — ut majus sit quod minus crat. Non-word for erat. Carried: so that greater be that which less [sic: crat]. The English preserves the Latin word order rather than supplying "was", because supplying it would translate the very type that is being carried. Note also that the paragraph ends here with no terminal stop in Migne — none is supplied (Pattern 8).
4. 0217A (chunk 0039) — vel fingit artifex imitant naturam. The sentence has an exact parallel eight lines earlier and four lines later (artifex imitans naturam, twice), so imitant is a defect and not a variant; but with it the clause does not construe (a plural verb form against a singular subject). Carried: or the craftsman [sic: imitant] nature fashions. Conjecture: imitans, certain. Judged as carried type rather than as a merely-wrong real word (the Pattern 12 boundary) because rendering it as a finite plural verb would produce English that reads as our own error rather than the plate's.
5. 0220D (chunk 0041) — tempus flendi, et tempus ridendi: templus plangendi, et tempus saltandi. Non-word for tempus; the same phrase is printed correctly three lines above and eight lines below. Carried: [sic: templus] to wail. Conjecture: tempus, certain.
6. 0222C (chunk 0041) — quoniam idcirco tempora augustiae data sunt. Non-word; augustia is not a Latin noun. Carried: the times of [sic: augustiae] have been given. Conjecture: angustiae, "of straitness / distress" — near-certain from the following ut caro afflicta … se temperet and from the parallel with the scattering of stones.
7. 0227B (chunk 0043) — et tota bonorum de malo in bonum cuntium. Non-word for euntium. Carried: and the whole life of the good from evil into good, of those [sic: cuntium]. Conjecture: euntium, "going" (gen. pl., agreeing with bonorum) — near-certain.
8. 0228B (chunk 0043) — quando multa desideria introicrunt in illud. Non-word for introierunt. Carried: when many desires [sic: introicrunt] into it. Conjecture: introierunt, certain.
9. 0228C (chunk 0043) — et marcent genimima bona in siccitate. Non-word. Carried: and the good [sic: genimima] wither in the drought. Conjecture: germina (the word Hugh uses six lines earlier, surgunt quippe germina virtutum) — near- certain; genimina ("offspring, produce") is a possible but less likely alternative, and neither is the form printed.
10. 0229D (chunk 0044) — ac deinde bona, vel sanctur infirma, vel mortua vivificetur. Non-word for sanetur. Carried: and then the good, either infirm, be [sic: sanctur], or dead, be made alive. Conjecture: sanetur, certain from the next sentence, quod male infirmum est, bene sanatur.
11. 0230C (chunk 0044) — ad tactum internae dulcedidinis admittatur. Non-word (dittographic syllable) for dulcedinis. Carried: admitted to the touch of inward [sic: dulcedidinis]. Conjecture: dulcedinis, certain. Not marked [d:]: Pattern 11 covers a repeated run the plate sets twice, not a doubled syllable inside a single word, which is broken type and belongs to Patterns 7/12.
12. 0232B (chunk 0044) — in cognitione veritatis peccata dilectione requiescit. See the separate entry below; this one is a judged exception to the Pattern 12 boundary.
13. 0232D (chunk 0044) — ut eos postmodum in judicie tanto gravius abjiciat. Non-word for judicio. Carried: that afterwards in [sic: judicie] he may cast them away. Conjecture: judicio, certain.

The one judged exception — 0232B peccata

Migne prints: altera vero, id est contemplativa, in cognitione veritatis peccata dilectione requiescit. peccata is a real word ("sins", or a feminine perfect participle "having sinned"), which under the Pattern 12 boundary would ordinarily be rendered as ordinary English rather than carried. It is carried anyway, and the departure is recorded here rather than hidden:

• Rendering it "sins" produces a noun with no verb to govern it and no case that fits (requiescit is intransitive), i.e. English that a reader would take for our error.
• Rendering it "having sinned" would assert of the contemplative life that it has sinned — a theological claim Hugh nowhere makes, in a sentence whose whole point is the contemplative life's repose. That is a silent repair of the worst kind: it construes, and it is wrong.
• The conjecture is pacata dilectione, "in a love that is at peace" (or pacata agreeing with altera), which restores both grammar and sense and is a single-letter class of error the plate commits repeatedly in this batch (items 3, 5, 10, 13 above).

So: in the knowledge of truth [sic: peccata] rests by love, and the conjecture lives here. Flagged for the merge as the one place in this batch where the Pattern 12 boundary was crossed deliberately.

Real words that are wrong — rendered literally, no marker (Pattern 12 boundary)

• 0213C (chunk 0038) — ut nihil maneat sub sole, laetitia quae de nativitate vel plantatione suboritur … temperetur. The connective between the two ut-clauses is simply absent in the plate. No conjunction supplied (Pattern 8): the English keeps the asyndeton exactly ("so that nothing may remain under the sun, the gladness which springs up … be tempered"). Conjecture: et before laetitia.
• 0215D–0216A (chunk 0039) — Propterea ergo permaneant opera Dei, quia stabile est … Subjunctive where the argument requires the indicative: the whole paragraph is a demonstration that God's works abide, not an exhortation that they should. Rendered literally as a jussive ("Therefore then let the works of God abide"), per Pattern 7. Conjecture: permanent.
• 0216C (chunk 0039) — Quando ergo naturam operatur Deus, et quando supra naturam operatur natura, Deo non cooperatur. As printed, natura is the subject of the second operatur and also of cooperatur, which makes nature work above nature. Rendered as printed with the subject taken from the printed word order ("When therefore God works nature, and when he works above nature, nature does not work together with God"); the printed non is carried. Conjecture: natura belongs only with cooperatur, i.e. a displaced word rather than a wrong one. Flagged because the agency here is exactly the fifth mechanism of Pattern 7a (reference or agency mis-bound) and a later reader should check the plate.
• 0225C (chunk 0042) — donec veniant ut vel sub judicio cadenti in perpetuum adversa sint omnia, vel feliciter consummatio, subjecta sint universa. consummatio is a real word in the wrong case (nominative where the construction wants an ablative absolute or a preposition). Rendered as printed, in place, without repair: "or, happily, consummation, all things be made subject." Conjecture: feliciter consummatione, "by a happy consummation". No marker (Pattern 12: the word can be rendered).
• 0227B (chunk 0043) — quaedam sunt in quibus commemorantur bona et mala - quaenam in quibus mala et bona. Two things at once: Migne prints a bare hyphen with spaces where the sentence needs a conjunction, and quaenam ("which indeed?", an interrogative) where the parallel demands quaedam ("certain"). Both are reproduced — the hyphen exactly as spaced, and quaenam rendered "which indeed" — so the reader sees a clause that does not construe, which is what the plate has. Conjecture: quaedam, near-certain from the parallel quaedam sunt immediately before and from Septem enim sunt … et iterum septem immediately after.
• 0221B (chunk 0041) — et causantur saepe Deum quando bene cum eis agitur. Rendered literally ("and often lay the blame upon God when it is dealt well with them"). The argument wants men to blame God when it goes badly; the plate says bene. No emendation, and no softening of the connective — this is precisely the class Pattern 7a mechanism 3 warns about. Conjecture: male, or an ironic bene ("when he is in fact dealing well with them"), which is defensible from the surrounding sentence about God's chastening and would need no emendation at all. Both left in the crux; the English states what is printed.

Negation and polarity (Patterns 7a, 7a′) — every printed negative carried

Checked clause by clause. Every non / nec / neque / nisi / nunquam / nullus / nihil the Latin prints appears in the English, and none appears that the Latin does not print. Three places where carrying it makes the sentence fight its own argument, and where the temptation to repair was live:

1. 0222D (chunk 0041) — Propter haec omnia tempus est amplexandi, et tempus longe fieri ab amplexibus: ut in tempore non quaerat homo quod aeternum est, et qui hominis viventis societatem diligit, cogitet semper morituri separationem. Carried: "so that in time man may not seek what is eternal". This states the opposite of the homily's argument — the entire passage exists to explain that embraces are loosed lest the heart of man rest in man and so that he may seek the eternal. Conjectures, in order of plausibility: (a) quod aeternum non est (the negative displaced, "may not seek in time what is not eternal"); (b) in tempore to be read "in a temporal way", so that non attaches to the manner and the sentence stands as printed. Neither is adopted. The printed non stands where it is printed, and this entry exists so the defect is attributable to Migne rather than to us.
2. 0231B (chunk 0044) — et non erat dilectus inter ubera ut fortiter astringeretur. Carried ("and the beloved was not between the breasts"). The negation is the point — the gentile philosophers embraced from afar — and it inverts the Cant. 1:12 allusion it depends on. Sound as printed; recorded only because a Song-of-Songs echo is exactly the kind of place a familiar-verse reflex would supply the affirmative.
3. 0211D (chunk 0037) — quia pro culpa patienti non ponitur quod non ipsius circa se voluntas, sed dispositio desuper justa moderatur. Double negative, both carried, with non ipsius … sed dispositio rendered as the correlative it is ("not his own will … but a just disposition from above"). Construes as printed; no crux beyond the record that both were checked.

Also checked and sound as printed, no repair made: nec grave omnino illud est (0212C); Nam sine Deo ab homine ut aliquid fiat, ita tempus nullum est, sicut justum nihil fieri sine Deo ab homine potest (0211D — two negatives in a comparative frame, both carried); nec sunt bona peccata, tametsi bonum est ut sint ipsa peccata (0214A — the paradox is Hugh's own and is rendered as it stands); non fit tempus ut fiant quia non bene fiunt (0214B); sed non cuncta exire volunt (0223D); nec potuit charitate extranea retineri (0231B).

Other notes

• Column-band sequence in this batch runs continuously 0210A→0233A with no text broken mid-word or mid-construction across any anchor. Anchors verified 1:1 against the Latin twins, in order, for all eight chunks. Per the standing rule no band arithmetic is reported.
• No [n:] in chunks 0038, 0039, 0041, 0042, 0043, 0044; 5 in 0037 and 3 in 0040, all reproduced verbatim, contents untranslated. No [nt:] conversions — every note in this batch is a bare scripture locator, not editorial prose (Pattern 15).
• No guillemets anywhere in chunks 0037–0044, in either language; parity holds trivially (Pattern 6). No " " supplied either — Migne's italic lemmata already carry the quotation function throughout these homilies.
• No [d:], no [ed:] in this batch. Nothing in the digitization is missing against the plate so far as the twins show, and no run is set twice.
• 0230A (chunk 0044) — the paragraph ends …sui consolatione laetificetur with no terminal stop; none supplied (Pattern 8). Same at 0216A, item 3 above.
• 0217D (chunk 0039) — Interim autem post nativitatem ante mortem dum in hac vita mortali vivit, quae sustentamento indiget, et alimentis nutritur, ut subsistat. A subordinate clause with no main verb; printed as a complete sentence. Rendered as printed, incomplete. No conjecture offered — the missing apodosis could be anywhere. # Cruces — 11065 (Hugh of St Victor, Homiliae in Ecclesiasten), chunks 0045–0052

Cols 0233B–0256C. Homilies XVI (concl.), XVII, XVIII, XIX (concl. of the work).

Work-wide conventions adopted in this batch (for the merge)

Pre-set vocabulary from the brief (vanitas → "vanity"; Ecclesiastes kept, never "the Preacher"; distendi → "be distended"; sententia → "sentiment"/"sentence"; narratio → "narrative"; littera/litteralis/moralis/mysticus/tropologia/ allegoria; contemplatio; occulta → "hidden things") is followed throughout. Decisions not pre-set, made here and recorded so peers can reconcile:

• Homily heads (full HOMILIA form throughout this batch, no abbreviations):
• XVI De spirituali intelligentia eorum quae de tempore dicta sunt. → On the spiritual understanding of those things which have been said concerning time.
• XVII De animorum confusione ex temporum transitu. → On the confusion of minds arising from the passing of times. (animorum = "minds", not "souls" — Hugh's subject through the homily is judicia mentium humanarum.)
• XVIII De perversis hominum moribus: et quid ex eis censuerit Ecclesiastes. → On the perverse morals of men: and what Ecclesiastes judged concerning them.
• XIX De innocentium oppressione, et derelictione: et vario ac stulto impiorum de hac vita judicio. → On the oppression and abandonment of the innocent: and on the various and foolish judgment of the impious concerning this life.
• (cont.) markers left untranslated and in place (chunk-division artifact).
• Ecclesiastes lemmata: tempus scindendi / consuendi → "a time of rending / of sewing" (Hugh's whole exposition is needle-and-thread; "rend/sew" carries it). tempus tacendi → "a time of keeping silence".
• disputatio / concertatio / rixari → "disputation" / "struggle" / "to quarrel", kept distinct and constant; they are the spine of Homily XVII.
• calumnia → "calumny" (Eccl. 4:1 lemma at 0252D–0253A), kept distinct from oppressio → "oppression", which Hugh prints alongside it in the gloss.
• laetari → "to be glad", gaudere → "to rejoice", kept apart wherever the pair occurs together (0245A, 0243A).
• irriguum superius / inferius (Jos. 15:19) → "upper watering / lower watering".
• non esse → "not-being" (hyphenated), the substantive of Homily XIX's argument.
• (Josue, XV) at 0233D is an inline parenthetical citation set in italics in the body, not an [n:] note. Reproduced verbatim, untranslated. Likewise (Eccles, IV) inside the lemma at 0253A (note the comma, not a period, after Eccles — printed thus). Not tagged [f:] — homiletic exposition is not a florilegium (standing false-positive rule 4).

Cruces

### 0239D — ut non invenirit homo omne opus invenirit is not a Latin form (inveniat / invenerit expected). Non-word type, carried untranslated under Pattern 7/12: "that man might not [sic: invenirit] every work…". Note per Pattern 7a′: the printed non governs translated text and is carried in the English even though the verb it negates is inside the marker; the sentence is deliberately left as a negative with an untranslatable verb rather than smoothed into "might not find".

### 0239D — et concortationibus muniunt errores suos concortationibus for concertationibus ("struggles"), the word Hugh uses correctly five times elsewhere in this homily. Non-word; carried, [sic: concortationibus].

### 0240C — et rigas coacitantes horrendas Whole run unconstruable. coacitantes is not a word (concitantes conjectured); rigas cannot be a 2nd-sg. present here, and horrendas (acc. fem. pl.) demands a feminine accusative plural noun — rixas concitantes horrendas, "stirring up horrendous quarrels", is the near-certain reading. Because the defect is the run and not either token, the run is carried whole per Pattern 10: "and horrendous [sic: rigas coacitantes]". horrendas is translated, since it is sound type.

### 0241B — cum fit quod male ipse non vult, dicit male fieri Printed non retained: "when that comes to pass which he himself does not evilly will, he says that it is ill done." The clause is awkward but construes as printed and was not repaired; the following cum bene fiat, et ipse male velit is rendered as the concessive it is.

### 0241B–C — putat secum bene agi. cum potius hoc fiat in malum ejus Migne prints a full stop followed by lowercase cum. Preserved (Pattern 8): "…he thinks that it goes well with him. when rather this comes to pass unto his harm". Not merged into one sentence.

### 0241B — ab initio ejusque ad finem illius ejusque where the sense requires ejus usque (so the parallel ab initio suo usque ad finem suum three lines later). Real words, wrongly divided; rendered as printed — "from his beginning and unto the end of him" — rather than emended. No [sic:], because the English on the page is ours (Pattern 12's operational test).

### 0240A–B — consumant in argumenta alterutrum adversum de se copia Does not construe: alterutrum and adversum cannot both be governed, and copia (abl.) hangs loose. Rendered as closely as the printed words allow — "and consume them into arguments one against the other about themselves out of its plenty" — with no word supplied and none suppressed. Flagged as unconstruable rather than adjudicated.

### 0243D — fluctuant eorda hominum eorda for corda. Non-word; carried, [sic: eorda].

### 0245B — qui neque amare vel approhare potest approhare for approbare. Non-word; carried, [sic: approhare]. The printed neque … vel is rendered "neither … nor"; the negation is on our page, not inside the marker.

### 0245B — cum ea quae operatur Deus homo mullatenus possit immutare mullatenus for nullatenus. This is the Pattern 7a′ case in this batch: the broken token is the sentence's only negation, and it governs text that is translated. Marking it alone would have left the English asserting that man can change what God works — the exact opposite of the sentence and of the whole homily. So the negative force is rendered AND the type is marked: "since man can by no means [sic: mullatenus] change the things which God works." Supplying the negative here is not an emendation; it renders the nullatenus Migne's type is a one-letter corruption of, which would otherwise be silently dropped.

### 0246B — licet intermissa ad templum videretur templum where the argument requires tempus ("although it seemed to have been interrupted for a time") — Hugh is explaining that God's ancient judicial truth only appears suspended while he tolerates the wicked; a temple has no place in the sentence. templum is a real word and can be rendered, so it is rendered as printed ("although it seemed to have been interrupted unto the temple") and NOT marked [sic:] (Pattern 12: only carried, untranslated type takes the marker). Conjecture: ad tempus.

### 0248C — Nam, si cognoverit se ex eo homo magnum est homo magnum est does not agree; either magnum est is impersonal ("it is a great thing") with homo the subject of cognoverit, or a word has dropped. Rendered on the first reading, which needs nothing supplied: "For, if a man shall have known himself out of that, it is a great thing…". Logged because a reader may wish to weigh magnus est (of the man).

### 0249A / 0249D / 0250A / 0250C — the three printed forms of Eccl. 3:19 Hugh quotes the verse-tail four times and Migne prints it three different ways. All three are rendered as printed and are not harmonized:

• 0249A et aeque utriusque conditio → "and equally the condition of both" (adverb);
• 0249D and 0250C et aequa utriusque conditio → "and the condition of both equal" (adjective);
• 0250A et similis utriusque conditio → "and the condition of both alike". The third is a different word, not a different case, and it is the one Hugh's gloss leans on (et sic homo moritur sicut illa moriuntur). A conventional English Ecclesiastes collapses all three.

### 0250A — et consummatio finis eadem moriuntur similis Unconstruable as printed: moriuntur (3rd pl.) has no expressed subject inside the clause and similis (nom./gen. sg.) agrees with nothing. Probable consummatio finis eadem; moriuntur similiter. Rendered with nothing supplied and nothing suppressed — "and the consummation of the end the same, they die similar, and they return to the earth of which they were first taken" — so the defect stays visible. similis is a real word and therefore takes no [sic:].

### 0250C, 0251C, 0252B, 0254A — quod habet homo jumento amplius Hugh's running lemma (Eccl. 3:19/3:21) is rendered uniformly "what man has more than the beast", including where Migne prints the negative form nihil habet homo jumento amplius (0250A, 0250C) — "man has nothing more than the beast". Every printed nihil / non / nec in the sequence is carried.

### 0252B — quod non, sic vita ista est, ut requies in ea esse possit Migne prints a comma after non. English word order cannot host it there; the clause is rendered "that this life is not so, that rest can be in it", keeping one comma inside the clause. No mark supplied, none deleted; only the position is unavoidably different.

### 0252D — Haec enim omnia, quae videbuntur tunc, alibi doceat esse verum bonum doceat (subj.) with a plural subject omnia. Number mismatch has no clean English exponent; rendered "may teach", leaving the hortatory force printed. Pattern 9 class — no marker, no churn.

### 0254A — quod homo ipse caligat ad vivendum se caligat ad vivendum se is idiomatically opaque ("grows dark toward living himself"). Rendered close to the printed words — "that man himself grows dark as to his own living" — rather than paraphrased into "man is blind to how to live".

### 0254C — vidit quod industriae bonorum, proximorum invidiae patuit industriae (nom. pl.) with patuit (3rd sg.), and Migne's comma after bonorum falls between the genitive and the noun it depends on. Both preserved: "he saw that the industries of the good lay open to the envy of neighbors." Number mismatch left unrepaired (Pattern 9 / Pattern 7).

### 0254C — paragraph ends ut ea in proximis carpant with no terminal stop Migne prints none. None supplied (Pattern 8). Same at 0236B, where the sentence Propter hoc tempus belli est, et tempus pacis ends the paragraph unstopped.

### 0255B — et otiocitas sua impinguat eum otiocitas for otiositas — and Migne prints otiositate correctly nine lines earlier (0255A, quasi pactum faciens cum otiositate), so this is this plate's own slip and not a period orthography. Non-word; carried, [sic: otiocitas].

### 0255B — quantum in utili vacatione pascitur Reads as printed "in useful leisure", which contradicts the sentence's own point (the flesh is wasted by the famine that follows) and the surrounding attack on idleness; inutili ("useless"), split by a compositor's space, is the near-certain reading and would be a Pattern 10 case. But both fragments are real Latin words forming a grammatical phrase, so the split is not demonstrable from the type alone; rendered literally — "as much as it is fed in profitable leisure" — with the conjecture here. No [sic:], since what the reader sees is our English.

### 0247D–0248A — Vidi sub sole impietatem in loco judicii Hugh's re-quotation inverts the word order of the lemma he printed at 0244A / 0246C (in loco judicii impietatem). Both orders rendered as printed; no normalization.

### 0239C — stultam fecit Deus sapientiam mundi [n: (I Cor. V)] Migne's note points to I Cor. V; the quotation is 1 Cor. 1:20. Note reproduced verbatim per the verbatim rule — not corrected, not commented on in the text.

Six Opus units, blind, 53/53 chunks, no sampling. 20 sites: 13 source-side, 7 ours. Unit JSONs in data/sweeps/polarity-11065/.

Recall was measured, not assumed. Every defect the translation pass had logged sat in a range whose polarity agent was forbidden to read this file, so each was a live recall test. Only u3's @0175C was named in its prompt; the rest were blind and were found anyway (u1 @0117D + @0132C, u2 all four of its range, u4 four of five). The clean chunks can therefore be trusted.

Every genuinely NEW finding was on our side, and all seven are the same failure: a negation or a word rendered as its opposite, reading smoothly, passing verify-english, leaving no trace. Three sat in columns this file had ALREADY logged for something else — the "a logged column is not a checked column" shape that translation-style.md 7a′ was written for this same morning.

Fixed in the English (ours — Wilson's standing policy: our errors are fixed in place)

• @0136C — non incovenienter is a LITOTES. Was "we can now not [sic: incovenienter] take the spirit…", a flat denial of what Hugh immediately goes on to argue (Hic ergo spiritus… omnia lustrat). The negative prefix had been quarantined inside the marker while non survived outside it. Now "not incongruously [sic: incovenienter]", matching his same idiom rendered correctly at 0127B. Textbook 7a′.
• @0171C — Nec possis … clare discernere. Was "And lest thou be able… to discern clearly" — warning the reader against seeing clearly. Hugh warns of LOSING sight (incipias pati caliginem), and the causal clause he attaches (quia imago quae apparet umbra, sola est, non veritas) is a reason one CANNOT discern. Migne prints a full stop before Nec. Now "Nor canst thou, set in the shadow, discern clearly".
• @0200D — male secura mens. Was "a mind ill at ease" — anxious, the opposite. secura = free from care; the passage is vain anxiety swapped for complacent idleness (vanitas mutata est, non dimissa), and the mind then per torporem, et otium… ad turpitudinem relaxetur. Now "a mind harmfully free of care".
• @0211C — negation scope. Was "many… do not die by will". The negation falls on quando tempus est, not on voluntate. Decisive: Etiam ii, qui ante tempus nascuntur, nascuntur quando tempus est — EVERY birth is timely, so the parallel is timeliness, and tempus moriendi aliquando eligit prava voluntas says deaths ARE willed. Now "by will die not when the time is".
• @0241B — scope of male. Was "which he himself does not evilly will", i.e. his willing was not evil. male modifies the not-willing: he wickedly does not want it. The next clause prints et ipse male velit. Now "which he himself evilly does not will". This column was logged and the entry DEFENDED the rendering — but it defended retaining non, which was never the question.
• @0253A — minus distributes over three verbs. et minus gemant, et suspirent, et desiderent ejus consolationem. Was "and groan less, and sigh, and desire his consolation", leaving the last two standing under ne — so the page had God leaving them helpless LEST they desire his consolation, against quia illis alibi reservat consolationes suas. Now "and less groan, and sigh, and desire…".

Logged, NOT changed

• @0249D — exire ab eo qui perverse inhaeserat (u6, low confidence). Our English renders the cui sense ("from that to which he had perversely clung"); Migne prints qui, which inverts who clings to whom. Flagged as a silent normalization of the named qui/cui mechanism. Left pending a reading, but it should not stay silent: either render the printed qui or say plainly why not.

@0150A — the crux STANDS, and here is the evidence that settles it

Two agents pressed the flat declarative Ergo non meditabimur… et requiescere poterit cor nostrum as an under-rendering, since the next three sentences all exhort to seek. The remedy proposed was to read it as a rhetorical question. Rejected, on measurement: Migne uses 403 question marks in this work and points rhetorical questions reliably — every Nunquid…? carries one. So the period here is his considered pointing, not carelessness, and re-mooding the English would emend his punctuation (Pattern 8). The contradiction is real and is the SOURCE's — most likely a dropped nec before requiescere. Rendered as printed, logged, not repaired.

Keep this finding: Migne's question marks in this work are reliable evidence. Where his pointing is declarative, that is a positive signal, not a shrug — do not re-litigate.