MIGNE.APP

Cruces

Adnotatiunculae in Joelem

Hugh of Saint Victor — read the work →

These are the edition's working notes, published as they were written. They record, work by work, where Migne's plate is defective or the reading uncertain: the printed form, what the English does with it, and why. They are not a groomed apparatus. Where a note argues with itself, withdraws an earlier decision, or leaves a question open, that is the record of how the reading was actually reached, and it is left standing — a conjecture that is visible can be checked, and one that has been tidied away cannot.

The English never silently repairs the page. A construable but wrong reading is rendered literally and the conjecture is recorded here, never applied to the text. How the English is produced and checked →

Merged from two agent batches (chunks 0000–0007 and 0008–0014), 2026-07-28.

Work-wide conventions (reconciled at merge)

Chunks 0000–0007 only. Chunks 0008–0014 are covered by a separate agent's cruces-0008.md; the two files are merged afterward.

Conventions: Pattern 7 (render the printed Latin literally, never the intended reading), Pattern 8 (punctuation follows the plate), Pattern 9 (attested medieval spellings and gender mismatches are NOT defects), Pattern 10 (carry the whole broken run), Pattern 12 (carried type is marked [sic: …]).


A. Carried type — non-words, marked [sic: …] in the English

These have no English exponent; the run is carried untranslated, in italics, except where it sits inside an already-italic scripture lemma (nested italics are not expressible, so the asterisks are dropped and only the marker remains).

ChunkColumnPrintedConjectureNote
00010326B (pre-anchor)diecretdiceretAc si diecret: — the formula runs Ac si diceret a dozen times elsewhere in the work. Metathesis of c/e.
00010326B (pre-anchor)sumnosomnoExpergiscimini … a sumno lethalis negligentiae.
00020330BverstrivestriInside the italic lemma de domo Dei verstri sacrificium; the same phrase is printed correctly (Dei vestri) twice within ten lines, in the lemma at 0331D and the repeat at 0332C. Asterisks dropped (lemma already italic).
00030333DdestructurdestrueturA, civitas destructur — parallel clauses either side read ducetur and evertetur, both future.
00030337BeadentcadentInside the italic Joel 2:9 lemma, Sed et per fenestras eadent. Migne prints cadent correctly at 0341C when Hugh re-quotes the same clause. Asterisks dropped.
00040339Dbaereticaehaereticaepluviam contra ariditatem baereticae perversitatis — b/h substitution, a common broken-sort defect.
00040340Am nemaneSplit type (Pattern 10). Dies autem haec quasi m ne expansum fuit super montesmane is printed whole in the lemma at 0336B and again in the gloss at 0340A ("mane illuminatio fidei"), so the reading is certain and the defect is a dropped a sort. Whole run carried, spacing included.
00060345CHincvoceHinc voceRun-together type (Pattern 10). Hincvoce Psalmistae — the formula Hinc voce Psalmistae occurs at 0339C. Both fragments are real words; the run is carried whole.
00060347A (lemma)santificate ×2sanctificatesantificate jejunium … santificate Ecclesiam, both inside the italic Joel 2:15–16 lemma. Migne prints Sanctificate correctly at 0330B and 0332C. Asterisks dropped.
00070347CnoxumnoxamInside the italic Exod. 32:31–32 lemma, aut dimitte eis hanc noxum. hanc is feminine; noxum is not a Latin form. Asterisks dropped.
00070347Dexpicturexpieturut … sacrificio expictur iniquitas.
00070349ADelDeiscientiam Del plusquam holocaustum (Osee 6:6). i/l substitution.
00070349Bamplectandaamplectendao virago, amplectanda, mediatrix peccatorum fidissima! Borderline — a/e in the gerundive stem could be argued an orthographic variant rather than broken type, in which case Pattern 9 would exempt it and the marker should be removed. Flagged for a human.
00070351Aligagavitligavitad quatuor aquilarum colla ligagavit — syllable doubled.

B. Real words, wrong form — rendered as ordinary English, no marker

Per Pattern 12's closing boundary: these can be rendered, so the English on the page is ours even though the reading is Migne's.


C. Punctuation and capitalization carried per Pattern 8


D. Migne's own bracketed matter and his own sic — reproduced, not translated


E. Cases needing a human's judgment

  1. 0006, 0345C — a citation that never became a note. Migne prints Instauras, inquit, testes tuos contra me, et multiplicas iram tuam adversum me, et poenae militant in me (Job X) — the locator sits inside the italic quotation with no [n: …] wrapper, unlike the ~110 other citations in this work. Reproduced verbatim in place, untranslated. I did not tag it [f: (Job X)]: Pattern 4's trigger is a work whose citing practice is inline tails, which is not true here — this is a one-off. But it means one genuine fons will not reach the index unless someone decides otherwise. This is the entry most worth a decision.
  1. 0003, 0336A — displaced punctuation collapses four glosses into two. bestiae, simplices, et idiotae fontes aquarum; doctores populorum ignis, cupiditas, flamma, luxuria. The series demands bestiae, simplices et idiotae; fontes aquarum, doctores populorum; ignis, cupiditas; flamma, luxuria — four lemma/gloss pairs. As printed, the pairs run together and the English necessarily reads oddly ("the beasts, the simple and the unlettered the fountains of waters; the teachers of the peoples a fire, covetousness, a flame, lust"). Rendered as printed per Patterns 7–8.
  1. 0004, 0337D — facies, erat exercitus, ordo, cultus, …. Same class: the comma after facies and the one after exercitus between them destroy the lemma/gloss boundary. Intended almost certainly facies erat exercitus: ordo, cultus, splendor, ritus, forma, modus. Rendered as printed.
  1. 0005, 0343A — Etham diabolus … ejus flamma perversarum sunt sectarum genera. flamma (singular) with sunt (plural), and the sense wants flumina — Hugh is glossing tu siccasti fluvios Etham (Psal. 73:15), where fontes has just been glossed and the fluvii are the natural next member. Rendered as printed, "his flame are the kinds of perverse sects."
  1. 0005, 0342C — conscientias singulorum, spiritum gratiae et precum obtineant. Two accusatives under one verb where the second is likely ablative (spiritu gratiae et precum obtineant — "obtain… by the spirit of grace and of prayers"), which is what the parallel spiritu fortitudinis dissolvant immediately before suggests. Rendered as printed.
  1. 0007, 0347C — in veteri rota. Hinc frequenter legitur in veteri rota, propter Isaac, propter Jacob…. rota ("wheel") is a real word and is rendered, but the sense plainly wants in veteri lege or in veteri Testamento. Possibly nota. Rendered "in the old wheel."
  1. 0007, 0349D — Obmutescat canina facundia saecularis eloquentia. Two nominatives in apposition where a genitive is wanted (saecularis eloquentiae). Rendered as printed, "Let the dogged fluency, secular eloquence, be struck dumb."
  1. 0007, 0348C — utrumlibet a te ponendo. Unconstruable as printed; the series is a te condito, a te redempto, [a te] …ndo, so a passive participle or gerundive of possession (possidendo?) is wanted. Rendered "in either case to be set by thee," which is as close as the printed words allow.
  1. 0004, 0340D — quia vanitatem deserti ipsius redegerunt in contemptum mundi. deserti has no antecedent that yields sense — the apostles are turning the hortus voluptatis into a solitudo deserti, so what they reduce is the vanity of the garden, not of the desert. Possibly horti. Rendered as printed.

F. Deliberate lexical choices (recorded so a sweep does not churn them)

Chunks 0008–0014 only. Chunks 0000–0007 are covered by cruces-0000.md; the two files are merged afterward.

Conventions: Pattern 7 (render the printed Latin literally, never the intended reading), Pattern 8 (punctuation follows the plate), Pattern 9 (attested medieval spellings and gender mismatches are NOT defects), Pattern 10 (carry the whole broken run), Pattern 12 (carried type is marked [sic: …]).


A. Carried type — non-words, marked [sic: …] in the English

Carried untranslated, in italics, except where the run sits inside an already-italic scripture lemma (nested italics are not expressible, so the asterisks are dropped and only the marker remains).

ChunkColumnPrintedConjectureNote
00080353BRubignemRubiginemRubignem dicens: Beati pacifici. Migne prints rubigo in the Joel 2:25 lemma at 0352A and rubiginem correctly in the parallel gloss eight lines earlier ("Contra rubiginem, perfectam patientiam"). Syllable dropped.
00090357AsculumscutumInside the italic Thren. 3:65 lemma, Dabis eis, Domine, sculum cordis laborem tuum. Asterisks dropped.
00100360Bculturomculturamad culturom idolorum compulerunt. Not a Latin form; o/a substitution.
00100361ApucilampuellamTunc pucilam, id est, cordis innocentiam vendunt pro vino. puellam is printed correctly in the Joel 3:3 lemma at 0359A and twice in the gloss at 0360C.
00100361Adicdiecum in dic salutis et in tempore placito visitare dignatur oriens ex alto (II Cor. 6:2). Borderlinedic is a real Latin form (imperative of dico), so Pattern 12's closing boundary would send it to section B; but it admits of no construction whatever in the slot (in + accusative singular imperative), so there is nothing to render and it is carried. Flagged for a human: if the ruling is that any real-word form goes to section B, the marker should be removed and the phrase rendered as broken English.
00100361BfaceritisfaceretisInside the italic Joel 3:6 lemma, ut longe faceritis eos de finibus suis. Asterisks dropped.
00110362BintepretaturinterpretaturSaba siquidem intepretatur captivitas. The verb is printed whole at 0362A (interpretantur), 0364D and 0369D. Dropped r.
00120365Bgloriossimagloriosissimagloriosi certaminis gloriossima praemia — the superlative syllable is contracted, and gloriosi stands correctly in the same phrase.
00130371AtranscunttranseuntNon transcunt, inquam, per eam alieni amplius. transeunt is printed correctly two sentences earlier in the same column and in the lemma at 0367D.

No split-type (Pattern 10) defects occur in this half of the work.


B. Real words, wrong form — rendered as ordinary English, no marker

Per Pattern 12's closing boundary: these can be rendered, so the English on the page is ours even though the reading is Migne's.


C. Punctuation carried per Pattern 8


D. Migne's own bracketed matter — reproduced, not translated


E. Cases needing a human's judgment

  1. Citations printed inside the quotation with no [n: …] wrapper — two more instances, both split by a column anchor. cruces-0000.md §E.1 flags the first (0345C, Job X). My half has: - 0008, 0353D…sicut disposuit mihi Pater meus, regnum (Luc. [0353D] XXII), etc. - 0013, 0370B…quoniam priora transierunt (Apoc. [0370B] XXI).

Both reproduced verbatim in place, untranslated, and not tagged [f: …], on the same reasoning the peer gives: Pattern 4's trigger is a work whose citing practice is inline tails, which this is not. But there are now three in one work, all of them genuine fontes that will not reach the index. Note also that in both cases the anchor falls inside the citation, between the book abbreviation and the numeral — this is the one anchor placement in my seven chunks that breaks a construction, and it is Migne's own line break, not a chunking artifact.

  1. 0008, 0354C — Unde infula dicitur quasi intus fula. fula is not a Latin word, which by the letter of Pattern 12 would make it carried type. It is not: it is Hugh's own etymological splitting of infula, the second half of a deliberate nonce-division, and marking it [sic: …] would blame Migne for the author's wordplay. Left in italics with no marker, alongside the infula it explains. The rulebook does not state this boundary (author's nonce-form vs. compositor's non-word) and probably should.
  1. 0013, 0368B — the Greek etymology of agios. Quod sanctum dicitur Latine, ἄγιος, id est agios dicitur Graece. Agios nomen est compositum ex α et γῆ, id est gi Gi, dicitur terra, a sine. The Greek is reproduced verbatim; the etymological Latin words are left untranslated; and the typography is carried exactly as printed — gi roman, then Gi, italic, with no stop between them, which is why the English sentence runs together. No [sic: …]: gi is Hugh's transliteration, not broken type. Same class as item 2.
  1. 0014, 0371B — demum ad contemplationis avolat. A genitive with no governing noun. The sentence's own opening gloss supplies the obvious word (contemplationum sublimia, five lines above), so the intended reading is near-certain — but a dangling genitive has no English exponent: any rendering either supplies the noun (silent emendation) or produces a phrase English cannot form. Rendered "at last she flies away to those of contemplation," which is the least-supplied option I could find. Worth a ruling, because the class (Latin genitive with elided head noun) will recur.
  1. 0010, 0361A — the dic marker. See section A. The only place in my half where I put a [sic: …] around a form that is technically a real Latin word.

<!-- corpus-polarity-sweep 2026-07-28 --> ## Corpus polarity sweep — 2026-07-28

Blind Opus read of the Latin/English pair for sense-reversing defects (one word or one letter asserting the opposite of what the context requires). 5 site(s). Migne's defects are recorded here and rendered literally in the English; our own errors were fixed in place. Full context: data/sweeps/corpus-polarity/REPORT.md.

0327B — added-negative (migne, medium confidence)

Printed: Rami sunt animae desideria, per quae extenditur usque ad [0327B] opera: qui albi facti sunt, quando nativum virtutis colorem non amittunt.

Conjecture: quando nativum virtutis colorem amittunt (delete non)

The lemma being glossed is a devastation image, and Hugh has just glossed the whole clause negatively in the two sentences immediately preceding: 'Sed hanc gens praefata decorticavit, nudavit, projecit. Decorticavit actu, nudavit sensu, projecit affectu. Actu probitatis, sensu honestatis, affectu bonitatis.' The branches are whitened (barked bare) precisely BY the enemy's stripping, i.e. when the desires lose the native green of virtue; as printed the gloss makes whitening the mark of virtue retained. Compare the parallel gloss on the same lemma at 0325A: 'quia persecutione ingravescente tepuerunt exercitia religionis, siluerunt verba praedicationis, virtutum merita attenuata sunt in multis.'

0339B — sense-reversing-letter (migne, medium confidence)

Printed: Nonnulli in divinitate Deitatis unitatem, sed disparem potestatem in aequalem dignitatem.

Conjecture: sed disparem potestatem, inaequalem dignitatem (word-division: inaequalem set as 'in aequalem')

The clause pairs 'disparem potestatem' (unlike power) with 'aequalem dignitatem' (equal dignity) in the same breath, and the very next sentence sets the opposite party against it: 'Plerique omnium aequalitatem, sed naturae negaverunt unitatem' — the group that granted EQUALITY is the other one, which denied unity of nature. The subordinationist group here must deny both equal power and equal dignity while holding the unity of Deity; 'in aequalem' is 'inaequalem' broken across a word division.

0341D — dropped-word (migne, medium confidence)

Printed: Sic cadendo non demoliuntur, sed potius eriguntur et consolidantur, juxta illud: Omnis qui se exaltat, humiliabitur [n: (Luc. XIV)].

Conjecture: juxta illud: et qui se humiliat, exaltabitur (Luc. XIV, 11b) — the second half of the verse, dropped or displaced by the first

The proof-text is adduced to support the immediately preceding clause 'sed potius eriguntur et consolidantur' — men who fall down (humble themselves) are raised up. As printed, the verse-half quoted asserts the reverse movement (the self-exalter is humbled), which supports nothing in the sentence it is attached to; the required half is 'qui se humiliat, exaltabitur'. The whole surrounding passage is about apostolic men who 'devota horum consideratione, Creatori sese humiliter prosternunt' and are thereby 'not broken'.

0352D — sense-reversing-letter (migne, high confidence)

Printed: Torcularia, ipsa sunt corda inter spem et timorem posita, constantiam inferius, patientiam superius habentia; constantiam in tentatione, patientiam in tribulatione…

Conjecture: constantiam SUPERIUS, patientiam INFERIUS habentia — the two adverbs are transposed in Migne's plate

The very next sentences state the opposite arrangement and are internally consistent with the wine-press image: 'Constantia siquidem premit, et DESUPER arcet; patientia INFERIUS jacet et pondus sustinet' ('For constancy presses down and holds off from above; patience lies beneath and bears the weight'). Constancy cannot be 'below' in one clause and press 'from above' two clauses later; in a torcular the upper beam presses and the lower part bears the weight.

0369A — sense-reversing-letter (migne, medium confidence)

Printed: quando per apostolos et eorum successores evangelica doctrina gentilem populum ad finem initiavit, ad credulitatem convertit, ad baptismi gratiam perduxit

Conjecture: ad FIDEM initiavit (one letter: fidem → finem)

'initiavit' means to begin/initiate, so 'ad finem initiavit' asserts an initiation unto the END — the reverse of what the verb requires; and the two parallel members that follow both name the entry-stages of conversion, 'ad credulitatem convertit, ad baptismi gratiam perduxit' (converted to belief, led to the grace of baptism). The triad is fides → credulitas → baptismus, not 'end'.