Expositio in Cantica canticorum

What this file is. The working editorial record for the English of Foliot's exposition of the Song of Songs: every place in the work where Migne's plate is defective, where his text diverges from the received one, or where the reading is uncertain — given by column, with what he prints, what the English does with it, and why. Declined calls are kept here with their reasons, because a place a reader already examined and let stand is worth as much to the next reader as a place that was marked.

What kind of work this is, and why it shapes the notes. The Expositio is a lemma-and-gloss commentary. Migne prints a verse or half-verse of the Song in italic as a numbered lemma, and Foliot then expounds it for anything from ten lines to two columns, re-quoting his own lemma inside guillemets as he goes — often four or five times, and often not in the same words. Most of the textual interest in this work is in those re-quotations, not in the lemmata: every named Song lemma in Caput III, IV, VI and VII agrees with the Vulgate word for word, while the glosses beneath them print the same clause three and four ways. So an entry below will frequently record that a lemma is clean and that the plate diverges from it eight lines later.

How to read the entries. They run in one ascending sequence by column, 1147A to 1304D, grouped under the work's own chapters — not by the order in which they were written, and not in batches. A reader arriving from a work page with one column in hand should be able to find that column and stop. Each entry opens with its column and (where it helps) the book or lemma at issue, then states the printed reading, then what was done. The markers named in them are the edition's standing apparatus:

Three things this file does not do. It never repairs the page: a construable but wrong reading is rendered literally and the conjecture is recorded here. It never corrects one of Migne's [n: …] citations, however plainly wrong — his notes ride verbatim, and a note that sits in an Ibid. chain is untouchable, because the chain resolves through it. And it does not harmonize this commentary with any other commentary on the Song; each stands on its own.

Work-wide conventions

Fixed once and held through all 53 chunks. Where two readers of different parts of the work reached different answers, the collision and its resolution are named — the resolution, not just the winner.

The Song's vocabulary of address and love

• amica = "friend" — 26 instances, whole-work scope. This was the largest collision in the merge and it was decided by Foliot, not by us. Two ranges fixed "friend" from two independent passages neither could see: at 1196A the gloss turns on the etymology (illam verbo amicabili convenit … dicens, « Amica mea »), and at 1231C Foliot puns on the word himself (ut amica nominum positio totum leviget) and calls the Bride's friends amicos at 1221C in the same breath as the vocative. A third range fixed "my love" on the sound but different ground that amica must stay a THIRD word, distinct from sponsa (Bride) and dilectus (beloved), with the vocative carrying the Bride/Bridegroom gender English loses. "My love" is not wrong about the vocative; it is wrong about who decides. A rendering that cannot carry Foliot's own pun makes his gloss unreadable, and "friend" keeps the third term just as distinct. Merged to friend everywhere, vocatives included ("my friend" where Migne prints amica mea). The masculine twin settles with it: at 1281C talis est Dilectus meus, et ipse est amicus meus puts both words in one line, rendered "such is my Beloved, and he is my friend"; and the work's last lemma, 1303C amici auscultant, is glossed fidelis amicitiae, so the friendship sense is what Foliot expounds at the end as at the beginning. Never "bride."
• sponsa = Bride · sponsus = Bridegroom, with Migne's own capitalization mirrored — he varies it, and the English varies with him (lowercase sponsa in a lemma reads "bride").
• dilectus = "beloved" · capital Dilectus = "the Beloved", again mirroring Migne's own varying capitalization, occurrence by occurrence, and never normalizing it. He prints both cases of the same phrase within a page — 1205B's capital Dilectus against 1205C's lowercase in the identical clause, 1294A's lowercase re-quotation of a lemma he capitalized eight lines above. Foliot's ordinary participial dilecta is "beloved" without promotion.

The praise-words — four Latin adjectives, four English ones

This was the second collision, and it took three ranges and a freed word to resolve.

• pulcher / pulchra = "beautiful", never "fair" — pulchritudo is glossed continuously for four columns (1207A–1210C), and "fair" would leave the gloss expounding a word not on the page. pulcherrima = "most beautiful", deliberately unharmonized where Migne's re-quotations drop the superlative (see 1194B). ⚠ One partial's vocabulary table (chunks 0010–0014) lists pulcher = "fair". No pulcher is rendered "fair" in the shipped English of those chunks — the two "fair"s there are ordinary prose — so the table is a stale entry, not a divergence in the text. Named so that a later sweep does not go hunting.
• decora = "comely" · decor = "comeliness" — Douay's word, and the precedent inside this work is Foliot's own prose at 1194C. The gloss depends on it at 1283B (unde patientiae decorem obtinet) and again at 1292D.
• speciosa = "fair" ("my fair one") — this is the resolution of the three-way collision. One range had fixed speciosa = "comely one" at 1236C; a later range needed "comely" for decora, which is glossed in three separate places and could not move. The word that freed the deadlock is species = "fairness" / speciosus = "fair", fixed at 1281A on 1210B's precedent (Haec forma, haec species, haec pulchritudo → "This is the form, this the fairness, this the beauty"). With speciosus = "fair", speciosa mea reads "my fair one" and all four adjectives stay distinct across the whole work. ⛔ This is a live repair, not a settled fact: the shipped English at 1236C still reads "my comely one." See the open items at the foot of this section.
• formosa = "beautiful" — the word of Nigra sum, sed formosa (1183C) and its fourteen re-quotations. It shares its English with pulchra, which is a flattening; the two never meet in one clause (formosa belongs to Caput I, pulchra enters at 1194A and takes over), so nothing in the work asks the reader to tell them apart.

The two hands

• laeva = "left hand" (a hand, answering dextera "right hand", which is what the gloss at 1221C–1222D argues about: Dominum laevam pariter habere et dexteram); sinistra = "on the left", the bare substantivized adjective with no "hand" supplied, because Migne prints no manus in the Proverbs III, 16 proof-text and Proverbs is describing what Wisdom holds on each side. Two Latin words stay visibly distinct without inventing a distinction of sense. dextera / dextra = "right hand" in both places, since Migne prints the same word in both. ⚠ The ruling was made at 1221C and the earlier columns predate it. At 1173A–1175A the two Latin pairs run through the argument about the two breasts and the two Testaments, and that range's own note records the decision it made instead — "English has one pair of words for two Latin pairs, and the alternation is recorded here rather than invented in the translation." So the work carries a declared flattening in Caput I and a declared distinction from Caput II onward. Not silently harmonized at the merge: the two passages argue different things, and no gloss in Caput I turns on sinistra against laeva. Named here because it is the kind of thing a careful reader will find.

The homograph the work turns on — mala

⛔ malus, -a, -um "evil" · malum "apple" · malus, -i (f.) "apple tree" · malum punicum / malum granatum "pomegranate" · pomum "apple". Every instance in the work was decided from case, gender and governing verb, never from context alone, and the decisions are tabulated range by range in the entries below. The distinction is not pedantry: at 1265B Foliot contrasts the two fruits (pomegranates = the martyrs, apples = the confessors), so a single wrong reading in that chain would destroy the argument and pass every automated check. The two senses collide inside one eleven-word clause at 1297D (ut sibi mala junguntur in malis, et nullis a te potuerunt malis averti, rendered apples–apples–evils) and Foliot slides across the homograph six times in four lines at 1298D–1299A. No apple occurs before 1220C; Caput I is clean, and every mala in it is "evil".

Other terms fixed across the work

• ferculum = "litter" — reconciled across the 0034/0035 seam, and the two readers agreed independently. The exposition proper is at 1252B–1253B, and Foliot's gloss is functional, not etymological (Surgens enim e lecto Rex, ferculo se suscipit, ut de loco feratur ad locum), so nothing in his argument depends on the word's root and every clause construes with "litter" (Douay's word at Cant. 3:9). ⚑ This is deliberately different from 11550, where Richard etymologizes the word (ferculum a ferendo) and it therefore had to stay in italic Latin. Both are right as they stand; each commentary stands on its own.
• capra = "goat" · caprea = "roe", kept distinct wherever they meet — the lemma at 1256B says greges caprarum (goats) and Foliot's gloss on it says capreae (roes). The slip from one to the other is his, and it is now visible instead of smoothed away.
• mammae and ubera both = "breasts" — a DECLARED flattening, not a silent one. Migne prints both Latin words in one verse at 1263A and the gloss uses them as plain synonyms in a single sentence (« quam pulchrae sunt mammae tuae! » id est, doctrina uberum tuorum). "Paps" for ubera would split a word the work treats as one. Stated here because a reader collating the Latin will see two words and one English word.
• custodes = "keepers" · vigiles = "watchmen" · custodiunt = "keep" · circumeunt = "go about". ✅ Checked at the merge and there is no collision: these are two different verses two chapters apart, and Migne agrees with the Vulgate in both — Cant. III, 3 (1247A) is vigiles qui custodiunt, Cant. V, 7 (1275D) is custodes qui circumeunt. They must stay visibly distinct because the Song says two different things and Foliot glosses them differently (at 1247A the vigiles are angels who teach the law of confession; at 1277A the custodes are the Church's prelates and the custodes murorum are the angels). The shipped English of 1247A reads "watchmen" throughout, as it must.
• mater = "mother" · genitrix = "her that bore her" — Foliot argues the distinction twice (1284C Haec vero genitrix, eadem quae et ejus mater est: diversitate tamen verborum, diversa denotat; 1299A), so collapsing them would leave those sentences expounding a difference not on the page.
• charitas = "charity", always — never "love", which belongs to amor and dilectio. The distinction is load-bearing at 1154C: Amor namque in Deum accensus, charitas est is only a statement if the two words differ. amor and dilectio are both "love": they are synonyms in this work, and inventing an English distinction Foliot does not make would be the mirror of harmonizing a lemma.
• lectulus = "little bed" · lectus = "bed" — the two stand side by side at 1252B. virgula = "little rod" against virga "rod", because Foliot argues the diminutive (quae non virga, sed « virgula » recte dicitur, twice at 1249C); Douay's "pillar of smoke" would make the page read "not a pillar, but a pillar". areola = "bed" (of spices) and not "little bed", because there the diminutive is not argued.
• racemus = "berry" · botrus = "cluster" (1284A, ex multis racemis botrus unus). species = "fairness", electus = "chosen", tornatilis = "turned", guttur = "throat" (never "palate" — the gloss apostrophizes O beatum guttur), suavis = "sweet" beside dulcis "sweet", pacificus = "the peaceable one", Sulamitis = "Sulamitess", sexaginta = "sixty" and octoginta = "eighty", never "threescore" — Foliot's gloss is arithmetical for a full column (sive sexies decem, aut decies sex), and "threescore queens" glossed by "the number sixty" would explain a word the page does not print. campus = "plain" against ager = "field", and villa / oppidum / civitas = "village" / "town" / "city", all three ranked in one sentence at 1295B.
• Latin kept visible where the gloss IS the etymology (the stylus/epistylium precedent): murenulae (1199B), monilia with monendo (1289B), collum with colligando (1198D), imber against pluvia in the defining sentence only (1232D), aurora / hora aurea (1285B), ager / agi (1294D), femina vel femora (1289C), ubera from ubertas (1173A), serum gathered sero (1173C), theoria beside contemplatio (1152D), Acoluthus and Sirasirim (the epistle and 1150A), pomum beside malum at 1213D, and the Greek θεοτόκον at 1267B with Foliot's own Latin gloss translated after it, as he wrote it.
• Second person is NUMBER, not reverence (Pattern 17), throughout. The corpus's cleanest worked case is in one sentence of the 1638 dedication — Amplissima Dominatio Vestra → "Your most Ample Lordship" (plural) against Amplitudini Tuae → "Thy Amplitude" (singular); Young switches and the English switches with him. The daughters of Jerusalem take "you"; the Bride, the reader and God take "thou". The number changes mid-quotation where the Latin changes (1258C, 1261A).
• Heads. CAPUT PRIMUM. → CHAPTER ONE., CAPUT II.–CAPUT VIII. → CHAPTER II.– CHAPTER VIII., following the sibling Song commentary 11550/10804 and the corpus majority. Migne's numbered section heads pass through unchanged — including his two lemmata both numbered "5." in one chunk at 1290D and 1291D, which are reproduced and not renumbered. (cont.) is our chunk division and is carried untranslated.

⬜ Open at the merge — two items the merge could not close from the files

1. The 1189A seam — RESOLVED 2026-08-17, and it was wider than one seam. Cant. I, 6 (Indica mihi, quem diligit anima mea, ubi pascas, ubi cubes in meridie) is the lemma of one chunk and is re-quoted and glossed in the next, and the two were written by readers who could not see each other:

Resolved to the left-hand column, and applied work-wide. quem diligit anima mea is not a two-chunk problem: it is printed thirteen times across four chunks (0014 ×4, 0015 ×1, 0032 ×4, 0033 ×4), and three ranges had settled it two ways — nine sites read "loveth", four read "loves". All thirteen now read "loves", and 1193C reads "in the midday" with its lemma at 1189A.

Two independent grounds, neither of them taste: the lemma governs its own re-quotations (the rule already applied at the 0048/0049 seam), and 0014 carries the phrase's FIRST lemma — Cant. I, 6 precedes 0032's Cant. III, 1. The corpus register agrees: across 11550 and 11066 the running prose runs loves 68 / loveth 8, says 480 / saith 3. translation-style.md PERMITS traditional wording where the quoted Latin matches the standard text; it does not require it, and it cannot license one work saying both.

⚑ How it was nearly missed, which is the part worth keeping. A first pass grepped for "pasturest" and "at midday", got almost nothing back, and came within one edit of recording that this entry was stale and the divergence imaginary. It was not: 1193C abridges ubi pascas away entirely (so "pasturest" was never going to appear), and the surviving "at midday" sat deep inside a 400-word line that the search window truncated. The apparatus said the site was there and the search said it was not — and the apparatus was right. Same failure as the Martyrs of Palestine case: an empty grep is not absence.

Both are internally consistent and they cannot both ship. This merge did not choose, because choosing is a translation decision and the fix is an edit to the chunk files, not to this record. The evidence in hand: the lemma-bearing chunk's own gloss runs on feeding (pastorem, pascit Deus, Spirituale pabulum) while keeping "pasture" free for pabulum; Migne's Latin agrees with the Vulgate here, so the scripture rule permits the traditional cadence, and Douay reads feedest … in the midday; 11550, the nearest control, renders the same verse at midday. Whichever is chosen must be applied in both places.

1. 1236C speciosa mea still reads "my comely one" in the shipped English, against the ruling above that frees "comely" for decora and gives speciosa "fair". One line of English has to move for the four praise-words to be distinct across the work.

⬜ STANDING ITEM 1 — THREE SENTENCES WITH NO APODOSIS. It is a class, not an accident.

1159D · 1269D · 1299C. Found by three readers who could not see one another, in three separate parts of the work, each rendering the sentence ending where Migne ends and each deliberately taking no marker. Two made it worth watching; three settles it: this plate drops limbs.

• @1159D — et ut homo de manu Domini bonum sine fine meruit, cum a veritate recessit. The ut clause has no sic clause. The argument requires "as man deserved good without end …, so he deserved evil without end when he departed from the truth", and the second limb is simply not on the plate.
• @1269D–1270A — Cui cum in corde vigilavit ad intelligentiam, qui sibi septem annorum futuram copiam; et septem annorum inediam declaravit. Two relatives (Cui … qui) and no principal verb: the limb that should say Joseph was the one who watched in heart for him is absent.
• @1299C — Haec enim si cohaerentia sunt, cor deliberans, et brachium id quod corde deliberatur exsequens. The protasis is complete, the two participial limbs are complete, and there is no main clause. The very next sentence starts a fresh conditional (Quae si ambo memoriale … tenent … cohibebunt) which does have its apodosis.

Why none of them is marked. [ed: …] is the edition's voice for what our digitization lost, and nothing here is lost on our side: at 1299C the TEI, the Latin chunk and the plate all stop in the same place. There is no marker class for a lacuna the printer left, so the English simply stops where Migne stops and the argument visibly breaks off.

⬜ For Wilson. Three independent sightings is a measurement, not an impression, and the question is what a reader should meet. The options, none free: leave them silent (current behaviour — the argument stops and nothing says so); log all three here (done, and cheap and honest, but invisible to a reader of the text); or open a marker class for a printed lacuna as distinct from a digitization loss. ⚠ The last is a real extension of the apparatus and would need a sweep of the other 105 works, because nothing has ever counted them. Do not open it casually. Note also that all three were found by translating, not by any check — no verifier in the pipeline can see this class.

⬜ STANDING ITEM 2 — "MIGNE AGAINST HIMSELF", and what Pattern 14 cannot say

The work quotes itself constantly, and it diverges from itself when it does. The apparatus has no vocabulary for that. A [var: …] must open by naming an EXTERNAL witness (Vulg. | LXX | Gk | Heb. | Douay | Vet. Lat.) — the validator rejects anything else — and it must not carry content that appears verbatim in the Latin twin, which is exactly what a self-quotation guarantees. So every one of these sites lives here, and the reader of the page sees nothing.

Fourteen sites, and for each the question that actually decides it: would it earn a marker ON MATERIALITY? That distinction is the one the 1199D repair established, and it must not be skipped: "re-opening the content with Vulg. would have made the marker validate while asserting a materiality it doesn't have… A marker that validates is not thereby earned." Passing the validator is a floor, not a warrant — the same shape as "a flagged passage is not a checked passage."

Two further sites of the same family, recorded where they fall: 1238A/1239A and 1248A, where the divergence is real in the Latin and invisible in English (tenui eum against « Tenui illum », introducam against inducam, a dropped nostra), so it lives only here.

⚑ The mechanism, diagnosed and bounded. The blocker is the LENGTH OF THE OVERLAP with the Latin twin, not self-quotation as such. The proof is 1270B, whose [var:] survived the mirror test on exactly the shape that blocked 1176B's — because Migne's own re-quotation there stops short of the full received phrase, so the overlap was partial. A second face of the same mechanism: a [var:] whose content is a single Latin word will trip the mirror test whenever that word occurs anywhere in the twin, which for a common word (Deus, rex) is nearly always. Two markers in this work (1215A, 1216B) had their contents rewritten from single words into phrases purely to clear the test, readings unchanged — so the apparatus silently dictated the marker's prose. Every later [var:] in the work was written phrase-length for that reason.

⚑ The proposal, for Wilson, in two parts. (a) Compare a marker's content against the sentence it sits in rather than the whole chunk. That keeps the protection the test exists for — we still cannot put a [var:] round our own words — while unblocking the case where a work quotes itself twenty lines later and diverges. (b) Consider whether the witness list should gain an internal term (something like Migne @<col>) so a self-divergence can be stated in the apparatus instead of buried here. ⛔ Do not change the validator without Wilson: it is the check that keeps a marker from claiming our prose is Migne's. And argue the case on 1176B or 1194B, which earn a marker on materiality — never on 1199D, which was mechanically blocked and unearned.

✅ STANDING ITEM 3 — THE PLATE-READ QUEUE: ALL FOUR READ 2026-08-18

Read from the archive.org Google scan of PL 202, patrologiaecurs139unkngoog, at 150 dpi for the page and 450 dpi for contested type. Mapping: PDF page = 600 + (column − 1187) / 2, established from p. 600 = cols 1187/1188 under the running head GILBERTI FOLIOT EPISCOPI LONDIN. Record it; it makes any future read of this volume a one-step job.

1. 1303C — Centum vs the lemma's ducenti: OUR TEXT IS RIGHT, AND SO IS MIGNE. The plate prints …per pacificos mille significas, et per ducentos, qui custodiunt fructus ejus. Hi namque perfecti illi sunt, qui ab Adam usque ad Christum… Centum vero qui supersunt, sunt ii qui a Christo usque in finem seculi… ⭐ It is not an arithmetic error at all: Foliot splits the Song's two hundred keepers into two hundreds — the first from Adam to Christ, and "the hundred that remain" from Christ to the end of the age. The number is coherent, the transcription is faithful, and no marker is warranted. Closed.
2. 1202D — pallium for the Vulgate's pullum: MIGNE'S, NOT OURS. At 450 dpi the plate reads « Ligans ad vineam pallium suum, et ad vitem, o fili mi, asinam suam (ibid.). » unambiguously. Our transcription is faithful and the [var:] stands as taken. Closed.
3. 1304D — the missing colophon: THERE IS NONE, AND NOTHING WAS LOST. The plate ends …certissima Sponsi sui promissione certificata conquiescit. followed directly by the printer's ornament. Matches sources/pl/tei/11613.xml, which closes at the same word. The work simply ends without a formula. Closed.
4. ⛔⛔ 1254B — quanam « Aaron »: MIGNE ANSWERED IT HIMSELF, IN A FOOTNOTE OUR SOURCE DROPPED. The plate prints the crux exactly as we have it — quanam « Aaron » (46) filiabus Sion consulitur egredi — and carries a numbered footnote at the foot of the page: “(46) F. earum.” F. is fortasse. Migne's own editors conjecture earum. quanam earum — "by which of them", i.e. of the egressiones just enumerated — construes perfectly, and it is the reading the translating agent independently favoured from the paragraph's taxonomy of goings-forth before rejecting it for want of a witness. ⭐ The agent was right to decline. It had no way to know the witness existed, because it is not in any file we hold. See the section below — this is not a one-site problem.

⛔⛔⛔ A WHOLE APPARATUS LAYER IS MISSING FROM OUR SOURCE — corpus-wide, found 2026-08-18

Migne prints numbered textual footnotes at the foot of his pages, and Corpus Corporum's TEI does not contain them. Three of the four pages rendered for the reads above carry one:

The numbering reaches 72 by col. 1303, so this work alone carries on the order of seventy-five of them across its 158 columns. We carry zero.

This is not a transcription failure of ours. sources/pl/tei/11613.xml holds 413 <note> elements and our chunks hold 413 [n:] — exact parity. Those 413 are Migne's inline parenthetical citations ((Gen. XLIX, 11)). The numbered foot-of-page notes are a different apparatus layer, and the TEI has no representation of it at all. ⭐ Measured across 400 TEI files: ZERO notes begin F. or f. — the conventional fortasse form. The layer is absent from the whole PL corpus we build on, not just this work.

⚑ Why it matters more than its size suggests: these notes are Migne's editors' own textual conjectures — precisely the apparatus that resolves cruces. 1254B is the worked example: we declined to conjecture at a site where Migne had already published his conjecture. Every [cj:] this project has ever declined may have had an answer printed at the foot of its own page, and no reader of our files could have known.

⬜ WILSON'S TO RULE, and it is a scope decision, not a marker decision:

• Narrow: mark 1254B with [ed: …] (the runbook's definition fits exactly — "only where the DIGITIZATION has lost text the plate carries") and move on. One site, one session.
• Medium: recover this work's ~75 notes from the plate now, while the scan is on disk and the page↔column mapping is written down, and add them as a new marker class.
• Wide: treat it as a corpus-wide gap like the PG band letters — 106 works, every volume a separate scan. ⚠ Large, and it would want a pilot on one work before any commitment.

⛔ Do not start the wide version on my reading of one volume. What is certain: the layer exists on the plate, it is absent from our source, and it contains conjectures that change how cruces are decided.

⬜ STANDING ITEM 3 — THE PLATE-READ QUEUE (original entries, now answered above)

Four sites where our files cannot settle the question and one look at PL 202 can. In priority order.

1. @1303C — Centum against the lemma's ducenti. The lemma prints ducenti, two hundred, and the gloss's first limb correctly renders it ("by the two hundred, who keep the fruits thereof") — and then the very next sentence divides them and speaks of « Centum vero qui supersunt », "the hundred that remain." Two hundred split into a first group and a hundred remaining. Deliberately unmarked: a number is neither broken type nor a negation nor a correlative, so no existing pattern reaches it, and [sic:] cannot wrap perfectly good type. The conjecture is that Centum stands for a second hundred, or that the plate has dropped a C. The cheapest decisive read in the work.
2. @1202D — pallium suum for Gen. XLIX, 11's pullum suum. The colt vanishes and a cloak takes its place — and pallium has just been correctly glossed two lines above, from the other half of the same verse, as vestem exteriorem, so the printed reading makes the verse say pallium twice. The work gives no independent witness either way: Foliot's gloss (ad quem asina alligata est) covers only the she-ass and never the colt, so it is compatible with what is printed. A [var:] ships on the received reading; the plate decides whether Migne's compositor or Migne's copy is responsible.
3. @1254B — quanam with no noun to agree with, and two good conjectures competing. quanam « Aaron » filiabus Sion consulitur egredi…? — a feminine ablative interrogative with nothing to agree with, and « Aaron », in Migne's own guillemets, standing where that noun should be, in a paragraph that has nothing to do with Aaron. The two readings: egressione ("by what going forth"), much the stronger contextually — the whole paragraph is a taxonomy of egressiones, the preceding sentence ends on egressio and the following clause answers non ea utique qua Cain egressus est; and ratione ("by what reasoning"), the ordinary idiom for quanam, palaeographically no worse. Two competing good readings is exactly when Pattern 18 says decline and log, which is what happened. A plate read decides it; a vote does not.
4. @1304D — the possibly-missing colophon. The work ends mid-thought on conquiescit. with no colophon, no Explicit, no Amen and no closing note. ✅ This was verified at the SOURCE, and the check is the point: sources/pl/tei/11613.xml closes straight from that word into </p></div2></div1></body> — no <trailer>, no <closer>, no trailing note, no further <pb/>. Nothing was dropped between source and chunk, which is the thing our files can actually settle. ⚠ What they cannot settle is whether Migne's plate carries a colophon that Corpus Corporum dropped, and a twelfth-century Song commentary ending with no formula at all is unusual. Worth one look, on the principle that where our text lacks something the genre always has, suspect OUR files first.

⚑ A locating heuristic worth testing while the plate is open. Three suspect words in one range — « Aaron » (1254B), « videbat » (1258B), « munitos » (1261A) — all sit inside Migne's own guillemets, and all three [cj:] calls on them were declined. That may be coincidence at n=3, but if it holds it locates defects: the guillemets mark where his compositor was setting quoted matter, which is where type most often went wrong. Test it before believing it — the corpus is its own control.

A note on Migne's citations, which are never corrected

His [n: …] notes are his, and they ride verbatim and untranslated wherever he prints them, including where they are plainly wrong and including where he sets them in the wrong place. Nothing in this work has been proposed for citation-corrections.json, for two standing reasons: agreement between readers is not evidence (two of six such reports collapsed on one-at-a-time checking in an earlier work, because the author's own prose named the book), and a reference that a following Ibid. depends on can never be touched, because the chain resolves through it. The wrong or misplaced references found in this work are listed at their own columns below — 1160C-area Matth. VIII, 30 · Matth. XV, 6 · Matth. XV, 41 · Act. IX, 45 · Isa. LIV, 4 · 1197B IV Reg. XIX, 39 · 1225D Psal. V, 17 · 1225C Psal. XCXVI, 5 · 1230B I Tim. I, 13 · 1240B Psal. XXVIII, 10 · 1244B III Reg. II, 11 · 1247B Matth. X, 18 · 1253C/1253D the two Genesis notes one sentence out of place · 1258C and 1261A Deut. XXVIII, 18, 19 for XVIII · 1265C Luc. XIV, 31 for XIV, 11 · 1269B Gen. XXVIII, 12 · 1269C ibid., 11 · 1272A Psal. CXVIII, 119 for 129 · 1274C Matth. V, 3 and the Matth. VIII Ibid. chain that follows it.

The dedicatory epistle and the Preface — 1147A–1150C

Patrick Young's 1638 dedication to William Juxon, Foliot's letter to Robert, and the Gilberti praefatio, before the exposition proper begins.

• @1147A — amico suo charissimo, Foliot to Robert. An ordinary masculine amicus, rendered "friend"; not the Song's amica and not "bride". Recorded because the count of amicus against amica is what the whole-work ruling rests on.
• @1147A (the epistle) — [sic: Reverendissim]. The vocative has lost its final -e in Young's dedication. It sits inside an already-italic run, so it takes no second italic marker.
• @1147A (the epistle) — Acoluthus kept in italic Latin. Young means the companion piece printed with the Expositio in 1638 ("for to go forth into public alone, without a companion, was neither seemly"); "acolyte" would carry a liturgical sense the sentence does not have.
• @1147A (the epistle) — Pattern 17's cleanest worked case, and both halves are in one sentence. Amplissima Dominatio Vestra → "Your most Ample Lordship", grammatically plural; Amplitudini Tuae → "Thy Amplitude", grammatically singular. Young switches; the English switches with him. Number, never reverence.
• @1148A · Cant. VIII, 6 — Fortis est ut mors dilectio, where the Vulgate has quia before it. Simple abridgement; no [var:].
• @1149A · Gen. I, 2 — [var: Vulg. inanis et vacua]. Migne prints terra autem erat informis, the Old-Latin/patristic form. Marked because Foliot's etc. stops the quotation immediately after, so the divergent word is the last one he prints and the reader has nothing else to go on.
• @1149A · Exod. I, 1 — ingressi sunt in Aegyptum for the Vulgate's ingressi sunt Aegyptum. Preposition only; no marker.
• @1149A — an unstated subject that inverts the sense. haec, quos irae filios, dictante justitia, perfecerat, in libertatem gratiae … reformat. The subject of both verbs is grammatically haec (= dilectio), which makes love the maker of the sons of wrath; the sense wants justitia, which the ablative absolute has already used up. Rendered with a neutral "it", preserving the ambiguity rather than choosing.
• @1149D — Seorsum vos, abstractus a vobis.* The one plural second person in Gilbert's own prose, addressed by the soul to the world's goods she is dismissing. Rendered "you"; everything else in his prose is tu* / "thou".
• @1150A — [sic: implieetur] for implicetur, ie for c. It is the clause's only verb, so the sense stands in plain words beside the marker: "be not entangled [sic: implieetur] in these things".
• @1150A — Sirasirim, twice, left as printed. Foliot's transliteration of the Hebrew title; he glosses it himself in the next breath (quod « Canticum canticorum » appellatur).
• @1150A ff. — Parabolae against Proverbia: a real distinction, not a synonym. Solomon's first book is « Parabolarum » here and Parabolas / Parabolarum volumine throughout the following column, and the lemma at 1150D is Parabolae Salomonis — all rendered "Parables", because the gloss is about the book's title. But at 1152D Migne writes jam edocta in Proverbiis, rendered "in the Proverbs". Foliot uses both names for one book; the English keeps both, because collapsing them would conceal that he does.
• @1150B — Animam igitur, vel Ecclesiam: the sentence that sets the work's allegory. Ecclesia = Church and anima = soul are fixed from here, and they meet head-on in this clause. Migne prints ecclesia lowercase through 1159B–C and capitalized elsewhere; followed, not regularized. Note also lowercase sponsum here against the capitalized Sponso/Sponsum of 1153B–C — again followed.

Caput I — 1149C–1211D

Lemma 1, Osculetur me osculo oris sui — 1149C–1172B

• @1149C · lemma 1 · Cant. I, 1 — Osculetur me osculo oris sui. Agrees with the Vulgate exactly; no [var:]. Fixed English: "Let him kiss me with the kiss of his mouth," held identical through twenty-one further re-quotations. Only the terminal mark varies, and it varies because Migne varies it, so it is carried each time: ? at 1149C and 1153B (both inside guillemets); : at 1149D, 1154A, 1163A; ; at 1162D (in the acsi dicat frame); . everywhere else. The lemma is also fragmented into its own words twice — « Osculetur » alone at 1163A and « Osculum » alone at 1163C — rendered « Let him kiss » and « A kiss », matching the fixed rendering word for word.
• @1151A · I Cor. XV, 46 — a missing non, checked under 7a and DECLINED. Migne prints Prius namque quod animale est, deinde quod spirituale, where the Vulgate states this by negation (sed non prius quod spiritale est). Nothing is asserted here that the Vulgate denies: the abridgement converts the Vulgate's own negative construction into the affirmative it entails, and Foliot's gloss (first the animal, then the spiritual) is what both forms say. Recorded because a reader comparing the two will notice the missing non — and because a dropped negation is the one class that must never be passed over silently.
• @1151B · Heb. XI, 6 — [var: Vulg. inquirentibus se]. Migne prints quod diligentibus se remunerator sit — seeking becomes loving. Load-bearing in a commentary whose entire argument is that love, not inquiry, reaches God; the gloss two sentences later builds on amor.
• @1151B and @1163A · Heb. XI, 6 — the same verse printed two ways within one work. sine fide placere Deo est impossibile here, sine fide Deo placere impossibile est at 1163A. Word order only. This is Migne's inconsistency, not ours; both rendered as printed, and the English differs in nothing but the same word order.
• @1151D · Gen. I, 28 — universis quae moventur for Vulg. universis animantibus quae moventur. No marker.
• @1152A · Eccle. XII, 1 — Memento Domini Creatoris tui for Vulg. memento creatoris tui. No marker.
• @1152A · Eccle. XII, 13 — hoc est omnis homo for Vulg. hoc est enim omnis homo. Declined deliberately: Foliot glosses the clause three separate ways at 1152C and none of the three turns on enim.
• @1152B–C — amor / dilectio / charitas, and why only one of the three is fixed. All three are in play here and at 1154C, where Foliot varies them for style within two sentences. charitas is always "charity", because Amor namque in Deum accensus, charitas est is a statement only if the words differ. amor and dilectio are both "love"; they are synonyms here, and inventing an English distinction Foliot does not make would be the mirror of harmonizing a lemma.
• @1152C — the lemma mandata ejus observa re-quoted as « mandata servare ».* Both rendered "keep the commandments": the verb is synonymous and the gloss turns on timere/amando*, not on the verb of keeping.
• @1152D — theoria kept in italic Latin. Foliot has contemplatio twelve times in the same chunk and builds an argument on it (amorem contemplatio suscitet); rendering theoria "contemplation" too would fuse two words he keeps apart at the very moment he is defining one of them.
• @1153B — an imperfect subjunctive among perfects. Sponsum prorsus nesciret in a chain of perceperit … contigerit. Rendered as a past ("knew nothing at all of the Bridegroom"); no marker, since the tense is expressible and nothing is asserted falsely.
• @1153C, @1153D and @1158A — three nunquid questions that close with a FULL STOP. (dicat tamen ad Patrem omnium · singultus poenitentiae trahens · in terram proni corruerunt.) Rendered with the printed stop. Not defects of ours; do not "restore" the marks — and expect the ?-parity check to be quiet here.
• @1153D · Rom. VII, 24 — O quis me liberabit a corpore for Vulg. quis me liberabit de corpore. No marker.
• @1154A · Joan. III, 29 — recast from the Vulgate's amicus autem sponsi qui stat et audit eum into cum amicis Sponsi stat, et gaudio gaudet ad vocem ejus. Note that amicus here is the Song's third love-word in its Johannine use ("friends of the Bridegroom") and is kept distinct from sponsa.
• @1154C — a stranded adjective, and the English strands it too. quasdam sibi praelibari velut arrhas expostulans, interno; solumque … audibili sono mentis exclamans. interno agrees with sono four words later, and Migne's semicolon severs them. The mark rides as printed, so the English reads "with an inward; and crying out with a sound of the mind…". It reads as broken because it is broken.
• @1154C — an ambiguous in ipsam. dedignans totum quod creatum est, in ipsam tota colligit, ipsi nexu charitatis … adstringitur. The feminine ipsam has two candidates, sapientiam (four words back) and the soul herself. Read as sapientiam, since ipsi in the next clause must be the same referent and the soul cannot be bound to herself by charity.
• @1155C · II Cor. XI, 2 — [var: Vulg. Christo]. Migne prints virginem castam exhibere casto — "to a chaste one" for "to Christ". Marked precisely because Foliot's point is the Bride's own cleanness, so the divergence is invisible in the argument and would have been silently smoothed by any conventional English.
• @1155C · Ephes. V, 27 — maculam, neque rugam for Vulg. aut rugam. The two nouns are then glossed separately, in guillemets, so both are held fixed as "spot" and "wrinkle". No marker.
• @1155C · Prov. IV, 23 — quoniam for Vulg. quia. Note also that Migne prints the full stop inside the quotation before the note and again after it (procedit. [n]. »); both stops are carried.
• @1155A and @1159A — tomus against tractatus, the work's own division-words. tomo (sequenti tomo 1155A, tomo superiore 1159A) = tome; tractatum (1159A, 1163C in tractatum sequentem) = treatise. Both appear within four lines of each other at 1159A, which is what forces the distinction.
• @1156B — [sic: eximiii] for eximii, three i's. An adjective only; nothing governed is lost, so the marker stands bare.
• @1156B · I Cor. X, 4 — no defect, logged so no sweep churns it. ad petram, qui Christus est: a masculine relative on a feminine antecedent. Latin gender has no English exponent; translated normally.
• @1156D · Phil. I, 21+23 — Dissolvi cupio for Vulg. desiderium habens dissolvi. No marker.
• @1156D · Psal. CXIX, 5 — the range's single !, and it is Migne's. « Heu mihi! quod incolatus meus prolongatus est », where the Vulgate reads quia for quod. The mark is carried.
• @1157A · Prov. XIV, 13 — novissima gaudii for Vulg. extrema gaudii. No marker.
• @1158A–B · Matth. XI, 11 — both halves recast into indirect/comparative form (Nemo … major sit; ipso longe sit major). longe is Foliot's, not the Vulgate's. No marker.
• @1158D — [nt: …], the only note in the range that is not a citation. Migne's (Nam totum quod in osculo misera peccat, et carnalis affectio, ab ejus mente tam longe est, quam ipsa procul a vitio, virtute purificata, secessit.) is an editorial statement, not a locator, so it is translated and re-marked [nt: …] at the same index.
• @1159B — Sponsi vero ad sponsam, the one place in the opening columns where the pair meets in a single clause. Bride / Bridegroom fixed from here.
• @1159D · Luc. XVI, 10 — [var:] DECLINED, twice over. Migne prints Qui modico fidelis non est, quomodo erit fidelis in magno?, a considerable recasting of qui fidelis est in minimo et in maiori fidelis est into a rhetorical question (his ? is preserved). Declined because the entailment Foliot draws — contempt in a small matter is contempt — is the verse's own, and because naming a single "received reading" would mean choosing between Luc. XVI, 10 and XVI, 11 on our own authority.
• @1159D ⭐ — A SENTENCE WITH NO APODOSIS. et ut homo de manu Domini bonum sine fine meruit, cum a veritate recessit. The ut clause has no sic clause: the argument requires "as man deserved good without end …, so he deserved evil without end when he departed from the truth", and the second limb is simply not on the plate. Rendered as printed, ending where Migne ends. No marker: nothing is broken and nothing is a wrong word — text is missing, and [ed:] is reserved for what our digitization lost. See Standing Item 1: this is the first of three.
• @1160C · Col. I, 26 — sacramentum for the Clementine mysterium, not marked. The words are synonyms in this register, Vulgate witnesses vary at this verse, and Foliot uses sacramentum as his own term throughout (Incarnationis hoc sacramentum, fidei Sacramentis), so an English that split them would misrepresent him. "Sacrament" wherever sacramentum stands.
• @1160C ⚠ — [var:] DECLINED on principle, and the reason is now the work's rule. Hoc consilium illud antequam est, de quo propheta ait, « Consilium tuum, Domine, antiquum verum fiat. » Migne prints no note here, and the wording matches no Vulgate verse; it is close to the Old-Latin form of Isa. XXV, 1 (Vulg. cogitationes antiquas fideles amen). Identifying the verse is our inference, not his, and a [var:] whose "received reading" rests on our own identification of an unlabelled quotation asserts more than we know. Rendered as printed. ⚑ This precedent is applied unchanged at 1242C, 1276D and 1292A. (Note that at 1214D, where Migne does print the locator, the same verse takes a marker.)
• @1160C — [cj: antequam; read antiquum, "ancient"]. Hoc consilium illud antequam est. antequam is a perfectly good conjunction, so nothing is broken and [sic:] has nothing to wrap; but the sentence will not construe, and the required word is the adjective antiquum — which Migne prints in the very quotation that follows, and which Foliot's own a seculis absconditum then glosses. Additive: "This counsel is that before [cj: antequam; read antiquum, "ancient"]".
• @1161A · Psal. CX, 2 and XVIII, 10 — word order only (opera Domini magna sunt; ipsius vera sunt judicia … in semetipsis), and ipsius for the Vulgate's Domini. No marker.
• @1161A — [sic: quitenebantur] for qui tenebantur, a word-division defect; the whole run is carried. The run swallows the relative pronoun and the verb, so the English says it in plain words beside the marker: "those who were held [sic: quitenebantur] in death".
• @1161A–B · Gen. III, 6 — abridged, with tulit mulier de fructu illius where the Vulgate's subject stands earlier in the verse. No marker.
• @1161C — caro / spiritus meet head-on: non delectatur hic caro; sensus ipse rationalis in spiritu carnem trahit post se. Flesh / spirit fixed from here.
• @1161D · Col. II, 14 — [var:] DECLINED. Migne prints inside guillemets a present-tense paraphrase (cruci Christi potenter affigitur, et deletur) of a verse the Vulgate gives as participles round delens … adfigens illud cruci. Rendered as printed. It is Foliot's own syntax, not a divergent reading of the verse, and Pattern 14 exists for the latter.
• @1162A — [sic: proculdio] for procul dubio, fusion plus loss. The clause keeps its predicate (satisfactum [est]), so no supply is needed; the conjecture is certain because the same phrase is printed correctly as proculdubio twice nearby (1159D, 1162B).
• @1163A · Jac. I, 6 — Postulet autem unusquisque in fide for Vulg. postulet autem in fide. No marker.
• @1163A — an active verb where a passive is wanted. quae sine fide est oratio, exaudire non potest. Printed active ("cannot hear"); the sense is exaudiri ("cannot be heard"). Rendered literally with God as the implied subject — "and the prayer that is without faith he cannot hear" — which is the closest English the printed word allows. [cj:] declined: the printed reading, so construed, is not false, only strained.
• @1163B — lowercase after a full stop. …nominatur. osculans, ipse Pater est. English has no exponent for a Latin lowercase mid-sentence-start, exactly as it has none for a gender mismatch; capitalized in the English, and the Latin column preserves it. Logged so a later sweep does not churn it.
• @1164B — « osculari » glossed as the verb, rendered « kiss », from the same fixed words.
• @1164C–D — the six kiss proof-texts, quoted loosely; all DECLINED. Gen. XXVII, 26–27 (et accessit for accessit); Gen. XXIX, 13 (complexus without eum); Gen. XXXIII, 4 (stringensque collum ejus); Gen. XLV, 14–15 (compressed: cum amplexatus recidisset in collum Benjamin … flevit, the Vulgate's mutual weeping dropped); Luc. XV, 20 (vidit pater ipsius, without illum). Sense unaffected in every case.
• @1163D — ipsam with no printed antecedent. qui de admisso unius pro omnibus membris suis ipsam quandoque destrueret: the feminine ipsam has nothing to refer to (admisso is neuter); probably mortem or culpam from the sense. Rendered "would … destroy that same thing"; nothing is supplied.
• @1165A / @1169B — the same Luke chapter quoted two ways by the same work. At 1165A Luc. XV, 6 is quoted in full and in Vulgate form; at 1169B « Congratulamini mihi de ove perdita. Congratulamini de drachma inventa » compresses vv. 6 and 9. Recorded as an internal inconsistency of the plate's quoting practice; no marker.
• @1165B · Gen. XLV, 1 — the quotation runs the verse into et osculatus est eos, which is Gen. XLV, 15. A conflation of the same scene; the gloss depends on the kiss, which is genuinely in the chapter. Declined.
• @1165B / @1174A · Matt. XXV, 41 — the work quotes one verse two ways, and the second takes a [var:]. Here Migne prints the Vulgate's Discedite a me maledicti … qui paratus est; at 1174A he prints Ite, maledicti, in ignem aeternum, qui praeparatus est. A work quoting one verse two ways is exactly the datum this edition exists to show. [var:] fired at 1174A. (Migne's note there reads (Matth. XV, 41) for XXV, 41 — kept verbatim.)
• @1165D · Matt. XXV, 35 — Esurivi, et dedistis without the Vulgate's enim. Declined.
• @1165D — cujus quidem bonus illa die et universis applaudet, et singulis. Read bonis ("will applaud her goods, both all and each"), which is what the next sentence goes on to enumerate — the six clauses of Matt. XXV. [cj:] declined: the printed reading is obscure, not false, and Pattern 18's bar is a sentence that says something the author is arguing against.
• @1166A — Misericordia quippe sunt opera. misericordia cannot agree with the neuter plural opera; read misericordiae (genitive, "works of mercy"). Rendered "For the works are mercy". A case slip, with no English exponent worth marking.
• @1166D · Matt. XXV, 36 / 43 — Nudus fui, et operuistis me and Hospes fui, et non collegistis me for nudus … et, hospes eram. Foliot's parallel phrasing across the six clauses; no sense at stake. Declined.
• @1166D · Matt. VIII, 20 — ubi caput suum reclinet; the note prints (Matth. VIII, 30) for VIII, 20, kept verbatim.
• @1167B · Apoc. III, 20 — Ecce, ego sto ad ostium pulsans for ecce sto ad ostium et pulso. Participle for finite verb; assertion identical. Declined.
• @1167B and @1167D — Migne prints NO full stop, at …omnimoda charitate et virtute Ipse namque ait and at …virtuti jam servire compellant De iis ipsa Veritas ait. The English does not supply the marks. Both read abruptly, and that is the plate.
• @1167D · Rom. VI, 13 — [var: Vulg. arma iniquitatis]. Migne prints « vasa iniquitatis ad iniquitatem » (and ad iniquitatem is from VI, 19). The image changes — vessels for weapons — and the gloss that follows turns on the members serving.
• @1167D · IV Reg. IV, 34 — condensed to Os suum ori ejus … applicat, present tense, the Vulgate's three posuit … super clauses collapsed. Recognizably the same act; declined.
• @1168A · Matt. IX, 12 — Non est opus sanis medicus for valentibus medico. Case and lexis both differ; the proposition is the same. Declined.
• @1168C · Matt. XXI, 18 — Veniens autem Dominus ad ficulneam, esurivit, Foliot's own weave of vv. 18–19. Declined.
• @1169A · Joan. XIV, 26 — Spiritum quem mittit Pater for the Vulgate's future mittet. Nothing in the gloss turns on the tense; declined deliberately rather than overlooked.
• @1169A · Joan. XV, 26 — Mittam vobis Spiritum, an abridgement. Declined.
• @1169A — Migne's full stop inside a participial construction. Sic ad Patrem loquens, os ejus, juxta hunc sensum. Filium nominans et … Spiritum sanctum intelligens. The second sentence is left with no finite verb. Carried as printed.
• @1169B — the note reads (Matth. XV, 6) for the Euge serve bone of Matt. XXV, 21. Kept verbatim.
• @1169D — [sic: carnem]. nuptiale carnem est quod tractatur. carnem is a real word with no slot here: a feminine accusative under the neuter nuptiale in a nominative predicate. Read carmen, "song" — the book is being glossed as an epithalamium two clauses earlier, and nuptiale carmen is the standard phrase. It is carried rather than rendered because English nouns show no case, so "it is a nuptial flesh that is treated" would be fluent English that silently repairs the disagreement.
• @1170A · Isa. LXIV, 4 — [var: Vulg. expectantibus te]. Migne prints quae praeparasti diligentibus te. Material, and it is the whole point of the sentence: Foliot immediately adduces I Cor. II, 9 with Juxta quod Paulus ait, and the two texts agree only because Isaias has been conformed to Paul's diligentibus se. (Migne's note prints Isa. LIV, 4 for LXIV, 4; kept verbatim.)
• @1170A · I Cor. II, 9 — the clause nec auris audivit is absent and his qui diligunt eum becomes diligentibus se. Abridgement; declined — and see the Isaias entry above, which is where the divergence does its work.
• @1170B · I Cor. XV, 24, 28 — Tradet enim Dominus Jesus regnum Deo et Patri supplies a subject the Vulgate's cum tradiderit leaves in the verb, and ipse erit omnia in omnibus for ut sit Deus omnia in omnibus. Declined.
• @1172A · II Cor. XII, 9 — [var:]. Migne prints Sufficit mihi gratia tua, the persons inverted from sufficit tibi gratia mea, because the frame makes the soul say it Domino Jesu. Rendered as printed; the marker records the received reading.

Lemmata 1–3, Quia meliora sunt ubera tua vino to Trahe me post te — 1172B–1178A

• @1172B · lemma · Cant. I, 1–2 — Quia meliora sunt ubera tua vino. 2. Fragrantia unguentis optimis. = Vulgate (fraglans/fragrantia is orthography only). Fixed English: "For thy breasts are better than wine. 2. Fragrant with the best ointments." The verse number 2 sits inside the italic line as Migne prints it and is kept there.
• @1172C, @1174C, @1175C, @1176A — the lemma re-quoted five ways, and the English follows each. 1174C and 1175C drop Quia, and the English drops "For" in exactly those two places and keeps it where Migne keeps it. At 1176A the lemma is split into glossed fragments — « Ubera », « meliora sunt, », « vino, », « fragrantia sunt unguentis optimis, » — the last inserting sunt, which the English carries ("they are fragrant with the best ointments").
• @1172C — dilecta mea is the ordinary participle, rendered "my beloved". The Song's Dilectus as the Bridegroom-noun does not occur in these columns.
• @1172D — [sic: praecontis]. Not a Latin word. Read praeconiis, "by proclamations, by heralding" — ab ipso quasi praeconiis extollitur, and Foliot uses praeconio of exactly this act eight columns later at 1176B. Carried; it swallows no verb, negation or conjunction, so nothing is supplied beside it.
• @1172D — Num ad ubera divertamus…? where the run of the argument wants Nunc ("Now let us turn aside"). Num renders in place as an interrogative particle and the sentence is a question either way, so no marker; the conjecture is recorded here.
• @1173A and @1173C — two etymological glosses keep their Latin beside the English, because the gloss IS the etymology: ubera from ubertas, and serum gathered sero.
• @1173A–1175A — laeva/dextera and sinistra/dextra both run through the argument about the two breasts and the two Testaments. English has one pair of words for two Latin pairs, and the alternation is recorded here rather than invented in the translation. ⚠ See the conventions above: the later ruling at 1221C distinguishes laeva "left hand" from sinistra "on the left", and these columns predate it. Not retro-harmonized, because no gloss here turns on the difference.
• @1173B · Exod. XXI, 23–25 — Reddat quis animam pro anima for the Vulgate's reddet animam pro anima: jussive plus indefinite subject where the Vulgate has a plain future. Rendered "Let a man render…"; the precept is unchanged. Declined.
• @1173D · Matt. XXIII, 13 — hypocritae absent. Declined.
• @1174C — quae Dei sunt printed twice, and it is NOT a [d: …]. ut quae Dei sunt negligat, et ea quae Dei sunt, haec inquirere et investigare contemnat. The phrase is not a compositor's immediate doubling but a resumptive repetition with its own syntax in each clause. Rendered twice, unmarked.
• @1175A · Eccle. VII, 30 — Deus creavit hominem rectum for fecit. A synonym; declined.
• @1175C · Isa. LV, 2 — argentum vestrum added. Declined.
• @1175D · lemma · Cant. I, 2 — Oleum effusum nomen tuum. = Vulgate. Fixed English: "Thy name is oil poured out." Re-quoted at 1176A, 1177B (« oleum effusum » esse) and 1178A.
• @1176B ⭐ — THE HEADLINE INTERNAL DIVERGENCE, and no marker could carry it. Foliot re-quotes his own lemma as « Nomen tuum est ut oleum » — est ut inserted, identity turned into simile — and then builds the entire following gloss (oil floating above every liquid) on the simile rather than on the lemma he printed four lines above. That is exactly the class [var:] exists for. It could not take the marker for a mechanical reason: a [var:]'s content must not appear verbatim in the Latin twin, and the received reading is printed as this chunk's own lemma, so any faithful marker here is a hard verifier error. Rendered as printed. ⚑ See Standing Item 2 — this is the site that earns the rule change, and it is one of the two to put to Wilson.
• @1176A — [sic: soquentes] for sequentes, "those following". Not a Latin word. The carried run is the object of potet et inebriet, so the sense is supplied in plain English beside the marker.
• @1176B — [sic: invensionibus] for inventionibus, "inventions". Not a Latin word; the same word is spelled correctly at 1174C (adinvenit, invenire), and the sentence needs the noun the gloss has been building for a column. Sense supplied beside the marker.
• @1176C · Luc. VI, 19 — Virtus enim de illo exibat, et omnes sanabat (word order). Declined.
• @1176D · Joan. XXI, 25 — Sunt et multa alia for sunt autem et alia multa. Declined.
• @1176D — [sic: saliari] for satiari, "to be satisfied"; the clause is about the spirit's food. It is the infinitive complement of praevaleat, so the English supplies "satisfied" beside the marker — the main predicate survives, but the marked run carried the only content verb.
• @1176D ⚠ — the most conspicuous factual oddity in the opening chapters, and it is NOT marked. Hoc panes duodecim ad novem millia hominum satianda ampliavit. The Gospels have five loaves and five thousand (Matt. XIV) or seven and four thousand (Matt. XV); twelve is the number of baskets of fragments and nine thousand is in no account. Rendered exactly as printed. [cj:] declined because this is Foliot's own prose, not a quoted scripture (so Pattern 14 does not reach it), and because no single printed word can be glossed as the error — both numerals would have to go, which is emendation, not annotation.
• @1177A · Heb. XI, 6 — Sine fide placere Deo impossibile est; the Vulgate's autem gone and the order rearranged. Declined.
• @1177B · lemma · Cant. I, 2 — Ideo adolescentulae dilexerunt te. = Vulgate (adulescentulae/adolescentulae is orthography). "Therefore the maidens have loved thee." Re-quoted at 1177C as « dilexerunt adolescentulae » (verb first) and at 1178A in full; the English keeps the same words.
• @1177C · Luc. IX, 35 — [var:]. dilectus … qui mihi per omnia complacet … hunc audite is a Matthean form standing under a Lucan citation, and electus against dilectus is the reading at issue in that verse's whole textual history.
• @1177D · Act. IX, 5 — Ego sum Jesus Nazarenus, from the parallel at Act. XXII, 8. Nothing in the gloss turns on it; declined. Migne's note reads (Act. IX, 45) for IX, 5 — kept verbatim.
• @1178A · lemma 3 · Cant. I, 3 — Trahe me post te. = Vulgate as far as it goes. ⚑ The Vulgate's curremus is NOT missing: Migne makes it the next lemma at 1180A (Curremus in odore unguentorum tuorum). This is lemma segmentation, not a variant, and no [var:] may be fired on either side of the seam. Trahe me is a singular imperative → "Draw me after thee".
• @1178A — [cj: Si; read Sic, "So"]. Si factus adolescentulus, audivit ab Anania verba quibus excrevit. Si is a perfectly good word, so nothing is defective for [sic:] to wrap, and the faithful English ("If, made a little youth, he heard…") turns Paul's actual instruction by Ananias into a hypothesis — in a sentence whose second half answers it with the correlative sic infinitus credentium numerus. A correlative weakened into a condition, which Pattern 18 already admits.
• @1178A — signa et mirabilia … quibus nec resistere, nec contradicere poterunt: future where the narrative is past (poterant). Renders in place; both negations carried.**

Lemmata 3–6, Curremus to Indica mihi — 1178B–1192C

• @1178B · Ephes. IV, 13 — in virum perfectum, et mensuram aetatis for Vulg. in mensuram. Preposition for conjunction, and not inside guillemets. Declined.
• @1178C · Joan. VI, 44 — [var: Vulg. Nemo potest venire ad me]. Migne prints « Nemo venit ad me, nisi Pater … traxerit eum », dropping the modal. "No man comes" is a statement of fact where the Vulgate states impossibility, and the whole paragraph argues about what a man can do without the Father drawing.
• @1180A · lemma · Cant. I, 3 — [var: Vulg. in odorem]. Migne prints Curremus in odore unguentorum tuorum — ablative for accusative, "in the odour" not "to the odour". Foliot's gloss depends on it: he expounds it instrumentally, id est, odoramento virtutum spiritualium, repeats in odore three times in the gloss and once in his own prose at 1179A. The marker concerns odore/odorem only; the curremus seam itself is segmentation (see 1178A).
• @1179C · Eccle. IV, 12 — « funiculus triplex non facile rumpitur » for Vulg. difficile rumpitur. Same sense by a different construction; Foliot glosses the cord, not the adverb. ⚠ The printed non is carried into the English ("is not easily broken") — the divergence is lexical, not a polarity change.
• @1179C and @1185A — the same verse punctuated two ways on two plates. Rom. VII, 24 reads « Infelix ego homo: quis me liberabit… » at 1179C and « Infelix ego homo! Quis me liberabit… » at 1185A. Both carried exactly as printed; the divergence is Migne's.
• @1179D — [sic: saptiam] for sapiam, a non-word: may inwardly savor [sic: saptiam] thee in my heart. The marked run was the clause's only predicate, so the sense ("savor") stands in plain words beside the marker. The conjecture is near-certain — the parallel members that follow are te et tua lingua denuntiem and opere devotus adimpleam, giving heart / tongue / work.
• @1180C — a guillemet opened and never closed, mirrored. « Trahe me post te; curremus in odore unguentorum tuorum. opens and Migne opens a fresh « at Introduxit me Rex. No » supplied.
• @1180D — ea quibus vult, et quando vult, et quamdiu vult, distribuit. Reads as though a word has dropped, but it construes as printed ("distributes them to whom he wills"). No action.
• @1181A · Bar. III, 24, 25 — [var: Vulg. non habet finem]. Migne prints « … et non habet consummationem, excelsus, et immensus ». The gloss is the etymology: 1181B turns immediately on the word — Sed nunquid cum creatura consummatus est Deus…? An English "end" would leave that question explaining nothing.
• @1181B/C · Joan. XVII, 5 — « … claritate quam habui antequam mundus fieret » for Vulg. priusquam mundus esset, apud te, with the opening Et nunc … tu and the closing apud te dropped. Declined: the argument at 1181B is about God's infinity, not about the mode of the world's coming-to-be.
• @1182A · Matth. XXVI, 52, 53 — An putas quia non possum rogare Patrem meum, et exhibebit mihi plus quam duodecim legiones; the Vulgate adds modo before plus. Declined.
• @1182A — Migne closes a DIRECT question with a full stop where the Vulgate prints ?. The mark rides as printed and the English ends with a period. This is "never supply", and it is not the indirect-question case: the Latin is a direct question and it is Migne's mark that is odd, not English grammar refusing one.
• @1182A–B — sed harum frequentatione laudum potius delectari magis est. Pleonastic potius … magis, and est with a bare infinitive where a gerundive is wanted. Rendered as closely as the printed words allow ("but rather it is the more a delight to be delighted by the frequenting of these praises"), which is clumsy in English because it is clumsy on the plate. No emendation, no marker: the words are all real and all occupy their slots.
• @1182C · Psal. XXXI, 11 — « Laetamini, justi, in Domino, et gloriamini … » for Vulg. Laetamini in Domino et exsultate justi. Migne compresses; an omission is not a divergent reading, so no [var:].
• @1183C · lemma 4 · Cant. I, 4 — Nigra sum, sed formosa, filiae Jerusalem. Agrees with the Vulgate exactly; no marker. Fixed English: "I am black, but beautiful, O ye daughters of Jerusalem."
• @1184A ⭐ — MIGNE RE-QUOTES HIS OWN LEMMA WITH et FOR sed, and it takes the marker. « Nigra sum et formosa, filiae Jerusalem. » — twenty lines below the lemma that prints sed. Marked [var: Vulg. sed, "but"]. Material, not cosmetic: the concessive sed is the whole hinge of the chapter — the exposition at 1184C–1186A rests on blackness being conceded and beauty asserted against it (Attendendum itaque nigredinem hanc plus habere formae in se), and 1186A's diptych is nigra sum → sicut tabernacula Cedar against sed formosa → sicut pelles Salomonis. A copulative et flattens that into a pair of coordinate attributes. Rendered as printed: "I am black and beautiful." ⚑ Nine other nigra and five formosa re-quotations print sed and hold the pair distinct (1184A's bare fragment « Nigra sum, filiae Jerusalem. », 1184C, 1185C, 1186A and the two diptych halves), each checked one by one. So this is a single-site defect in the plate, not a drift in the work. The bare fragment that drops sed formosa entirely is a fragment, not a variant, and takes no marker.
• @1183D · I Reg. XV, 22 — « Melior est obedientia quam victimae » for Vulg. Melior est enim obedientia. Declined.
• @1184B/C — a second unclosed guillemet, mirrored. « Justus prior est accusator sui, (Prov. XVIII, 17) opens and is never closed before the next « at Nigra sum. No » supplied.
• @1184C · Matth. XVI, 21 — DECLINED deliberately. Quia « oportet me ire Jerosolymam … » for Vulg. oportet eum ire, with the closing et tertia die resurgere dropped. Foliot's own frame (Dominus noster Jesus Christus affuturam declarat, dicens, Quia…) converts the evangelist's third person into Christ's direct speech, so the first person is the commentator's construction and not a text-critical fact. Named because it is exactly the shape a blind read would flag.
• @1184C — mundumque sibi, et se mundo contemptibilem parat. The subject of parat is ambiguous between the Holy Spirit and the soul, and sibi between the two. Rendered without resolving it; the ambiguity is Migne's syntax, not a defect.
• @1185A · Act. XII, 11 — « … quia emisit Dominus angelum suum » for Vulg. misit. Compound for simplex, same sense. Declined.
• @1185A · II Cor. XI, 23 — « … in plagis super modum » for Vulg. supra modum. Declined.
• @1185B · Rom. VIII, 38, 39 — [var: Vulg. neque virtutes]. Migne prints « … neque Angeli, neque Principatus, neque Potestates ». Foliot's own next clause (nec caetera quae sequuntur, quae apostolus connumerat) makes the enumeration itself load-bearing, and virtutes is a term he uses throughout these columns for the angelic order.
• @1185B and @1186C · Rom. VIII, 39 — DECLINED, both. « eam separare queant a charitate Dei » and « Scio quod nulla omnino creatura separabit me a charitate Dei » are Foliot's own adaptations inside guillemets, with the object changed to fit his sentence (eam = the Bride; me for the Vulgate's nos). The reference of the pronoun is manifestly the commentator's, not a witness to a divergent text, and Pattern 14 forbids speculating about why readings differ. Rendered exactly as printed.
• @1185C · Gen. VI, 5, 6 — « Videns autem Deus … cuncta cogitatio cordis esset intenta ad malum » for Vulg. intenta esset; word order only. Declined.
• @1186B — sic et hae, ab amore Dei … averti. Feminine plural after a discussion of the masculine tabernacula Cedar; the antecedent must be the perfected souls or churches. Rendered "so neither could these women be turned away"; gender has no English exponent.
• @1186D — a full stop that breaks a relative clause, preserved in place. Summus enim decor … huic affuit; quem. nullo carere quo vita hominum adornari queat, quisquis bene sanus intelligit. The English carries the mark in the same position: "…was present to him; whom. whosoever is well in his wits understands to lack nothing…". It reads as a defect because it is one; never deleted.
• @1187A — « Formosus sum sicut pellis Salomonis; ». The Bride speaks of herself with a masculine Formosus. Latin grammatical gender has no English exponent, so there is nothing to render either way and nothing to mark. Note also that this is Foliot's own construction, not a Song lemma.
• @1187C · lemma · Cant. I, 5 — Filii matris meae pugnaverunt adversum me for Vulg. contra me. Synonym; no sense change. Declined.
• @1187C — [sic: justistitiae] for justitiae, a dittographic syllable: the Sun of justice [sic: justistitiae] himself. Sense supplied beside the marker, and the conjecture is settled by Migne's own next sentence, qui Sol est justitiae, six words later.
• @1187D–1188B · Ose. I, 10 / Rom. IX, 25, 26 — « Vocabo non plebem meam … et non dilectam meam, dilectam » for Vulg. et non dilectam, dilectam, and in hoc loco for in loco. ⚠ dilectam here is the ordinary participle "beloved", not Dilectus the Bridegroom; rendered lowercase throughout. Declined.
• @1188B/C · Act. V, 17, 18 — « Exurgens Princeps Sacerdotum … » for Vulg. Exsurgens autem princeps. Declined.
• @1188C — caecis postmodum denuntiaverunt — [cj:] CONSIDERED AND DECLINED. The dative caecis has no plausible referent: the apostles were not blind, and the subject is the Sanhedrin. Rendered literally, "they afterwards gave charge to the blind, that they should by no means speak…". The bar is that the printed word makes the sentence assert something the author is arguing against, and this produces an odd referent rather than a false claim; it is not a negation and not a correlative. A reader who wants the conjecture — caeci, nominative, of the accusers themselves — has it here. Flagging without acting is deliberate.
• @1189A · lemma 6 · Cant. I, 6 — Indica mihi, quem diligit anima mea, ubi pascas, ubi cubes in meridie, ne vagari incipiam post greges sodalium tuorum. Agrees with the Vulgate exactly. ⛔ This lemma is glossed across a chunk boundary and the two renderings do not agree — see the open items in the conventions above. It is the one live translation collision the merge could not close.
• @1189C · I Cor. VI, 17 — [var: Vulg. Qui autem adhaeret Domino]. Migne prints « Qui conjungitur Domino, unus spiritus est ». The gloss is the word: the very next sentence is Haec igitur Sponsa conjuncta Deo, habens omnes … animas conjunctas Deo. Foliot builds two clauses on conjungi.
• @1190D · Luc. XI, 4, 5 — « Quis vestrum habebit amicum … dicens, Amice … quoniam amicus meus venit ad me » for Vulg. et dicet illi, with de via dropped. Migne's own citation numbers the verses 4, 5 where the passage is Luke 11:5–6; the note passes through verbatim. Declined. ⚑ amicus here is the Gospel's friend, rendered "friend", never "bride" — another count for the amica ruling.
• @1191B · Matth. XIV, 19 — « Cum, acceptis quinque panibus … »; the Vulgate adds et duobus piscibus, and Migne adds suis after discipulis. The omission is worth a reader's eye because Foliot's numerology at 1191B–C counts only loaves (five = Pentateuch, seven = Heptateuch) — but an omission is not a variant reading, so no marker.
• @1191D — « Paulus etiam apostolus. » in guillemets with no citation and no governing verb, and the sentence beginning Pentateuchum vero, id est quinque libros Mosis… has no main clause. Rendered as printed, quotation marks included; nothing supplied.
• @1191D — « in castitate perimat; » likewise in guillemets although it quotes nothing. Carried 1:1.
• @1191D — ut sint haec quasi semina, et multa ex his fruges oriatur — MARKER DECLINED FOR A REASON WORTH KEEPING. Plural fruges against singular oriatur. English "may arise" is number-neutral, which is the case a marker exists for — but the marker must wrap the defective run and it is not determinable which token is defective: fruges for frux, or oriatur for oriantur, or multa pulling either way. Marking an arbitrary guess would misattribute the compositor's error.** Rendered "that many fruits may arise from them" and logged instead.
• @1192A · Psal. CI, 5 — [sic: suum], and this is the dangerous class. Migne prints « Percussus suum ut foenum, et aruit cor meum » for Percussus sum. suum is a perfectly good Latin word, but it has no grammatical slot — the clause is left with no finite verb at all — so it is carried: I am smitten [sic: suum] as hay. The copula is supplied in plain words outside the marker; quarantining suum silently would have left "smitten as hay" with no subject and no verb. Conjecture: Vulg. sum. A second [var:] beside the [sic:] was considered and declined as telling the reader the same thing twice. ⚠ A real word takes no automated flag — this is the shape that must be found by reading.
• @1192B/C · III Reg. XIX, 7, 8 — « Surge, comede; grandis enim restat via. Surrexit ergo, et comedit, ambulavitque… » for Vulg. grandis enim tibi restat via. Qui cum surrexisset, comedit et bibit, et ambulavit. Declined.
• @1180C, @1182C, @1183A, @1183B, @1186A, @1186C, @1187B, @1188D, @1190B, @1190D, @1191A, @1191B, @1192A, @1192B, @1192C and the lemma at 1183C — checked word by word and AGREEING with the Vulgate exactly. Recorded so the check is visible: Introduxit me Rex in cellaria sua, Memores uberum tuorum super vinum, Recti diligunt te, Nigra sum, sed formosa, filiae Jerusalem, Sicut tabernacula Cedar, sicut pelles Salomonis, Nolite me considerare, quod fusca sim, quia decoloravit me sol, Vineam meam non custodivi, Indica mihi…, and Joan. XII, 32 · Gen. XXXII, 26 · Rom. VII, 24–25 · Joan. XIV, 2 · Psal. XXXII, 1 · Psal. XCIX, 3 · Psal. XXXVII, 7 · Job X, 11 · Psal. CXXXI, 15 · Psal. LXXVII, 25 · Psal. CXVIII, 130 · Luc. XXIV, 35 · Psal. XLI, 4 · Psal. CI, 9–10 · Luc. XI, 3. (exultabimus at 1180C and Exultate at 1182C are orthography, not divergences.)

Lemmata 7–13, Si ignoras te to Botrus cypri — 1192D–1206B

Every printed lemma line in these columns agrees with the Vulgate. All the divergence sits in the re-quotations inside the gloss.

• @1192D — « Ubi pascas, et ubi cubes », where the lemma at 1189A and the Vulgate have no et. A connective added in re-quotation; declined.
• @1193A / @1193B · Gen. XVIII, 1 — quoted in full at 1193A and re-quoted at 1193B as apparuit Dominus sedenti in tabernaculo, abridging in convalle Mambre … in ostio tabernaculi sui in ipso fervore diei. Abridgment, not variation; declined.
• @1193B · Gen. XVIII, 4–5 — drops et requiescite sub arbore and the et before ponamque. Abridgment; declined.
• @1193B — cum ipsum in trinitate personam tertiam, et Deum unum in ipsa trinitate intellexit. Rendered literally: Abraham "understood him in the trinity to be the third person". The argument requires only that Abraham saw three persons and one God; tertiam makes the Son the third person, which is not the order Foliot has just laid out. Possibly tertium with personam attracted, possibly a plate slip. No marker: tertiam is a real word occupying its slot and the English asserts what the Latin asserts.
• @1194A · lemma 7 · Cant. I, 7 — Si ignoras te, o pulcherrima inter mulieres, egredere. = Vulgate. "If thou know not thyself, O most beautiful among women, go forth." ("Among women" here is kept distinct from 1281D's genitive pulcherrima mulierum, "of women".)
• @1194B ⭐ — [var: Vulg. pulcherrima, "most beautiful"]: the lemma's superlative is lost in EVERY re-quotation. All four glosses print pulchra — « Si ignoras te, o pulchra inter mulieres, egredere » · « O pulchra inter mulieres! » · « O pulchra inter mulieres, si ignoras te » (1194B–C) · « Si ignoras te, o pulchra, egredere » (1195D). The English follows: "most beautiful" in the lemma, "beautiful one" in all four glosses, with one marker at the first gloss occurrence. Not harmonized in either direction. ⚑ This is the site to put to Wilson on the "Migne against himself" question — it is the one that would EARN an internal marker on materiality, because the superlative is lost and the gloss is built on the weakened form. It currently ships a [var:] only because the gloss's pulchra also diverges from the Vulgate externally, which is a genuinely different fact. ⚠ The marker's content was tightened from [var: Vulg. and the lemma above pulcherrima, …] to the form above: it validated either way, but it carried an internal comparison inside its content, and no shipped [var:] in these columns now leans on one.
• @1194C — quarum comparatione decora. Foliot's own decora, rendered "comely" — the precedent inside this work that later fixes decora = comely at 1283B and 1292D.
• @1195A — I Cor. II, 9 woven in with no guillemets and no note. Migne marks it as nothing at all, so nothing is supplied.
• @1195B — « haedos suos » where the lemma has haedos tuos (second person); the gloss converts to the third person for the exposition. Foliot's own recasting; declined.
• @1195C · lemma 8 · Cant. I, 8 — Equitatui meo in curribus Pharaonis assimilavi te, amica mea. = Vulgate. "To my company of horse in the chariots of Pharaoh have I likened thee, my friend." equitatus = "company of horse".
• @1195C — [sic: eonsilium] for consilium, the c broken to e. Carried; the marker is a noun object, and quarantining it alone would leave "the Bridegroom gives to the Bride who was lifted up ⟨hole⟩", so the sense stands beside it in plain words — "gives counsel [sic: eonsilium]". Near-certain: the same sentence goes on si boni pertaesa consilii, spelled correctly.
• @1195D — moderate responderit, perfect subjunctive where the narrative wants the indicative, among a run of indicatives. English has no clean exponent for the mood ("he may have answered" would assert something the passage does not), so the rendering is "answered with moderation". A mood anomaly that merely fails to construe stays declined.**
• @1196A ⭐ — the amica gloss, and the reason the whole work reads "friend". illam verbo amicabili convenit … dicens, « Amica mea » — "he accosts her with a word of friendship … saying, « My friend »". With amicus Caesaris at 1202C in the same columns and Foliot's own pun at 1231C, this is the evidence that fixed the term.
• @1196A — sic innuens, sibi jam in te gratum esse totum quod erat a gratia imminutum. A third-person sentence about the Bride slips into the second person mid-clause. Rendered as printed ("is now pleasing to him in thee"); it is a live possibility that Foliot is quoting the Bridegroom's own words without marking them, which is why nothing is supplied.
• @1196B · Isa. XIV, 14 — [var:]. Migne prints Ponam sedem meam ad Aquilonem, et ero similis Altissimo, compressing vv. 13–14 and replacing exaltabo solium meum / in lateribus aquilonis. Rendered as printed; the marker records the received reading, and no speculation is offered as to cause. ⚠ Compare 1211D and 1217C, where the same boast is printed the same way and the marker is declined — see those entries for why the two treatments differ.
• @1196C — ut hinc egressio, aut electos probent, aut … praecipitent. egressio is a nominative singular standing against two plural subjunctives whose subject must be the demons. Rendered so that the disagreement survives in English ("so that hence a going forth, they may either prove the elect, or hurl headlong…") rather than smoothed to a singular verb, which would repair it invisibly. No marker: the type is not defective, the syntax is.
• @1196D–1197A — the Pharaoh sentence. …saevientem Pharaonem plagis durissimis exercens Dominus … immisso equitatu coelesti, vindictam Domino exercuit. Dominus is nominative with a participle and then the finite verb takes Domino as its dative, so the subject of exercuit has to be the equitatus. Rendered that way ("the heavenly company of horse being sent in, it executed vengeance for the Lord"), which is the only reading the printed cases allow; logged because the sentence reads at first as though the Lord were the subject throughout.
• @1197B · Matt. XXII, 30 — [var: Vulg. erunt]. Migne prints sed sunt sicut angeli Dei in coelo. Present for future, inside an argument about what the Bride is being made like now in predestination (jam similem te feci, 1197D), so the tense is not idle.
• @1197B — the note (IV Reg. XIX, 39). 4 Kings XIX has 37 verses; the Angel and the 185,000 are at verse 35. Reproduced verbatim — notes are his — and recorded for the index.
• @1197C · Matth. XXVI, 52, 53 — [var:]: the sword-word is John's, not Matthew's. Migne prints « Mitte gladium in vaginam. An putas, quod non possum rogare Patrem meum… » under a note reading (Matth. XXVI, 52, 53.). Matthew has converte gladium tuum in locum suum; mitte gladium tuum in vaginam is Joan. XVIII, 11. The note is reproduced verbatim; the marker records that the wording printed is John's. (quod for Matthew's quia, and modo dropped before plus quam, are immaterial and take nothing.) ⚑ Converte gladium tuum is singular to Peter, so "thy".
• @1197C — Qui quidem numerus octoginta millia continet, octo minus. Rendered literally: "Which number indeed contains eighty thousand, eight less." Twelve legions at 6,000 to the legion is 72,000, which is eighty thousand less eight thousand — so octo is elliptical for octo millia rather than wrong. No emendation, no marker; the arithmetic is recorded here so a reader is not left thinking 79,992.
• @1197D · lemma 9 · Cant. I, 9 — Pulchrae sunt genae tuae, sicut turturis: collum tuum sicut monilia. = Vulgate. monile = "necklace" here (contrast 1289B, where the same word is kept in italic Latin because Foliot etymologizes it).
• @1198B · Phil. II, 6 — omits se (esse aequalem Deo). Immaterial; declined.
• @1198B · Act. IV, 12 — the nearest of the declines to the bar. Recast into indirect discourse (Nullum aliud nomen sub coelo esse, in quo salvos fieri oporteat homines) inside guillemets, against nec enim aliud nomen est sub caelo datum hominibus…. Structural recasting to fit asseverant et adstruunt; sense intact, datum hominibus lost. It takes no marker only because nothing in the gloss turns on the dropped words.
• @1198C · Psal. I — opens Beatus cujus in lege Domini voluntas ejus, compressing Beatus vir qui non abiit … sed in lege Domini. The note (Psal. I, 1, 2, 3) shows Migne knew verse 1 was in play. Compression; declined. (Contrast 1302D, where the same psalm loses Beatus and does take a marker, because there the argument runs on the word.)
• @1198D — Collum enim a colligando dicitur: an etymology the English cannot carry. "Neck" shares no root with "bind", so both Latin words stand in italic inside the rendering: "For collum, the neck, is so called from colligando, binding together".
• @1199A — Quam custodiendam suscepi, casta sis…. The necklace's speech opens with an accusative relative and continues with jussive subjunctives addressed to a second person; there is no antecedent and no main verb. Rendered keeping the case visible ("Whom I have undertaken to guard, be thou chaste") rather than converting to a nominative address, which would tidy away the anacoluthon.
• @1199B · lemma 10 · Cant. I, 10 — Murenulas aureas faciemus tibi. = Vulgate (Migne prints only the first half of the verse). murenulae kept in italic Latin — the gloss IS the etymology, murena the fish — which is the precedent later applied to monilia at 1289B.
• @1199C — unde auditorum conviventia suscipiuntur. conviventia (from convivere) is attested in medieval Latin for fellowship or living-together, so it is not a defect; rendered "into the fellowship of their hearers". Recorded because convenientia and conniventia are both a single letter away and either would give a smoother sentence.
• @1199D · lemma 11 · Cant. I, 11 — Dum esset Rex in accubitu suo, nardus mea dedit odorem suum. = Vulgate (Migne capitalizes Rex). accubitus = "repose", nardus = "spikenard".
• @1199D ⭐–1200C — THE FOUR-WAY RE-QUOTATION, and the marker that was WITHDRAWN. Migne re-quotes his own lemma four further ways, and the first three change the word:

Each is rendered as printed and none is harmonized to any other; a reader meets all four forms differing in the English itself. A draft [var:] at 1200A read [var: the lemma above and Vulg. Dum, "While"] and was rejected by the validator for opening on an internal comparison instead of a named external witness. It was WITHDRAWN rather than re-opened with Vulg., and the second reason is the one that matters: (1) the comparison is internal — the Vulgate agrees with the lemma's Dum, so what the marker recorded was Migne diverging from himself; (2) re-opening it with Vulg. would have made it validate while asserting a materiality it does not have — dum and cum with the imperfect subjunctive are near-synonymous here, so the divergence fails Pattern 14's own bar. ⛔ A marker that validates is not thereby earned.

• @1199D — quae totum legem utramque statuit. totum against feminine legem utramque. Gender has no English exponent; "the whole of both laws", nothing emended.
• @1200B · Joan. I, 17 — [var: Vulg. facta est]. Migne prints per Christum, gratia et veritas donata est (and drops Jesum). Given against made: the whole paragraph is about what the Father disposed and grace completed, so "given" is doing work.
• @1200B — ab Eva usque in Adam: the order reversed against the history. The delay runs from Adam to the fullness of time, and the same sentence goes on ob Adae offensam. Rendered as printed ("from Eve unto Adam"). Declined for a marker: no word is broken, the English asserts exactly what the Latin asserts, so there is nothing for [sic:] to wrap, and Pattern 18's bar is not clearly met — the sentence is odd, not contradictory. ⚠ Flagged as the site in these columns most likely to be judged differently.**
• @1200C — dum ille, quem regem super omnia judicio, Filius…. judicio is a real word (ablative of judicium) but leaves the relative clause with no verb; the obvious reading is judico, "whom I judge king over all", which would also match the first-person voice of the whole speech. Rendered as closely as the printed words allow, leaving the clause verbless ("while he whom king above all things by judgment, the Son, …"). Not glossed, because the defect is a stranded case, not a negation or a correlative.
• @1200D · Gen. III, 14 — Quia fecisti haec for Vulg. hoc. Number only; declined.
• @1201A–B — [cj: si; read sic, "so"]. …ut nullus eam tentandi … serpenti locus extiterit: si totum quod movere mortalium corda consuevit … statim virtutis calcaneo contritum extiterit. si is a real word, so there is nothing to wrap; but the conditional it opens has no apodosis anywhere, and the clause is plainly the second member of the ita … sic correlative begun in the previous clause. A correlative weakened into a condition. Additive: Migne's si keeps its English ("if"), the conjecture stands beside it.
• @1201B — nec ejus sint suggestionibus aut deflecti, aut certe omnino submitti. sint with two passive infinitives will not construe (sinit? sinat? a lost fas?). Rendered as closely as the words allow, negation carried: "nor are they to be turned aside by his suggestions, or indeed to be wholly made subject."
• @1201D — Hic enim Dei Patris, instar Noe, obdormisse dicitur. A genitive where the sentence needs a nominative Deus Pater. English nouns carry no case, so there is no exponent; rendered "here God the Father … is said to have fallen asleep", logged, not emended.
• @1202A — in quam pater, vibrato jam gladio, collo filius extenso consensit. The ablative absolute wants collo filii extenso; Migne prints the nominative filius. Rendering it as a genitive would be a silent repair, so the English keeps it stranded: "to which the father consented, the sword being now brandished, the neck stretched out, the son." The page reads oddly on purpose.
• @1202C · Joan. XIX, 12 — Si dimittis hunc for Si hunc dimittis. Word order; declined. (amicus Caesaris stands here — see the amica ruling.)
• @1202D ⭐ · Gen. XLIX, 11 — [var: Vulg. pullum suum], and it is a PLATE-READ CANDIDATE. Migne prints Ligans ad vineam pallium suum. The most consequential divergence in these columns: the cloak is tied to the vineyard and the colt vanishes — and pallium has just been glossed two lines above, in its right place, as vestem exteriorem from the second half of the same verse, so the printed reading makes the verse say pallium twice. Foliot's gloss (ad quem asina alligata est) covers only the she-ass and never the colt, so the work gives no independent witness either way. Rendered as printed ("Tying his cloak to the vineyard"); marker gives pullum suum. See Standing Item 3.
• @1203B · lemma 12 · Cant. I, 12 — Fasciculus myrrhae dilectus meus mihi, inter ubera mea commorabitur. = Vulgate; dilectus lowercase here, rendered "my beloved to me". fasciculus = "bundle".
• @1203B–1206B — dilectus / Dilectus, occurrence by occurrence, mirroring the plate. 1203B Dilecto suo "her Beloved" · 1203B lemma dilectus meus mihi "my beloved to me" · 1203C Dilectus meus "My Beloved" · 1203D dilecti sui "her beloved" · 1204B « Dilectus meus mihi » "My Beloved to me" · 1204C Dilectus "the Beloved" · 1205A Dilectus suus ×2 "her Beloved" · 1205A the full Song quotation, capital · 1205B lemma 13, capital · 1205C « mihi dilectus meus commorabitur » lowercase, and with the word order changed from the lemma · 1205C « Botrus cypri dilectus meus mihi. » lowercase, against the capitalized lemma eight lines above · 1205D and 1206B capital again. Rendered as printed at every one; not harmonized. Foliot's own participial dilecta (1194B, 1195A) is the ordinary word, "beloved", without promotion.
• @1203C · Joan. XIV, 23 — [var: Vulg. diliget]. Migne prints et Pater meus diligit eum — present for future again, in a sentence whose other verbs (veniemus, faciemus) stay future.
• @1203D · Apoc. III, 20 — drops mihi after aperuerit, and prints introibo for intrabo, ille for ipse. Immaterial; declined.
• @1205A — the note (Joan. XIX, 39) set INSIDE the Song quotation. Migne puts the citation for the myrrh-and-aloes of the preceding sentence inside the guillemets of « Dilectus meus mihi inter ubera mea commorabitur », where it documents nothing. Reproduced verbatim and in place; recorded so the index does not bind John XIX to Canticles I.
• @1205B · lemma 13 · Cant. I, 13 — Botrus cypri Dilectus meus mihi. = Vulgate (Migne capitalizes Dilectus and splits off in vineis Engaddi as a separate lemma at 1206B). botrus = "cluster".
• @1205B — an unclosed quotation, and it is Migne's. Hic enim « myrrhae fasciculus, multa passionis Christi memoria est; opens a guillemet that is never closed, and the quotation runs on into what is plainly Foliot's own second-person exhortation (quam si teneas…). The English mirrors the plate and supplies no ».
• @1205C — Cyprus enim, insula vitibus abundans est, and why the lemma reads "Cyprus". Foliot reads cypri as the island, not the henna or cypress shrub, and builds the whole gloss on its vines and clusters. The lemma is therefore rendered "A cluster of Cyprus" so that the gloss explains it; the conventional "cypress" would leave the island paragraph attached to nothing. Recorded rather than marked — this is a translator's reading of cypri, not a divergence in Migne's text.
• @1205D — a comma where the sentence ends, preserved. …quod faciem hominis exhilarat, Ait itaque Sponsa haec… — a comma followed by a capital A. Preserved exactly: never delete a mark he prints, never supply one he does not. The English carries the same comma splice.
• @1206A · Phil. II, 8 — [var:]. Migne prints Christus factus est pro nobis obediens usque ad mortem; the Vulgate has humiliavit semetipsum, factus obediens usque ad mortem. pro nobis is not in the verse and humiliavit semetipsum is gone.
• @1206A · Phil. II, 9 — exaltavit eum for illum and dedit for donavit. Immaterial; declined.

Lemmata 14–16, Ecce tu pulchra es to laquearia nostra cupressina — 1206C–1211D

• @1206C · Luc. XXIII, 34 — [var: Vulg. dimitte illis: non enim sciunt]. Migne prints Pater, ignosce illis, nesciunt enim quid faciunt.
• @1206D · lemma 14 · Cant. I, 14 — Ecce tu pulchra es, amica mea, ecce tu pulchra. = Vulgate. Rendered with amica = "my friend" per the ruling. "Thine eyes are those of doves" is the fixed rendering of the bare genitive here, and the same rendering is held at 1255C.
• @1206D ⭐ and @1209B — the gender English cannot carry, and what carries it instead. The two lemmata stand consecutively and are grammatically opposed: Ecce tu pulchra es, amica mea (feminine, the Bride) and Ecce tu pulcher es, Dilecte mi, et decorus (masculine, the Bridegroom). English adjectives carry no gender, so the referent is carried by the vocative Migne prints and by nothing else — amica mea can only be the Bride, Dilecte mi can only be the Bridegroom. No word was invented to disambiguate. Foliot's own gloss reinforces it independently at 1209C: Sic amans amanti pari respondet elogio.
• @1207A ⭐–1210C — "beautiful", not "fair", and this is where the rule is earned. pulchritudo is glossed continuously for four columns (what beauty is, whose beauty, the beauty of the Trinity), so a lemma reading "fair" would leave the gloss explaining a word not in it. The familiar Douay cadence is exactly the pressure that had to be resisted here.
• @1207A — a full stop mid-sentence, preserved. …triplici jam magnificavit elogio. dicens, — a period followed by a lowercase participle. The English reads "…with a threefold commendation. saying,". Almost certainly a comma in the compositor's copy; the same shape is preserved at 1208C (…quae poterit esse vexatio? qui nec praesens aderit) and 1213B (« Ego flos campi. » qui etiam mundo…).
• @1207B · Isa. II, 12 + 16 — DECLINED, and the page adjudicates it. Dies Domini exercituum super omnem superbum et excelsum, et super omne quod visu pulchrum est conflates v. 12 and v. 16 — and Migne's own note says (Isa. II, 12, 16), so the divergence is already adjudicated on the page.
• @1207B · Isa. III, 18–19 — clause order rearranged, et lunulas → lunulas. Word order and a dropped connective; declined.
• @1207C · Psal. XLIV, 14 — circumamicta varietate for Vulg. varietatibus. Number only, and the gloss (virtutum omnium varia diversitate) works either way. Declined.
• @1207D · Luc. XV, 21 — Pater, non sum dignus … fac me sicut unum ex mercenariis tuis conflates v. 21 with v. 19 and reads ex for de. A composite of one speech, adduced for the shame of the prodigal; no clause asserts anything the Vulgate does not. Declined.
• @1208A — Sic sic pulchram nominat is NOT a dittography. Emphatic sic sic is idiomatic Latin, and the doubling is rhetorical, matching the doubled lemma it glosses (ecce tu pulchra … ecce tu pulchra). Rendered "So, so does he name her beautiful". Declined deliberately.
• @1208D — a clause with no verb, rendered gap and all. quae irrationabiliter non irascitur, nec dicto, nec facto quemquam sed quantumcunque offensa… — quemquam is an accusative with nothing to govern it; a verb of harming (laedit, matching nec rostro laedit eight words earlier) has dropped out. Rendered "…who is not angered unreasonably, nor anyone by word, nor by deed, but however greatly offended…". No verb supplied; no [sic:], because nothing is broken — something is ABSENT, and there is no run to carry. Not [cj:] either: that opens on a word Migne DID print, and here he printed none.
• @1209A — [sic: animi]. Sitivit animi mea ad Deum. A real form with no grammatical slot: mea is feminine singular and sitivit singular, so nothing in the clause can govern it. Almost certainly anima. The verb survives outside the quarantine and the clause still predicates, so no sense had to be supplied beside it: "My [sic: animi] hath thirsted for God…".
• @1209A · Psal. XLI, 3 — [var: Vulg. fortem]. Migne prints ad Deum fontem vivum — "fountain" for "strong". ⚠ Compare 1272A, where the same verse drops fortem altogether and the marker is declined; the two entries state their own grounds and should be read together.
• @1209A–B · Gen. III, 6–7 — the nearest decline to the bar in these columns. Word order transposed and qui comedit dropped after deditque viro suo, so the plate's ellipsis has the eyes of both opened without the man's eating being stated. Declined because the omission reads as citational abridgment (Foliot's argument is about Eve's eye, not Adam's eating) rather than a divergent reading. Flagged here so a blind read can reverse the call.
• @1209B · lemma 15 · Cant. I, 15 — Ecce tu pulcher es, Dilecte mi, et decorus. = Vulgate; capital Dilecte mirrored, rendered "my Beloved".
• @1210A · Psal. CI, 27 — [var: Vulg. omnes sicut]. et haec ut vestimentum veterascent.
• @1210B · I Cor. II, 9 — [var: Vulg. quae praeparavit]. quam manifestabit Deus — "will manifest" is not "hath prepared".
• @1210B · Joan. XVII, 3 — [var: Vulg. solum Deum verum]. ut cognoscant te verum Deum — solum dropped. ⚠ Compare 1255B, where the same verse loses solum and prints videant for cognoscant, and takes its own marker on the verb.
• @1210B ⭐ — where species = "fairness" is fixed. Haec forma, haec species, haec pulchritudo → "This is the form, this the fairness, this the beauty." This precedent is what later frees "comely" for decora and gives speciosa "fair" — see the conventions.
• @1210B · I Pet. I, 12 — in quam desiderant et angeli prospicere, case and connective adapted to Foliot's own relative clause. Weaving, not divergence; declined.
• @1210D–1211A — lectulus "little bed" against lectus "bed", both used within one paragraph and kept distinct. floridus = "flowery", because the gloss turns on flower-not-yet-fruit.
• @1211C · Matth. XVIII, 10 — [var: Vulg. Videte ne contemnatis unum ex his pusillis]. Migne prints Nolite scandalizare pusillos hos. (The plural vobis inside this quotation takes "you"; the singular reader eight lines away takes "thou".)
• @1211D · lemma 16 · Cant. I, 16 — laquearia nostra cupressina; the re-quotation eleven lines later prints laquearia cupressina, without nostra. Both rendered as printed ("our panelled ceilings of cypress" / "the panelled ceilings of cypress"). The difference is Migne's, not ours.
• @1211D — an ambiguous purpose clause, rendered with the ambiguity intact. …malignos spiritus enecat, ne bonorum corda perturbent, ne a bonis initiis ad majora consurgant. The grammatical subject of the second ne clause is still spiritus, which makes the devils the ones rising from good beginnings; the sense wants the good to be prevented from rising. Rendered literally, keeping both negatives: "lest they disturb the hearts of the good, lest they rise from good beginnings to greater things." No negative was added or removed to resolve it.
• @1211D and @1217C · Isa. XIV, 14 — DECLINED, twice, and the consistency is the point. Ponam sedem meam ad aquilonem, et ero similis Altissimo stands twice against Vulg. sedebo in monte testamenti, in lateribus aquilonis … ero similis Altissimo (vv. 13–14). This is the standing patristic form of the boast, the asserted content is identical, and no part of Foliot's gloss rests on ponam sedem. Recorded because it recurs, and because a marker on one occurrence and not the other would be incoherent. (At 1196B the same reading DID take a marker; the two calls sit in different arguments, and both grounds are stated where they fall.)

Caput II — 1212A–1243B

• @1212B — seraphin for seraphim, three columns after Seraphim at 1210A. Attested medieval spelling; normalized to "seraphim" in the English, preserved in the facing Latin. No marker.
• @1212D · II Cor. II, 15 — [var: Vulg. Christi]. Migne prints quia Christo bonus odor sumus Deo — a good odour TO Christ is not the good odour OF Christ.
• @1213A · Joan. VI, 71–72 — Filius Dei vivi, an added word (harmonizing to Matt. XVI, 16) and a well-attested Vulgate variant. Declined.
• @1213A · Jac. IV, 6 — humilibus vero for autem. Declined.
• @1213A · Psal. LXIV, 14 — the guillemet opens BEFORE quia, so Foliot's own connective sits inside the quotation. Rendered inside it ("that the valleys shall abound…").
• @1213C — sponsam lowercase, where Migne capitalizes Sponsa everywhere else in these columns. Rendered lowercase ("the bride"), mirroring the plate, on the same principle as Dilectus/dilectus.
• @1213C — 29 open / 28 close guillemets, and the imbalance is Migne's. « In propria venit, et sui eum non receperunt [n: (Joan. I, 11)]: quos « suos, » non affectu… — the Joan. I, 11 quotation is opened and never closed; the inner « suos, » closes normally. Mirrored exactly. Do not supply the missing ».
• @1213D ⭐ — sive pomum dicatur, sive malum: two Latin names, one English word. Foliot distinguishes two names for the same fruit, and an English rendering would make the sentence say the fruit is called an apple or an apple. Both words are kept in italic Latin. No marker: this is an editorial retention, not carried defective type.
• @1213D–1215D · lemma · Cant. II, 3 — Sicut malus inter ligna sylvarum, and its five re-quotations. The apple tree (feminine malus; Malus enim arbor est, Malus enim ista … arbor est at 1215D), with fructum malorum at 1213D and istius mali fructus at 1215C both apples. ligna sylvarum = "the trees of the woods", fixed across 1213D–1214D. ⛔ The two senses of mala come within four lines of each other at 1215C–1216A (istius mali fructus … omni malitia) and again at 1220A–B; each was decided from case, gender and governing verb, not from context.
• @1214D · Isai. XXV, 1 — [var: Vulg. cogitationes antiquas fideles]. Migne prints Consilium tuum, Domine, antiquum fiat verum. ⚑ Here he prints the locator, which is precisely why this takes a marker where the same reading at 1160C could not: there he printed none, and naming a received reading would have rested on our identification.
• @1215A · Psal. XLIV, 7 — [var:]. Sedes tua, Domine for Deus — the verse is adduced for the enthroned Christ. ⚠ The marker's content was rewritten from the single word to a phrase, because a one-word content occurring verbatim in the Latin twin trips the mirror test. The reading is unchanged; only the marker's prose is.
• @1215B · III Reg. XXII, 19 — [var: Vulg. super solium suum]. super solium excelsum — and excelsum is Isai. VI, 1's word, which Migne quotes three lines later.
• @1215B · Matth. XXV, 31–32 — condensed from two verses to one clause. Declined.
• @1215C — ut mihi Sponsa promissa compleret. Sponsa is nominative (or ablative) where the sense wants a dative agreeing with mihi. Rendered "that he might fulfil to me, being his Bride, the things promised", reading Sponsa as a loose nominative in apposition to the speaking Church. No marker: a case slip has no English exponent and the reading is recoverable.
• @1216A — O beatum guttur, which is why guttur = "throat" and never "palate." The gloss apostrophizes the word; the same rendering is held at 1281C and 1293D.
• @1216B · lemma · Cant. II, 4 — [var:]. Migne prints Introduxit me rex in cellam vinariam; the Vulgate has no rex. ⚠ Marker content likewise rewritten from one word to a phrase, reading unchanged. cella vinaria = "wine cellar".
• @1216B · Deut. XXXII, 32–33 — botri amarissimi (pl.) for botrus amarissimus. Declined.
• @1216C — a dangling genitive, rendered with the gap visible. Hic enim sensus est; Reproborum, qui sanctas Scripturas prave intelligendo… — Reproborum depends on a noun (vinum) that is not printed. Rendered "For this is the sense; of the reprobate, who…". The word was not supplied.
• @1216C — praeduces. …in inferno praeduces ipsi mansiones habere compellunt. An unusual form, taken as praedux used adjectivally with mansiones, "foremost dwellings". Not marked: it construes.
• @1216D · Psal. V, 11 — [var: Vulg. judica illos]. judica illo Deus — the accusative object is lost; rendered "judge thereby, O God".
• @1216D–1217A · Prov. IX, 1 — two different verbs for one lemma, IN MIGNE. The quotation reads Sapientia aedificavit sibi domum, and Foliot's own gloss six lines later re-quotes it as sibi « domum construxit, ». Both rendered as printed ("hath built" / "constructed"). No marker: the second is Foliot's paraphrase inside his own sentence, not a scripture citation.
• @1217A — a quotation left open across a whole gloss, mirrored. « Misit ancillas suas, ut vocarent ad arcem, et ad moenia civitatis quia, ut ait apostolus, Non multi sapientes… — the Prov. IX, 3 quotation is not closed before Foliot's gloss and the I Cor. I citation begin; the whole run closes once, at destrueret. ». Reproduced, so the English also carries a gloss inside the quotation marks.
• @1217B — an inline italic citation inside a quotation, with a column anchor inside it. « Ego sum vitis, vos palmites (Joan. XV, [1217B] 5). » — reproduced verbatim, untranslated, anchor in place, on the same principle as [n: …]. No [f: …] tag: this is a lemma-and-gloss commentary, not a florilegium. It is the only inline citation in these columns. (vos palmites is plural — "you" — eight lines from a singular "thou"; the distinction is deliberate.)
• @1217C — 25 open / 24 close, and a » printed after the note. « Deus » enim « charitas est [n: (Joan. I, 4, 8)], omnem procul dubio diligens creaturam… — the second quotation opens and runs on unclosed into Foliot's own clause; and …et ero similis Altissimo) [n: (Isai. XIV, 14)] » closes after the note and after a parenthesis that opened outside the quotation. Reproduced in that order.
• @1218A · III Reg. XXII, 20–22 — the Lord's In quo? and the spirit's Ego decipiam illum are dropped; and Migne closes cadat in Ramoth Galaad with a period where the Vulgate has ?. The mark is NOT supplied. Declined.
• @1218A — non horum approbans malignitate. malignitate (ablative) where approbans wants the accusative; the parallel quod ab ipso est approbans follows immediately. No marker, no repair: English nouns have no case exponent, so "their malignity" is the only possible rendering either way. Logged so a later sweep does not churn it.
• @1218B · Exod. XXIII, 22 — [var: Vulg. affligam affligentes te]. et adversabor adversariis tuis.
• @1218B · Sap. XI, 25 — word order, and Domine added. Declined.
• @1219A · Matth. XXII, 37 — recast by Foliot into the first plural (Diligamus) inside guillemets; his adaptation, rendered as printed. Declined.
• @1219B · Luc. X, 31 — [var: Vulg. descenderet]. ut sacerdos quidam ascenderet eadem via — the Jericho road descends.
• @1219D · Prov. VI, 16 — quae odit Deus for Dominus. Declined.
• @1220B · Exod. XX, 12 — truncated at super terram; matrem without tuam. Declined.
• @1220B · lemma 5 · Cant. II, 5 — stipate me malis = apples, and the lemma is left un-glossed by the plate: Foliot's exposition of flores (the saints' words) against mala (their works) begins in the next columns. The two must stay distinct there — flores = "flowers", mala = "apples", held through 1220C–1221A (quae mala quaerenda sunt, « Stipate me malis »).
• @1220D — nec contepescat. Not classical, but a transparent compound of tepescere; translated normally ("grow lukewarm"). Recorded because the form is rare enough to be mistaken for broken type.

Lemmata 6–14, Laeva ejus sub capite meo to Surge, amica mea — 1220C–1238A

• @1221C ⭐ · lemma 6 · Cant. II, 6 — laeva against sinistra, and the ruling for the whole work. Migne prints two different Latin words for the left within four lines, and Foliot's argument stands on the pairing: the Song lemma Laeva ejus sub capite meo, et dextera ejus amplexabitur me, and the Proverbs proof-text Longitudo vitae in dextera ejus, in sinistra vero divitiae et gloria. Ruled: laeva → "left hand" (a hand, matching dextera "right hand", which is what the gloss argues — Dominum laevam pariter habere et dexteram); sinistra → "on the left", the bare substantivized adjective, no "hand" supplied, because Migne prints no manus and Proverbs is describing what Wisdom holds on each side. dextera is "right hand" in both, since Migne prints the same word in both. The English keeps the two Latin words visibly distinct without inventing a distinction of sense, and without flattening both into "left hand." ⚠ The alternative — laeva "left" / sinistra "left hand" — reads worse against the gloss; flagged for a second opinion.
• @1221C · lemma 6 — Migne prints ejus for the Vulgate's illius in the second member. No sense change (both are simply "his" in English), no [var:]; recorded because the whole gloss turns on the two hands and the pronoun is part of the lemma chain. ⚠ At 1297D the same verse is printed with dextera illius, and that entry explains why the English does not differ there either.
• @1221C–1222D — the lemma held through six re-quotations. « Laeva ejus sub capite meo, et dextera ejus amplexabitur me: » (verbatim) · the paraphrase quod laeva ejus sub capite jam meo est … quod ejus me dextera amplexabitur · Laevam itaque sub capite fore intelligit · recte laeva Dei sub capite Sponsae · laevam Dei sub « ejus, » id est, Ecclesiae capite · « Et dextera ejus amplexabitur me. » and Quid enim Dei dextera. Every one renders laeva "left hand" and dextera "right hand."
• @1221C · Prov. III, 16 — [var: Vulg. longitudo dierum, "length of days"]. Migne prints longitudo vitae and in sinistra vero for in sinistra illius; the marker takes the substantive divergence only, and the connective difference is recorded here.
• @1221C — amicos of the Bride's friends, rendered "friends". One of the two independent sites that fixed amica = friend.
• @1221D — inofficiosum membrum rendered "a member without office". The legal sense (inofficiosus, "undutiful") is possible, but the context is the body's service to the head. Flagged, not marked.
• @1223A · lemma 7 · Cant. II, 7 — Adjuro vos, filiae Jerusalem, per capreas, cervosque camporum, ne suscitetis, neque evigilare faciatis dilectam, quoadusque ipsa velit. Two differences from the Vulgate, neither marked: quoadusque for donec (same force, "until" in both, no English exponent), and the comma Migne sets after capreas in the lemma and drops in every re-quotation — punctuation rides as printed, so the lemma reads "by the roes, and the harts of the fields," and the re-quotations "by the roes and the harts." Negation fidelity: the lemma prints ne … neque and the English prints exactly two, in the lemma and in every re-quotation. No third negative introduced, none dropped.
• @1223A ⭐ ff. — THE ADJURO VOS VERSE IS PRINTED SIX DIFFERENT WAYS ACROSS THE WORK, and none of them takes a marker. Collected here because they are scattered over four chapters:

Declined at every site, on two grounds that hold for all of them: these are the work quoting itself, not diverging from an outside authority; and English has one word ("until") for quoad / quoadusque / quousque / donec, so the marker would have nothing to show a reader. The one difference English CAN carry — Migne's capitalization of Dilectam — IS carried, at 1232B and 1297D against the lowercase elsewhere.

• @1223B–C and @1228A — caprea / capreolus / cervus / hinnulus cervorum = roe / roebuck / hart / young hart, held distinct.
• @1224B — ab haec vocatione. ab + accusative. No English exponent for the case error; rendered "from this calling", no marker, logged so a later sweep does not churn it.
• @1224B — the broken parenthesis, reproduced in Migne's order. (de qua dictum est, « Generationem ejus quis enarrabit?) [n: (Isai. LIII, 8)], » — the parenthesis closes inside the quotation and the guillemet closes after the note. Reproduced exactly, in that order.
• @1225A · lemma 8 · Cant. II, 8 — [var: Vulg. iste, "this one"]. Migne's lemma prints Ecce ille venit saliens in montibus — and he prints iste himself at 1227C, and drops the demonstrative altogether at 1226B. Three printed forms of one clause in two columns, kept distinguishable in English: "Behold, that one cometh" (lemma, 1225B) · "Behold, he cometh" (1226B) · "Behold, this one cometh" (1227C). The lemma also begins after Vox dilecti mei, which Migne glosses separately — lemma segmentation, not a divergence.
• @1225A — [sic: moi]. Vox dilecti moi haec est. Not a Latin word; the conjecture is mei, near-certain from the Vulgate and from the lemma the sentence introduces. Consequence recorded: with the possessive quarantined, the English reads "the voice of the beloved," not "of my beloved." Nothing that moi governs is left un-rendered, because moi is itself the whole of what is lost.
• @1225B · II Reg. I, 21 — [var:]. Gelboe, nec ros, nec pluvia super vos descendat for Vulg. Montes Gelboe, nec ros nec pluvia veniant super vos. The gloss turns on descendat: quanto plus ascendis, tanto plus corruis atque descendis, and Gelboe is glossed « descensus ». Montes is dropped from the quotation and restored by Migne in the gloss (« Montes Gelboe »); carried as printed both times.
• @1225C — the note (Psal. XCXVI, 5). XCXVI is not a Roman numeral (the psalm is XCVI); it is inside a [n:] and rides verbatim, untranslated. No marker.
• @1225D · Psal. LXVII — DECLINED. « Selmon mons Dei, mons pinguis » conflates v. 15 and v. 16, and Migne prints the received text of v. 16 correctly four lines later. A conflation is not a variant reading. The note on « Ut quid suspicamini montes coagulatos? » reads (Psal. V, 17.) where the verse is Ps. LXVII, 17 — his citation, ridden verbatim.
• @1225B–D and @1226B — umbra / lumen = shadow / light, fixed (Selmon … « umbra », then lumine).
• @1227A — an unclosed quotation, and the range's single !. « Petre, macta, et manduca [n: …]; et diceret, « Absit a me… — the first « is never closed. Mirrored, so the chunk stands at 27 open / 26 close in both Latin and English.
• @1227B and @1236D — rhetorical questions closed with a PERIOD. Quis in eum non dixerit insiluisse Dominum, cum … baptizavit. and Quis enim ignoret … « Petra autem erat Christus ». Both keep the period; no ? supplied.
• @1227C — a second unclosed quotation. « Transiliens colles. Quod quare sic, videamus. Mirrored; 21 open / 20 close in both.
• @1227D · Isa. II, 2 — [var:]. « Elevabitur Dominus super colles », where the Vulgate's subject is mons domus Domini, not the Lord. The subject is changed, and the sentence is being used to prove that some hills are good.
• @1228A · lemma 9 · Cant. II, 9 — [var: Vulg. dilectus meus, "my beloved"]. Migne's lemma omits meus: Similis est dilectus capreae, hinnuloque cervorum. A whole word of the lemma is gone, and it is the word that makes the clause the Bride's own speech; the English lemma accordingly reads "The beloved is like…", not "My beloved."
• @1228B — ut qui Patrem videt, et vivendo Patrem cuncta contuetur in Patre. vivendo with an accusative object is odd; rendered literally ("living the Father"). No emendation offered.
• @1228C ⭐ · Cant. II, 9 re-quoted — [var: Vulg. respiciens], the most load-bearing marker in these columns. The lemma at 1228B prints only En ipse stat post parietem nostrum, which agrees with the Vulgate; the divergence enters with the re-quotation « Ecce enim ipse post parietem nostrum stat, despiciens per fenestras ». The whole following exposition depends on despicere as looking DOWN — Hos itaque despicit, quasi deorsum laborantes at 1229C — and Migne prints despiciens again at 1230A.
• @1229A — odoratu, gustu, dictu et tactu — [cj:] DECLINED for want of a reading. The five senses have just been listed (visus, auditus, gustus, odoratus et tactus), and dictu — speech — stands in this second list where a sense belongs. The word is real, the syntax is sound, and there is no confident conjecture to propose: visu would duplicate what the eye-window paragraph has covered, auditu is likewise spent. The list simply names one thing that is not a sense. Rendered "of speech" and left.
• @1229B · Isai. LX, 8 — DECLINED, and deliberately. « Qui sunt hi qui ut nubes volant…? » for Vulg. isti — and Migne prints isti himself two lines later (« Qui sunt isti, » etc.). Kept distinguishable in English ("these" / "those") and logged. ⚑ Contrast 1225A, where the same class of divergence WAS marked because it sits on a numbered lemma.
• @1229C · Matth. VII, 13, 14 — [var: Vulg. ad perditionem]. « Lata enim est porta, et via spatiosa, quae ducit ad mortem » — and Migne writes quae ducit ad perditionem in his own gloss at 1229D, so the two readings stand a few lines apart on the same page.
• @1229C — a paragraph that ends with no terminal stop. …quae supra Dominum die qua baptizatus est apparuit. The English ends without one.
• @1229D — et arctam inchoans viam. A nominative participle whose subject could be the Lord or those he leads. Rendered with the same ambiguity ("and, beginning the narrow way").
• @1230A — mala non male agunt, "evil things not evilly", where both senses of the root stand in one clause. Rendered so.
• @1230B · I Tim. I, 13 — DECLINED. « Ecclesiam Dei persecutus sum: sed tamen misericordiam consecutus sum… » — I Tim. I, 13 has no such clause; it is I Cor. XV, 9 / Gal. I, 13. A conflation of two apostolic sayings under one citation is not a variant reading. Rendered as printed.
• @1230D ⭐ — a doubled clause, [d: …] taken BY THE LETTER OF THE RULE, with the analysis recorded here. Migne prints manifeste bonos et malos videns per fenestram; hos vero qui se occultant, intuens per fenestram; hos vero qui se occultant, intuens per cancellos. The clause is set twice, differing only in its final noun. Both are rendered, and the second occurrence of the shared run is wrapped. ⚠ The reservation matters: the intrusion is almost certainly the FIRST tail (per fenestram), not the second — the sense the paragraph has been building for two columns is that the openly good and bad are seen through the window and the hidden through the lattices, and the window clause has already been used. The rule is written for identical runs and does not decide a doubling that varies in one word; the marker was placed by the letter of the rule and the analysis left here rather than in the text.
• @1230D · I Cor. XIII, 3 — [var: Vulg. nihil mihi prodest]. …charitatem autem non habeam, nihil sum » — "I am nothing" for "it profiteth me nothing". The rest of the quotation is loose throughout and is rendered as printed.
• @1230D · Hebr. IV, 12 — an inline italic citation, not a note. Migne sets the reference as (Hebr IV, 12) in italics rather than as an [n: …]; reproduced verbatim in that form, and deliberately not tagged [f: …] — this is a lemma-and-gloss commentary. The connective divergence (nec est for et non est) is trivial and takes nothing. ⚑ The same Hebrews passage is quoted again at 1237B (there with et non est) and a third time, shorter, at 1251B; all three are carried as printed.
• @1231C · Rom. VIII, 38–39 — compressed from the Vulgate's ten members to five neques. Five negations printed, five rendered. Compression logged, no marker.
• @1231C ⭐ — FOLIOT PUNS ON amica HIMSELF, which is the second independent ground for "friend". ut amica nominum positio totum leviget ("the friendly setting of the names"), and Ait enim, « Amica; » nam ex hoc cognovi te amantem me. A rendering that cannot carry the pun makes the gloss unreadable.
• @1231C and @1236C — formosa "beautiful" against speciosa "comely", the two praise-words that collide across these columns. ⛔ See the conventions: the merge resolution moves speciosa to "fair", and 1236C's shipped English has not yet been changed.
• @1231D · Eph. V, 14 — surge / exsurge held apart ("arise" / "rise up"), because the verse prints both.
• @1232A · Cant. I, 3 re-quoted — [var:], and the divergence is a MOVED STOP. Migne prints « Trahe me post te; curremus in odore unguentorum tuorum » where the Vulgate points Trahe me: post te curremus. The stop moves and the sense moves with it: Migne's punctuation gives "draw me after thee", the Vulgate "after thee we will run."
• @1232B · lemma 11 · Cant. II, 11 — Jam enim hiems transiit. = Vulgate word for word, Migne splitting the verse into two lemmata (Imber abiit et recessit. follows at 1232D). No marker. ⚠ The re-quotation at 1236C inverts it: « Hiems transiit; abiit imber, et recessit », rendered "The winter is past; gone is the rain, and departed," so the inversion is visible. Jam enim is dropped there and the English drops it too.
• @1232C — gelu malitiae = evil, decided from the syntax.
• @1232D ⭐ — imber against pluvia: the gloss IS the lexical distinction. Imber non quaelibet pluvia dicitur, sed tantum hiemalis, and English has one word. Following the stylus/epistylium precedent, imber is kept in italic Latin in the defining sentence only ("Not every rain is called imber, but only the wintry"), while the lemma and the rest of the gloss read "rain". Logged rather than marked.
• @1232D — Lamech « homicida »: Migne guillemets the single word, and the guillemets are kept on the word.
• @1233B · Matth. VII, 27 — « domus illius ruina magna » for et facta est ruina illius magna. Word order and an added domus; no sense change. Declined.
• @1233D · lemma 12 · Cant. II, 12 — Flores apparuerunt in terra nostra. = Vulgate, as are the two following sub-lemmata (Tempus putationis advenit., Vox turturis audita est in terra nostra.). No marker. flores = "flowers" throughout, held distinct from mala "apples".
• @1234A ⚠ — fructus … malos, "bad fruits", and it is NOT apples. The most dangerous mala site in these columns: it sits in a paragraph about fruit, but the trees are the human race and the fruits are works. Decided from the syntax, recorded site by site because an error here would be undetectable.
• @1234C — [cj: Praedicatur; read Praedicator, "a preacher"]. Praedicatur is a real form ("is preached"), so nothing is broken to wrap; but the four nominative participles that follow (exquirens … nuntians) and the finite gemit … habitat … fugit have no subject, and the sentence's own opening clause names the praedicator. Additive, with Migne's word rendered first.
• @1234C — hiemis abcessum.* Assimilated spelling of abscessum*, attested medieval orthography; translated normally, no marker.
• @1235A · lemma 13 · Cant. II, 13 — Ficus protulit grossos suos. = Vulgate. grossi rendered "green figs" and held fixed through the gloss (1235A, 1235B ×2, 1236C). The companion lemma Vineae florentes dederunt odorem suum. (1235D) likewise agrees.
• @1235B · Luc. XIII, 6 — Arborem ficum habebat quidam plantatam in vinea sua for Arborem fici; an unguilleted paraphrase, no marker. The quoted portions (« Succide eam », « Domine, dimitte illam hoc anno… ») agree with the Vulgate but for hoc anno against et hoc anno. Logged.
• @1235D — Unde subditur with no colon or stop before the lemma. Carried.
• @1236A — Vineae vero Domini et una est, et multae sunt. Plural subject, singular verb, then plural. Rendered "of the vineyards of the Lord there is one, and there are many"; no English exponent for the mismatch.
• @1236C · lemma 14 · Cant. II, 13b–14 — [var: Vulg. in caverna maceriae, singular]. Migne prints the plural cavernis where the Vulgate has the singular — and the gloss immediately argues for ONE hollow, the spear-wound: In hac etiam maceriae cavernam inesse, quis nesciat (1236D). The divergence is material precisely because the exposition needs the singular. (The verse number 14 sits inside Migne's italic run and is reproduced there.)
• @1236C ⚠ — Migne's own re-quotation reads « Surge, amica mea, sponsa mea, et veni », where his lemma four lines above prints speciosa and the Vulgate has speciosa. Rendered as printed — "my bride" in the re-quotation against the lemma's praise-word. No [var:]: this is Migne's internal inconsistency, not a variant reading of the received text, and the two English words show it plainly.
• @1237A, @1237B — capitalization carried, including in one sentence. Assistenti autem Dilectae … dilectus ait — the English keeps the capital where Migne keeps it.
• @1237C · Psal. LXXIX quoted three times, cited (Psal. LXXIX, 4), (ibid., 15), (ibid., 20); the second is v. 8, not v. 15. Migne's citation, ridden verbatim, no marker. The three petitions themselves agree with the received text and are rendered with their three distinct vocatives kept apart.
• @1237D — velut evaserit. "Shall have as it were let them go"; the verb is intransitive in classical use and the object is supplied by the sentence. Rendered without emendation.

Lemmata 14–17, Capite nobis vulpes parvulas to Revertere — 1238B–1243B

• @1238D · lemma · Cant. II, 14 — Vox enim tua dulcis, et facies tua decora. = Vulgate, verbatim. ⭐ decora = "comely" here, following Foliot's own decora at 1194C — the decision that later forces speciosa to move. decor / decore = "comeliness" with it (1242A, 1250A).
• @1238D · lemma 15 · Cant. II, 15 — Capite nobis vulpes parvulas, quae demoliuntur vineas: nam vinea nostra floruit. = Vulgate, verbatim. No marker.
• @1238D — inter amicos suscipi, "received among friends" — the same usage as 1221C, and further support for the amica ruling.
• @1238D — coelestium novem ordines, hominemque eis adjunctum, et parem eis per gratiam factam. Feminine factam against masculine hominem. Latin gender has no English exponent**; no emendation, no marker.
• @1238A/1239A — « Sonet vox tua in auribus meis » re-quoted with the word order changed (« Et sonet in auribus meis vox tua »). English word order is fixed by grammar, so both read "let thy voice sound in mine ears"; the divergence lives only here.
• @1239A ⚑ — the seven signs of spring re-quoted with two words dropped, and NO marker. The lemma chain at 1233D–1235D reads flores apparuerunt in terra nostra … vox turturis audita est in terra nostra; this re-quotation prints in terra both times and reorders abiit imber to imber abiit. The English follows each — "in our land" there, "in the land" here. This is the work quoting itself, and the received reading is printed as its own lemma one chunk earlier, which is exactly the shape that blocks the marker. Third sighting of the gap; see Standing Item 2.
• @1239C · Hebr. I, 14 — Migne drops the Vulgate's sunt (Nonne omnes administratorii spiritus…). No [var:]: English supplies a copula either way, so the marker would annotate a difference the reader's page cannot show.
• @1239D — suos in hoc successus reputant, si Ecclesiam … impedire valeat. Singular verb against the plural subject hi … potestates contrariae. Rendered with the singular ("if it be able in anything to hinder the Church"). Named here because it is the site in these columns most likely to be read as OUR error.**
• @1239D — quam ad Deum tendentem … impedire valeat. The relative quam is picked up by participles with no finite verb of its own. Rendered "the Church, her that tends toward God…", keeping the anacoluthon visible rather than converting it to a clause.
• @1240A — an unclosed quotation, mirrored. The Luke XIII, 32 quotation (« Ite, et dicite vulpi illi…) runs straight into ac si dicat and never closes; the chunk stands at 25 open / 24 close on both sides.
• @1240B — the note (Psal. XXVIII, 10) on Introibunt in inferiora terrae. That clause is the continuation of Ps. LXII, 10, quoted two lines above with its own correct note; Ps. XXVIII, 10 is Dominus diluvium inhabitare facit. This looks like a repeated-verse split that took the wrong number, and it is the strongest candidate of the work's wrong locators — and it is still not corrected, per the standing rule.
• @1240C — de allophylis rendered "from the Philistines". allophylus is the Vulgate's own word for the Philistines through 1–2 Kings; this is translation, not substitution.
• @1240D — titillatione mala excrescente = evil (ablative feminine singular agreeing with titillatione).
• @1241C · lemma 16 · Cant. II, 16 — Dilectus meus mihi, et ego illi. = Vulgate, verbatim.
• @1241D · lemma · Cant. II, 16b–17 — Qui pascitur inter lilia. 17. Donec aspiret dies, et inclinentur umbrae. = Vulgate, verbatim.
• @1241D ⭐ — [cj:] DECLINED because the printed text construes, and this one is worth the space. illi soli cui commendata sunt, nobis cum noluerit aperienda, commendamus. First read, this looks like the classic negation defect: the sense seems to want voluerit, and "to be opened to us when he shall be UNwilling" is the opposite of the argument, which would bring it inside the rule that refuses to excuse a failed negation on grounds of incoherence. Declined, because the printed text construes without emendation: aperienda is a gerundive serving as the object of noluerit — "we commit them to him alone to whom they have been committed, since he has not willed that they be opened to us." That reading is coherent and continuous with the preceding ne tentemus ingredi, nec … praesumamus aperire. A [cj:] here would have glossed away a good sentence. Logged rather than passed over silently, because the alternative is the sort a blind reader will propose.
• @1242A — six nons in one sentence, all carried. non inflammet libido, non stimulet avaritia, non effeminet lascivia, non sternat ambitio, invidia non maceret, non secularis cura … sollicitet. None supplied, none dropped.
• @1242B — bonis … bona, malis mala quidem sine fine permansura = "to the evil, evil things", in explicit antithesis with bonis … bona.
• @1242C · Matt. XXII, 30 — [var: Vulg. neque nubentur, "nor be given in marriage," passive]. Migne prints cum jam non nubent homines, neque nuptum tradent. The divergence is one of AGENCY: the Vulgate's second limb is passive, Migne's is active, which puts the men in the position of those disposing of brides rather than of those disposed of. (Migne also drops the closing in caelo; not marked — no sense turns on it and the note locates the verse.)
• @1242C · Psal. XCV, 13 — DECLINED, on the 1160C precedent. Judicabit autem omnes in aequitate, et populos in veritate sua, where the psalm reads judicabit orbem terrae in aequitate — and the whole world has just been named in the preceding clause. Declined because Migne prints no locator here: naming a "received reading" would rest on OUR identification of the verse rather than on his. This should stay the rule.
• @1242D — pallio terrae noctis obducuntur. Three genitives in a row with no obvious governor; rendered "are drawn over with the cloak of the night of the earth", following the printed order. Recorded because a smoother "with the earth's cloak of night" would be a reading, not a rendering.
• @1243A · lemma · Cant. II, 17b — Revertere, similis esto, dilecte mi, capreae, hinnuloque cervorum super montes Bether. = Vulgate, verbatim.
• @1243A — an unclosed quotation, mirrored. The II Tim. IV, 7 quotation runs into Unde supplex orat and never closes; 16 open / 15 close on both sides.
• @1243B · II Tim. IV, 7–8 — DECLINED, and the divergence is recorded in full so it is not lost with the marker. Migne prints reposita nobis est haec corona justitiae, quam ecce nobis retribuet justus judex against Vulg. reposita est mihi corona justitiae, quam reddet mihi Dominus in illa die justus judex — singular turned plural, in illa die gone. A real divergence — but the printed note reads [n: (II Tim. IV, 7)] and stops at fidem servavi, and the divergent material is v. 8, which carries no locator of its own. Same principle as 1160C and 1242C.

Caput III — 1243C–1255B

⚑ Every Song lemma in this chapter agrees with the Vulgate. The three [var:] markers taken here are all on other books — Romans, Matthew, Matthew again — which is the shape to expect once the lemmata themselves are clean: the commentary's own text is stable and the proof-texts adduced round it are not.

• @1243C · lemma 1 · Cant. III, 1 — In lectulo meo per noctes quaesivi quem diligit anima mea; quaesivi illum, et non inveni. = Vulgate (the Vulgate sets : where Migne sets ; — punctuation only, carried as printed). lectulus = "little bed".
• @1243C — a full stop that breaks the sentence in two, and the English is as broken as the plate. Exoravit haec Sponsa, vel fidelis quaecunque anima. Dilectum suum, ut coelum non deserens… — the stop leaves Dilectum suum, the accusative object of Exoravit, stranded at the head of the next sentence. Rendered with the stop where he prints it: "…has entreated. Her Beloved, that, not deserting heaven…". Not marked: no type is defective and nothing is missing. Deliberately visible.
• @1243D — [sic: perfecto], and it is the dangerous class again. ibi perfecto luxus exuberat. perfecto is a real Latin form with no grammatical slot — no noun to agree with and no construction that takes it. The near-certain reading is profecto, "assuredly", the intensifier the clause wants. A real word takes no automated flag, and this is the same shape as 1192A's [sic: suum]. The quarantined run carries no negation, no conjunction and no verb — deleting it leaves ibi luxus exuberat, which is complete — so nothing needed supplying beside the marker.
• @1244B · Rom. VIII, 38 — [var: Vulg. Certus sum enim, "For I am sure"]. Migne prints Scimus enim. Person and modality both move: Paul's personal certainty becomes the community's knowledge — and Foliot is at that moment arguing about what a whole ORDER of souls may say cum Apostolo, so the plural helps his argument, which is exactly why the reading must not be silently corrected.
• @1244B — the note (III Reg. II, 11) on the boys' mockery of Eliseus. 3 Kings II, 11 is the death of David; the fiery chariot is 4 Kings II, 11 and the bears are 4 Kings II, 23–24, which Migne cites correctly nine words later. ⚠ Foliot's own prose names Elias's chariot in the preceding clause, so the collapse-on-checking risk is live: the note may be Migne's citation of the chariot, misplaced, rather than a wrong number. Not corrected.
• @1244D — a ? inside the Joan. I, 14 quotation (plenum gratiae et veritatis?), on what is a statement in every text of the Gospel. Carried.
• @1244D — an unclosed quotation, mirrored. The Matt. XVI, 19 quotation (« Et tibi dabo claves…) runs into Et maledictus erit and on to the paragraph break; 30 open / 29 close on both sides. (Its two ibid. notes depend on the Matt. XVI note, which is correct — one of the reasons nothing in the chain may be touched.)
• @1244D — hos sequentes divites mali = "evil rich men"* (nominative plural masculine agreeing with divites*).
• @1245C · I Cor. XII, 28 and @1246C · Isa. XXVIII, 20 — each drops one connective (Et / enim) against the Vulgate. No marker.
• @1246B — si mala nobis occurrant et pessima = "evil things" (nominative neuter plural, subject of occurrant, with the superlative pessima).
• @1246C — eam malus si spiritus occupaverit = "an evil spirit" (nominative masculine singular with spiritus, split word order).
• @1246D · Matt. VII, 7 — [var: Vulg. et dabitur vobis, "and it shall be given to you"]. Migne prints Petite, et accipietis. Migne's reading makes the asker the agent of receiving; the Vulgate's makes the gift the act of another — and the whole paragraph turns on the triad (petitioni nil negatur, quaestioni nil clauditur, pulsationi promptius aperitur), so the first limb is load-bearing.
• @1246D — the three-part gloss on Isa. XXVIII, 20, kept in MIGNE'S order. His sequence is « pallium » … « utrumque » … « operire » … non potest. English wants the verb before its object, which would reorder his quoted words. Kept in his order ("the « cloak » … « both » … is not able to « cover »"), fronting the object, so that the guillemet sequence on the English page matches the Latin twin position for position.
• @1246D — the lemma re-quoted a third time, and now as a question. quid est quod ait, « Quaesivi, et non inveni illum? » — object moved to the end and the mark inside the quoted lemma. Rendered "I sought, and found him not?" so that the difference from the lemma is visible.
• @1247A · lemma 2 · Cant. III, 2 — Surgam, et circuibo civitatem: per vicos et plateas quaeram quem diligit anima mea: quaesivi illum, et non inveni. = Vulgate, verbatim.
• @1247A · lemma 3 · Cant. III, 3 — Invenerunt me vigiles qui custodiunt civitatem. = Vulgate, verbatim. ⭐ vigiles = "watchmen" and custodiunt = "keep", and they must stay distinct from Cant. V, 7's custodes qui circumeunt at 1275D. Checked at the merge: these are two different verses, Migne agrees with the Vulgate in both, and Foliot glosses them differently — here the vigiles are angels who teach the law of confession. No collision.
• @1247A — bonis et malis communis = "to the good and to the evil" (dative plural masculine, of persons).
• @1247B · lemma — Num quem diligit anima mea, vidistis? = Vulgate, verbatim.
• @1247B — the column anchor set flush against the word, [1247B]Qui sunt hi vigiles? Mirrored exactly, no space inserted.
• @1247B — the note (Matth. X, 18) on Videte ne contemnatis unum ex his pusillis. That is Matt. XVIII, 10; the digits are transposed. Matt. X, 18 is ad praesides et ad reges ducemini. Kept verbatim.
• @1247B — quid boni, quid mali quis egerit = "what evil" (partitive genitive neuter, answering quid boni).
• @1247C — nec amplius dixisse, vel quid egisse, carried as "and has said nothing further, nor done anything."
• @1247D · lemma 4 · Cant. III, 4 — Paululum cum pertransissem eos … in cubiculum genitricis meae = Vulgate (the Vulgate sets : for Migne's ;; genitricis/genetricis is spelling). ⚠ Migne prints NO terminal stop, and the English does not supply one.
• @1247D — tenui eum, nec dimittam, the answering positive-by-negation, and its gloss at 1248B piles up four more (Hunc itaque non dimittet, quia non silebit, non a praedicatione, non a contestatione … cessabit). All five carried.
• @1248A ⚑ — a divergence that is REAL in the Latin and INVISIBLE in English. The lemma prints tenui eum and donec introducam; the gloss prints « Tenui illum, » and « donec inducam ». Both render identically ("I held him", "till I lead him"), so the divergence lives only here. Recorded because a reader collating the Latin twin will see it and should not read the English as a conformation.
• @1248B · lemma 5 · Cant. III, 5 — the Adjuro vos verse, third printing. donec ipsa velit, agreeing with the Vulgate; re-quoted eleven lines below without per capreas cervosque camporum. No marker — see the six-printing table at 1223A. The lowercase dilectam of both printings is mirrored ("the beloved").
• @1248B — Sequitur. with a full stop, where every other instance in these columns prints Sequitur:. Rendered "There follows." Not a defect, and not normalized. (Sequitur: = "There follows:" is fixed across all 53 chunks.)
• @1248C · lemma 6 · Cant. III, 6 — Quae est ista, quae ascendit per desertum? = Vulgate, verbatim. ⚠ This is a DIFFERENT verse from 1298A's Quae est ista, quae ascendit de deserto — "by the desert" here, "from the desert" there. The frame is deliberately identical because Migne's frame is identical; the preposition differs because his does.
• @1248D — quae sic ascendit per desertum? for the lemma's quae ascendit. Rendered "who thus goeth up by the desert?", so the added word is visible.
• @1249A — « Quae » igitur « ista haec, » with no copula. English cannot omit the verb; rendered "« Who » therefore « is this she, »", supplying only the copula English grammar requires and leaving both quoted fragments in their printed places.
• @1249A — two exclamation marks, both Migne's, in the Quae est ista gloss (qualis, et quanta est ista! and « ista est! »), where the sentence is interrogative in form. Carried; not converted to ?.
• @1249B · lemma · Cant. III, 6b — Sicut virgula fumi ex aromatibus myrrhae, et thuris, et universi pulveris pigmentarii. = Vulgate, verbatim.
• @1249C ⭐ — virgula = "a little rod", because Foliot argues the diminutive. Douay gives "a pillar of smoke", and that rendering cannot be used here: the gloss is built entirely on the diminutive — quae non virga, sed « virgula » recte dicitur, twice — quia recta et extenuata, agilis quoque compunctio cordis intelligitur. With "pillar" the page would read "not a pillar, but a pillar",* and the two guilleted occurrences would be expounding a word not on the page. virga* = "rod" with it.
• @1250A · lemma 7 · Cant. III, 7 — En lectulum Salomonis sexaginta fortes ambiunt ex fortissimis Israel. = Vulgate, verbatim. ⭐ sexaginta = "sixty", never "threescore" — Foliot's gloss is arithmetical for a full column (sexagenarius ex duobus numeris perfectis constat, sive sexies decem, aut decies sex, legis denarium, Evangelii senarium), and "threescore queens" glossed by "the number sixty" would explain a word the page does not print. octoginta = "eighty" with it.
• @1250C · Matt. XXV, 35–36 — collegistis for collexistis and cooperuistis for operuistis. Ordinary medieval variants of the same verbs with the same sense, not divergence. No marker.
• @1250C — bona bonorum, et mala malorum = "the evil things of the evil."**
• @1251A · lemma 8 · Cant. III, 8 — Omnes tenentes gladios. = Vulgate, verbatim.
• @1251A — an unclosed quotation, mirrored. The Ephes. VI, 14–17 quotation runs into Haec quidem armatura and never closes; 18 open / 17 close on both sides.
• @1251B · Hebr. IV, 12, the SECOND quotation of the verse in this work, and shorter. At 1237B Migne prints Vivus est enim sermo Dei, et efficax, et penetrabilior…; here he prints Vivus enim Dei sermo est, … and omits et efficax and the et before discretor. The English differs accordingly ("living, and more piercing" here against "living and effectual, and more piercing" there). Migne prints no note on this occurrence, so no [var:].
• @1251B — quod verbo « contexunt, » hoc manibus exercere intelliguntur. contexunt sits inside guillemets as though quoted, but no source is given and the sense wanted is closer to detexunt ("unweave") or simply "profess". Rendered literally, "that which they « weave together » in word". No marker: the word is real, it construes, and the sentence asserts nothing false — it is merely odd. Flagged as the site in these columns most likely to be judged differently.
• @1251C · lemma — Et ad bella doctissimi. = Vulgate, verbatim.
• @1251D · lemma — Uniuscujusque ensis super femur suum. = Vulgate, verbatim.
• @1251D — qui parcit malis indulta venia = "the evil" (persons), dative plural masculine, object of parcit.
• @1251D — hostis … cui poenas meruit, impunis evadit. Rendered "the enemy … escapes unpunished from him at whose hands he has deserved punishment", which is the only reading the printed cases allow. Recorded because cui invites the opposite binding — the shape a polarity read must watch.
• @1252A · lemma — Propter timores nocturnos. = Vulgate, verbatim. The Belial paragraph beneath it prints nulla … nulla … nulla … nullus … nulla — five negations, five carried.
• @1252B · lemmata 9/10 · Cant. III, 9–10 — Ferculum fecit sibi rex Salomon de lignis Libani … media charitate constravit propter filias Jerusalem. = Vulgate, verbatim.
• @1252B ⭐–1253B — ferculum = "litter", and the corpus was consulted first. In 11550 (Richard on the same book) the word was kept in italic Latin at all eleven occurrences — and that was right there, because Richard etymologizes it: ferculum a ferendo, quia fertur de loco ad locum, and an English word sharing no root with "bear" would leave his gloss explaining nothing. Foliot does not do that. His gloss is functional: Surgens enim e lecto Rex, ferculo se suscipit, ut de loco feratur ad locum — the King rises from the bed and takes to the ferculum to be borne about; it has silver pillars, a golden reclinatorium, a purple ascent, and an area in the middle paved with charity. Nothing in his argument depends on the word's root, and every clause construes with "litter" (also Douay's word at Cant. 3:9). ✅ The seam closed itself: the reader who held 1253B, without the passage that explains the word, reported "litter" independently, and the reasoning here was written before that report arrived. ⚑ Deliberately different from 11550, and both are correct as they stand. Do not sweep either.**
• @1252B — lectulus "little bed" against lectus "bed", side by side. Surgens enim e lecto Rex, of the lectulus Salomonis.
• @1252B — the consequential renderings in the same passage. reclinatorium → "reclining-place" (not italic Latin, and not Douay's "seat"): the gloss is quasi reclinatorium ex auro sapientiae composuit, a place to lean, with no etymology argued. media … charitate constravit → "the middle parts he paved with charity", neuter plural, held fixed through « Media, » id est, area quae in medio ferculi est and media namque charitate constrata sunt — rendering media as a singular "the midst" would break the second re-quotation. area ferculi → "the floor of the litter".
• @1252C — Ferculum fecit rex Salomon…, dropping the lemma's sibi. Rendered without "himself" accordingly.
• @1253B · lemma — Egredimini, et videte, filiae Sion, regem Salomonem. = Vulgate, verbatim.
• @1253C and @1253D — two note-citations printed one sentence out of place, kept exactly where Migne prints them. [n: (Gen. IV, 8)] closes the diademate quo coronatus est clause, one sentence before the Cain material it belongs to; [n: (Gen. XXXIV, 1)] closes the Cain sentence and belongs to the Dina material that follows. Contents untranslated, positions untouched — and note that an Ibid. elsewhere may depend on the printed sequence.
• @1253D — Egressiones enim istae malae sunt = evil, and it is nowhere near a fruit.
• @1254A · Gen. XII, 1–3 — et in te benedicentur for the Vulgate's atque in te benedicentur. Conjunction only; no marker.
• @1254A · Gen. XXVIII, 10–14 — Bersabe / Haram for Bersabee / Haran are spellings, not defects. The lapidem magnum ("a great stone") is Foliot's own narration, not a quotation — Gen. XXVIII, 11 has no magnum — so it takes nothing. The quotation proper (Et benedicentur in te…) agrees with the Vulgate.
• @1254B ⭐ — quanam « Aaron » — [cj:] DECLINED, and this is a PLATE-READ CANDIDATE. quanam « Aaron » filiabus Sion consulitur egredi, ut Regem gloriae in diademate suo contemplari valeant et intueri? quanam is a feminine ablative interrogative with nothing to agree with, and « Aaron » — the patriarch's name, in Migne's own guillemets, in a paragraph that has nothing to do with Aaron — stands where that noun should be. Rendered literally, "by what « Aaron » is it counselled to the daughters of Sion to go forth…?", with the guillemets and the ? as printed. Two good conjectures compete, of different kinds: egressione ("by what going forth") — contextually much the stronger, since the whole paragraph is a taxonomy of egressiones, the preceding sentence ends …passibus accedens egressio, and the following clause answers non ea utique qua Cain egressus est; and ratione ("by what reasoning"), the ordinary idiom for quanam, palaeographically no worse. Choosing between them would be a guess dressed as a reading, and the marker is additive precisely so that it never has to be one. The conjecture is recorded in full so a reader of the plate can settle it. ⚑ Note also that Migne's guillemets round the word suggest his own editors already regarded it as suspect — a further reason not to stack a marker of ours on top. See Standing Item 3.
• @1255A · Joan. XX, 28 — [var: Vulg. couples the two titles, "my Lord and my God"]. Migne drops the conjunction: Dominus meus, Deus meus, leaving the two titles in bare apposition.
• @1255B · Joan. XVII, 3 — [var: Vulg. reads that they may know thee the only true God]. Migne prints ut videant te verum Deum — seeing for knowing, and solum dropped. Not cosmetic: the whole paragraph is about the daughters of Sion SEEING the King (ad hunc videndum … videant et attendant), so the altered verb is doing work in Foliot's argument, whoever introduced it. ⚠ Compare 1210B, where the same verse loses only solum and the marker addresses that.

Caput IV — 1255C–1267D

Every named Song lemma in this chapter agrees with the Vulgate, apart from Migne's own segmentation and the truncation at 1255C.

• @1255C · lemma 1 · Cant. IV, 1a — Quam pulchra es amica mea, quam pulchra es! oculi tui columbarum. = Vulgate. Rendered "How beautiful art thou, my friend, how beautiful art thou! thine eyes are those of doves." — amica = "friend", pulchra = "beautiful", and the ! rides as printed. "Thine eyes are those of doves" reproduces the rendering already fixed at 1206D for the same bare genitive, so the two sites agree inside the work. Migne's only other departure is segmentation: he breaks the verse after columbarum and sets Absque eo quod intrinsecus and Capilli tui sicut greges caprarum as separate lemmata. Not a variant.
• @1255C ⭐ — THE LEMMA IS TRUNCATED, AND FOLIOT'S OWN GLOSS PRINTS THE FULL FORM. Absque eo quod intrinsecus. — no verb. The Vulgate reads absque eo quod intrinsecus latet, and four lines below Migne's own gloss quotes « Absque eo quod intrinsecus latet; » — and the whole exposition (what lies hidden within the dove-soul) rests on latet. ⛔ [var:] DECLINED for the mechanical reason: the received reading is printed in this very chunk as Migne's own gloss, so no faithful marker can be written. What the English does instead: it renders the truncation visibly. The lemma reads "Besides that which within." with no verb supplied, and the gloss and the second occurrence at 1257B read "« Besides that which lieth hid within; »". A reader meets the defect on the page even though no marker could carry it. See Standing Item 2.
• @1256B · lemma — greges caprarum, "flocks of GOATS", against the gloss's capreae, "ROES". ⭐ Two Latin words, two English words, on the laeva/sinistra precedent: the lemma says goats and every gloss says roes, including de monte Galaad, and the next chapter's lemma has hinnuli capreae (young roes). The slip from one to the other is Foliot's and is now visible to the reader instead of being smoothed away.
• @1256C · lemma 2 · Cant. IV, 2 — Dentes tui sicut greges tonsarum … et non est sterilis inter eas. The Vulgate inverts to sterilis non est. Word order only; the negation is present and carried ("and there is not one barren among them"). No marker.
• @1256C — internal re-quotation drift, recorded, unmarked. The gloss glosses gemellis fetibus ("twin young") with gemellis ipsos ditant fructibus ("enrich them with twin fruits"). Foliot has slid from offspring to produce. Rendered as printed in both places, distinct words kept distinct; no marker, because the second is his own exposition and not a re-quotation of the lemma.
• @1256C — a missing full stop, preserved. sancti … ad superna ascenderunt Sequitur: — no stop between the sentence and the formula. No mark supplied: the English reads "…have ascended to things on high There follows:". It looks like our error and is not.
• @1256D · lemma 3 · Cant. IV, 3a — Sicut vitta coccinea, labia tua; et eloquium tuum dulce. = Vulgate. Re-quoted at 1257A as « Eloquium tuum dulce. », held to the lemma rendering. dulcis = "sweet".
• @1257A · lemma · Cant. IV, 3b — Sicut fragmen mali punici, ita genae tuae, absque eo quod intrinsecus latet. = Vulgate. Re-quoted at 1257B as « ut fragmen mali punici: » and « Absque eo quod intrinsecus latet; », both held. ⛔ Homograph check: mali punici is the FRUIT (pomegranate) in all five of its occurrences here, governed by fragmen / malum punicum Ecclesia est. Nothing in these columns is malus, -a, -um "evil".
• @1257B — a full stop inside a quotation frame, preserved. Hae tamen genae sic sunt. « ut fragmen mali punici: » — Migne stops the sentence before the quotation that completes it, and starts the quotation with a lowercase ut. Both preserved.
• @1257C · lemma 4 · Cant. IV, 4 — Sicut turris David collum tuum … omnis armatura fortium. = Vulgate. Re-quoted at 1257D as « mille clypei, » and glossed fortium armatura twice.
• @1258A · lemmata 5–6 · Cant. IV, 5–6a — Duo ubera tua sicut duo hinnuli capreae gemelli, qui pascuntur in liliis. Donec aspiret dies, et inclinentur umbrae. = Vulgate but for Migne's punctuation (hinnuli, capreae gemelli). ⭐ aspiret is rendered "BREATHE", not the conventional "break", precisely because Foliot re-uses the verb — cum dies judicii aspiraverit at 1259A, "when the day of judgment shall have breathed". A "break" in the lemma would leave that gloss expounding a word not on the page. Re-quotations « ut duo hinnuli capreae, » (1258D) and « pascuntur in liliis, » (1259A) both held.
• @1258B — qua terram « videbat » serpens insidians — [cj:] DECLINED. Another word set in guillemets by Migne himself. The clause construes as printed — the serpent, lying in wait at the heel, "saw" the ground by that outermost part — oddly but without a false assertion, so it is rendered literally and left alone. See the guillemet heuristic in Standing Item 3.
• @1258B / @1262C · Gen. III, 15 re-quoted with the words transposed. Inimicitias ponam inter te et mulierem here; at 1262C, calling it principali et prima prophetia, Migne prints Ponam inimicitias. Word order only; both rendered "I will put enmities", so the two passages agree in English as they agree in sense. No marker.
• @1258C ⭐ · Deut. XVIII, 18–19 — [var:]: THE SPEAKER'S PERSON IS INVERTED, and the control is inside the work. Migne prints Prophetam suscitabit vobis Deus … similem mei … ponet verba sua … ipse ultor existet — God spoken of, in the third person. The Vulgate has God speaking as himself (suscitabo … ponam verba mea … ego ultor existam, and similem tui). [var: Vulg. has God himself speaking in the first person throughout … and "like to thee"]. ⚠ The decisive control is inside the work: 1261A prints the SAME verse in the FIRST person under the same note, and 1282C prints it in the first person a third time. Three witnesses inside the work, two first-person and one third — so 1258C is a single-site plate divergence, not the author's habit, exactly the Nigra sum ET formosa shape at 1184A. Both are rendered as printed, so the English of the two passages differs, deliberately. Do not "correct" it.
• @1258C and @1261A — the note (Deut. XXVIII, 18, 19) is printed twice for what is Deut. XVIII, 18–19. Kept verbatim both times, not "fixed": agreement between readers is not evidence, and a following Ibid. may depend on the printed reference.
• @1258C · Jer. XXXI, 22 — [var: Vulg. puts the verb in the perfect]. Migne prints Faciet Dominus novum for creavit Dominus novum super terram. The verse is being adduced as a prophecy of the Incarnation, so the tense is the point of citing it. (The mulier/femina substitution is not separately marked: English has one word.)
• @1258C · Dan. IX, 24 — [var:], and the printed line is not the Vulgate at any point. Migne prints cum venerit Sanctus sanctorum, cessabit in vobis unctio. Dan. IX, 24 reads ut … ungatur Sanctus sanctorum — the Holy of holies is anointed; there is no ceasing anointing in the verse. The form printed is the medieval testimonium against the Jews. ⚑ Because Migne DOES print a locator here, the received reading can be named without our supplying the identification — which is exactly the distinction that made 1160C decline.
• @1258C · Isa. VII, 14 and Ezech. XLIV, 2 — Isaias truncated after pariet Filium (no et vocabitur nomen ejus Emmanuel); Ezechiel abridged in the middle and printing transiet for transibit. Abridgement and a tense form; no marker. Both negations (non aperietur, vir non transiet) carried.
• @1259A — umbrae hae igorantiae, carried untranslated. igorantiae for ignorantiae is not a Latin word: "these shadows of igorantiae". The conjecture is certain — the phrase answers the lemma's inclinentur umbrae and the gloss's quae nos ad praesens impediunt — and it lives here, not in the text.**
• @1259B · Col. III, 5 — concupiscentiam malam = evil**, and it is nowhere near a fruit.
• @1259C · lemma 7 · Cant. IV, 7 — Tota pulchra es, amica mea, et macula non est in te. = Vulgate. The negation is carried twice, in the lemma and in the re-quotation at 1259C ("all beautiful, and there is not a spot in thee"), in the same words both times.
• @1259D · lemma 8 · Cant. IV, 8 — Veni de Libano, sponsa mea … de cubilibus leonum, de montibus pardorum. = Vulgate.
• @1259D–1260A — head → summit, and three Latin words for the top of a mountain. The gloss quotes Coronaberis de capite Amana (matching the lemma) and then immediately corona de cacumine Amana. capite · vertice · cacumine occur within two lines and are kept visibly distinct in English as head · top · summit, on the laeva/sinistra precedent. Recorded, unmarked.
• @1259D — an ORPHAN CLOSING guillemet, mirrored. Migne prints cum dicitur, Coronaberis de capite Amana; » — a closing guillemet with no opening one. The English reproduces the orphan close in the same place, which is the whole of the chunk's 28-open / 29-close asymmetry.
• @1260B — the gloss re-quotes « De cubilibus leonum, et de montibus pardorum. » where the lemma has no et. Conjunction only; recorded, unmarked.
• @1260C · lemma 9 · Cant. IV, 9a — Vulnerasti cor meum, soror mea sponsa, vulnerasti cor meum. = Vulgate. ⭐ The repetition is rendered AS a repetition, in identical words, here and at every re-quotation (1261D the half-line, 1262D the bare « Vulnerasti cor meum; », 1262D the doubled form again in full) — because Foliot's argument at 1262D–1263A IS the doubling: he counts the two woundings and asks for a third. Any elegant variation would destroy the gloss.
• @1260D · Psal. I, 1 INVERTED — by the author, not by the plate. quae dudum abiit in consilio bonorum, et in eorumdem via stetit, et in cathedra prudentiae sedem sibi firmavit — the psalm's consilio impiorum … via peccatorum … cathedra pestilentiae turned positive. This is Foliot's own rhetorical inversion, not a quotation and not marked by Migne as one; no marker. Recorded because it looks like a corruption and is not.
• @1261A — Dei Spiritu Deo dudum « munitos » — [cj:] DECLINED. munitos ("fortified") sits oddly with the dative Deo, and unitos ("united to God") would suit the run of the sentence, which is about the elect being conjuncti … et coadunati … unus cum eo spiritus. Declined for two reasons: "fortified for God by the Spirit of God" construes without asserting anything false, and Migne has already flagged the word himself with guillemets, which is a reason not to stack our marker on his. Recorded because a plate reader may want to test unitos.
• @1261B — a space before a semicolon, preserved. quibus « vivere Christus est, et mori lucrum » ; Deo dicentes assidue: — the spaced semicolon is Migne's and is reproduced.
• @1262A — [cj: aeternitas; read aeternitatem]. quo contemplatur aeternitas. contemplatur is deponent and active in sense, so the printed nominative makes eternity the one doing the contemplating — in a sentence whose whole subject is the Church's two eyes and what each of them looks at (unus quo intuetur proximum, alter quo videt Deum). The word is a real form, so nothing is broken to wrap, and the faithful English asserts something Foliot did not. Rendered "wherewith eternity contemplates [cj: …]".
• @1263A · lemma 10 · Cant. IV, 10 — Quam pulchrae sunt mammae tuae, soror mea sponsa! pulchriora sunt ubera tua vino … = Vulgate. The ! rides as printed, twice.
• @1263A ⭐ — mammae and ubera: A DELIBERATE FLATTENING, DECLARED. Migne prints two Latin words for the breast in one verse, and the gloss uses them as plain synonyms in a single sentence — « quam pulchrae sunt mammae tuae! » id est, doctrina uberum tuorum. Both are rendered "breasts". ubera was already fixed as "breasts" at 1258A, no argument anywhere turns on the difference, and "paps" for ubera would split a word the work treats as one. Flagged rather than silently done.
• @1263A — Migne's ? closing what is in sense an exclamation (… et sustentationem parvulorum?). Carried as printed, recast as a direct question English can hold ("how beautiful is the teaching … ?"). The Latin is already direct.
• @1263B · lemma 11 · Cant. IV, 11 — Favus distillans labia tua, sponsa. / Mel et lac sub lingua tua. / Et odor vestimentorum tuorum sicut odor thuris. = Vulgate; Migne divides the verse into three lemmata.
• @1263B — Foliot runs his own word straight out of the closing guillemet. « Favus distillans labia tua, sponsa » Christi. Rendered so: "« A dropping honeycomb, thy lips, O bride » of Christ." Guillemet positions unchanged.
• @1263D · Matth. XI, 28 — Migne drops omnes (Venite ad me, qui laboratis). One word, no sense change; no marker.
• @1264B · lemma 12 · Cant. IV, 12 — Hortus conclusus soror mea sponsa, hortus conclusus, fons signatus. = Vulgate. Re-quoted in full at 1264B and as « hortus conclusus » and « Fons signatus; » at 1264C.
• @1264D ⭐ · lemma 13 · Cant. IV, 13a — Emissiones tuae paradisus MALORUM punicorum, cum pomorum fructibus. = Vulgate — AND THE HOMOGRAPH AT ITS SHARPEST. Decided from the syntax at every one of the nine sites in these columns. malorum punicorum / mala punica / Malum punicum are the fruit (genitive plural of malum punicum, governed by paradisus and glossed sub rubeo cortice multitudinem granorum claudit) — pomegranates. pomorum / poma beside it is the second fruit-word — apples — and the distinction is load-bearing: at 1265B Foliot CONTRASTS them, Mala namque punica sunt, qui … sanguine laureati discedunt: illi vero fructus, qui solum poma dicuntur — the pomegranates are the martyrs, the apples are the confessors. A single "evil" anywhere in this chain would destroy the argument and pass every check.
• @1264D — emissiones = "sendings forth", not "plants" or "shoots", because Foliot glosses the noun with its own verb three times (fructus emittit · odorem virtutum … emittunt · nec solum mala punica emittuntur a te). An unrelated English noun would leave the gloss explaining a word not on the page.
• @1265C — the Sirach citation is INLINE ITALIC, not a note. Migne prints (Sirac III, 20) inside the quotation, in italics, where every other citation in the chunk is an [n: …]. Reproduced exactly as printed, verbatim and untranslated, in the same place — deliberately NOT converted to [n: …], because marker positions are matched positionally across the pair and a conversion would desynchronize them. Not tagged [f: …].
• @1265C — the note (Luc. XIV, 31) on Qui se humiliat exaltabitur. The quotation is Luke XIV, 11; XIV, 31 is the king going to war. Kept verbatim, unmarked.
• @1266B · lemma 15 · Cant. IV, 15 — Fons hortorum, puteus aquarum viventium quae fluunt impetu de Libano. = Vulgate.
• @1266C — a SECOND opening guillemet with only one close, mirrored. Migne opens at « Fluminis impetus … and opens again at « Quae fluunt cum impetu de Libano … , closing only once. The English reproduces both opens and the single close in the same places; this is the whole of the chunk's 27-open / 26-close asymmetry. (The Ps. XLV, 5 text itself agrees with the Vulgate.)
• @1266C — the phrase in three forms inside four lines. The lemma fluunt impetu de Libano, the psalm-tail fluunt cum impetu de Libano, and the gloss « cum impetu, » id est impetuose. Rendered "with force" throughout so the gloss lands; the added cum is not a sense change and took no marker.
• @1266D · lemma 16 · Cant. IV, 16a — Surge aquilo, et veni auster, perfla hortum meum, et fluant aromata illius. = Vulgate. ⚠ At 1267B the gloss quotes « Fluant ejus aromata, » for the lemma's aromata illius. Same possessive in English; the two Latin pronouns are indistinguishable in translation. Recorded, unmarked.
• @1267C ⭐ · Cant. IV, 16b — [var: Vulg. has the singular, "the fruit of his apples"], and the plate diverges from the received text AND from itself at the same site. Migne's lemma prints the plural comedat FRUCTUS pomorum suorum, and his own gloss eight lines below re-quotes the singular (« ut comedat fructum pomorum suorum, »). The two readings are rendered as printed in their two places — "the fruits of his apples" in the lemma, "the fruit of his apples" in the gloss. The marker content is written as a phrase, per the mirror-test finding. dilectus is lowercase here and is rendered lowercase, mirroring Migne's varying capitalization. See Standing Item 2 — this one earned its marker only because the divergence is external as well as internal.

Caput V — 1268A–1281D

• @1268A · lemma 1 · Cant. V, 1 — Veni in hortum meum, soror mea sponsa; messui myrrham meam cum aromatibus meis. = Vulgate. ⭐ messui rendered "I have reaped", not "I have gathered", because the gloss is built on the harvest image: Quadam falce mortis praescidi, "with a certain sickle of death I have cut them off." "Gathered" would leave the sickle explaining nothing. Re-quotations all held: « Veni in hortum meum. » · « Soror mea sponsa, » (twice) · « Messui myrrham meam, … » · « Cum aromatibus meis, ».
• @1268B — a guillemet opened and never closed, mirrored. « Messui myrrham meam, opens and Migne never closes it before the next opening « Cum aromatibus meis, ». 31 open / 29 close across the chunk, and the imbalance is his.
• @1268B — abcessisse for abscessisse. Attested medieval orthography; translated normally, no italics, no marker. Listed only so a later sweep does not churn it.
• @1268C · lemma · Cant. V, 1b — Comedi favum cum melle meo; bibi vinum meum cum lacte meo. / Comedite, amici, et bibite, et inebriamini charissimi. = Vulgate apart from charissimi (orthography) and Migne's DROPPED COMMA before charissimi, which rides as printed: the English reads "and be inebriated most dear ones" with no comma, and the gloss's re-quotation is punctuated the same way. ⭐ amici → "friends", charos → "dear", charissimi → "most dear", so the gloss's own three-step ladder (amicos … charos … charissimos) survives in English.
• @1268D · lemma 2 · Cant. V, 2 — Ego dormio, et cor meum vigilat. = Vulgate. Re-quoted « Ego dormio. » four times, all "I sleep".
• @1269A · Gen. XV, 13–16 — a long quotation, agreeing with the Vulgate throughout. Recorded because it is the longest scripture citation in these columns and a silent divergence here would have been costly.
• @1269B — the note (Gen. XXVIII, 12) is MISPLACED. It closes the sentence about Abraham (Gen. XV), and Gen. XXVIII, 12 is Jacob's ladder, which Migne narrates in the following sentence with no note. Kept verbatim, no correction proposed.
• @1269C — the note (ibid., 11) on the sun, moon and eleven stars. The dream is Gen. XXXVII, 9, not 11 — but the ibid. resolves to the preceding (Gen. XXXVII, 7), so the book is right and only the verse is off. Kept, and explicitly not proposed for correction: this ibid. and the two that follow it (Ibid., 8. and Ibid., 10., both of which ARE correct) are one anaphoric chain.
• @1269C — anima qua cordis sedes est. qua where quae is wanted ("the soul which is the seat of the heart"). Rendered as an ablative of place, "the soul wherein is the seat of the heart," which is what the printed case allows. Compare 1270B, where Foliot writes the construction the other way round and correctly: Dei sapientia, cui cor meum sedes est. Recorded, unmarked.
• @1269D ⭐–1270A — A SENTENCE WITH NO APODOSIS, THE SECOND OF THREE. Cui cum in corde vigilavit ad intelligentiam, qui sibi septem annorum futuram copiam; et septem annorum inediam declaravit. Two relatives and no principal verb: the limb that should say Joseph was the one who watched in heart for him is simply absent from the plate. Rendered ending where Migne ends, deliberately unmarked — [ed:] covers only what OUR digitization lost. See Standing Item 1.
• @1270B — the lemma split into two guillemet units, and NOT merged. « Cor » tamen « meum, » … ipsa « vigilat, ». The split is Migne's and is preserved as two units: "« The heart » however « that is mine, » … this same « watcheth »". Merging them would have read better and would have destroyed a guillemet pair.
• @1270B ⭐ · Cant. V, 2 — [var: Vulg. Aperi mihi, soror mea, amica mea, "Open to me, my sister, my friend"], AND IT IS THE MARKER THAT PROVES THE MECHANISM. Migne's lemma prints Vox dilecti mei pulsantis, aperi, soror mea, amica, columba mea, immaculata mea, dropping both the dative mihi and the possessive mea after amica. The missing mihi is not cosmetic: Foliot's whole gloss is that the Bridegroom asks to be admitted to himself — sibi postulat aperiri … « Aperi mihi, soror mea; » … habitaculum quaerenti in te — and he restores mihi in his very first re-quotation, four lines below the lemma he printed without it. ⚑ This is the same "work quoting itself" shape as 1176B, and here the marker went through cleanly, because the received phrase is NOT printed verbatim in the Latin twin: Migne's own re-quotation stops at soror mea, so the overlap was partial and the mirror test did not fire. That is the evidence that the blocker is the LENGTH OF THE OVERLAP, not self-quotation. See Standing Item 2.
• @1270C — the re-quotation reorders the vocatives, and the order is preserved. The lemma runs soror, amica, columba, immaculata; the gloss runs « soror mea, amica mea, immaculata mea, columba mea es. » — dove and undefiled transposed. Internal re-quotation variance, no second marker, per the Adjuro vos precedent.
• @1270C · lemma · Cant. V, 2b — Quia caput meum plenum est rore, et cincinni mei guttis noctium. = Vulgate. ⚠ Second unclosed guillemet of the chunk: « Quia caput meum plenum est rore. opens at 1271A and runs to the end of the sentence without a closing mark. Mirrored.
• @1270D — a four-line hexameter on dew, LINEATED because Migne lineates it. Ros etenim de nocte venit … — one English line per printed line, with the terminal full stop kept outside the closing italic asterisk exactly as the plate sets it, and the double blank line before the next anchor preserved. ⚠ The Latin is metrically rough and quanto distenditur illa tenore / Hanc sequitur is genuinely obscure; rendered literally rather than construed into smooth English, and flagged here rather than repaired.
• @1271B · lemma 3 · Cant. V, 3 — Exspoliavi me tunica mea, quomodo induar illa? lavi pedes meos, quomodo inquinabo illos? Two differences from the Vulgate, neither a reading: Exspoliavi for Expoliavi (orthography), and Migne's commas where the Vulgate sets colons, carried as printed. Both question marks are Migne's and both ride, and the gloss adds four more direct questions of its own, so the chunk carries 11 question marks against the Latin's 11 — verified by count, not by eye.
• @1271D · lemma 4 · Cant. V, 4 — Dilectus meus misit manum suam per foramen, et venter meus intremuit ad tactum ejus. = Vulgate. ⭐ venter rendered "belly", not the familiar "bowels", because the gloss is anatomical and literal: Venter namque osse caret, firmitate nulla participat, quod intromittitur hoc recipit, et decoctum reddit — a thing with no bone that receives food and renders it digested. "Bowels" would make Foliot's physiology unreadable. foramen = "the hole" for the same reason: Ambrose glosses it as oculus mentis, and Douay's "key hole" imports a lock the gloss does not have.
• @1272A · Psal. XLI, 3 — [var:] DECLINED, and it is a consistency question. Migne drops fortem (Sitivit anima mea ad Deum vivum). Declined: a dropped divine epithet in a proof-text whose whole point is the verb sitivit does not change the sense. ⚠ Recorded so the merge can overrule — but taking it here would require taking the next item too, and the two should stand or fall together. (Contrast 1209A, where the same verse prints fontem for fortem and the substitution, not an omission, takes a marker.)
• @1272A · Psal. CXVIII, 129 — [var:] DECLINED, and a wrong citation kept. Migne inserts a vocative Domine the psalm does not have: Mirabilia testimonia tua, Domine, ideo scrutata est ea anima mea. Declined on the same ground as the item above. ⚠ The citation is also wrong: Migne prints 119, and the verse is 129 — Ps. CXVIII, 119 is Praevaricantes reputavi omnes peccatores terrae, which has nothing to do with the quotation. Kept verbatim. ⛔ Do NOT enter this in citation-corrections.json without a plate check: a digit transposition is the most plausible reading, but "most plausible" is not evidence — and no Ibid. depends on it either way (the next note is a fresh Psal. XCII, 5), so nothing downstream is at risk from leaving it alone.
• @1272B — a third unclosed guillemet, mirrored. The final « Venter meus intremuit ad tactum ejus. is never closed; 8 open / 7 close.
• @1272C · lemma 5 · Cant. V, 5 — Surrexi, ut aperirem dilecto meo: manus meae stillaverunt myrrham. / Et digiti mei pleni myrrha probatissima. = Vulgate (Migne splits the verse into two lemmata — segmentation, not a variant, the same class as the curremus seam at 1178A/1180A). ⭐ probata / probatior / probatissima rendered "approved / more approved / most approved" rather than Douay's "choicest", because the gloss builds an explicit three-step comparison (cum uxore … cum vidua continente … cum virgine sancta permanente) that a superlative-only English cannot carry.
• @1273A · lemma 6 · Cant. V, 6 — Pessulum ostii mei aperui dilecto meo. = Vulgate. ⚠ The re-quotation INVERTS the word order — « Ostii mei aperui pessulum, » — and the inversion is preserved in English ("Of my door I opened the bolt"). Not harmonized: Migne prints two orders and the reader should see both.
• @1273A — [cj: discerptionis; read discretionis, "of discretion"]. Quo in loco virtutem « discerptionis » aperit: per digitos enim, discretio demonstratur. discerptio is a real Latin word ("a tearing to pieces"), so nothing is broken to wrap — and the English then says Foliot opened "the virtue of tearing apart," which is not a virtue and is not what the next six words of his own sentence say. discretio stands in the very next clause, and the whole paragraph is about measure and discernment. Additive: Migne's word keeps its English and the conjecture stands beside it.
• @1273A — quia post digiti tactu tangitur, id certa quadam ratione mentis intelligitur. Does not construe: post is stranded and the antecedent of id is unclear. Rendered as closely as the printed words allow. No conjecture offered — more than one repair is available (quod for quia, a lost quod after post), and guessing between them would be the silent smoothing the rule forbids.
• @1273B — possulum carried untranslated, and NOT wrapped in [sic:]. For pessulum, which Migne prints correctly in the lemma one paragraph above and again at 1273C. Not a word, so it is carried in italics in a sentence that keeps its verb and its sense on either side of it (the run is a bare object, it governs nothing, and ostii amovisse survives). No [sic:] because the type is carried, not glossed — the italic possulum IS the marking.
• @1274C · Matth. IV, 17, noted as Matth. V, 3 — the notes have slipped one quotation to the left. The quotation is Poenitentiam agite, appropinquavit enim regnum coelorum (Matth. IV, 17); Matth. V, 3 is Beati pauperes spiritu, which Migne quotes in the next sentence with no note at all. Kept verbatim. ⚠ Note that Migne prints an inline italic (Matth. IV, 17) inside the guillemets a few words earlier, correctly locating the Novissime … in Filio clause — so the plate carries the right reference and the wrong one within two lines. Not tagged [f:].
• @1274C–1274D — the Matth. VIII healing chain, off by one member throughout. Mundare is Matth. VIII, 3 ✓; confestim mundata est lepra ejus is noted ibid., 13 (it is VIII, 3); Vade, et sicut credidisti, fiat tibi is noted ibid., 14 (it is VIII, 13). All kept verbatim, and again no correction proposed: three consecutive ibid. notes are precisely the anaphoric shape that may never be touched.
• @1275C · Matth. VII, 8 — Omnis qui petit, accipit, dropping the Vulgate's enim. Connective only; no marker.
• @1275D ⭐ · Joan. XVII, 24 — [var: Vulg. ut ubi sum ego, et illi sint mecum]: two verses fused. Migne prints Volo, Pater, ut ubi ego sum, illic sit et minister meus — the close of Joan. XII, 26 fused onto the Volo, Pater opening of XVII, 24. This changes the sense and the argument: Foliot is proving that the elect are brought into the Father's glory, and the received XVII, 24 says exactly that, while the printed conflation narrows it to the single minister. ⚑ No claim is made in the marker about WHY the two texts have fused — the XII, 26 identification is recorded here in the crux, where it belongs.
• @1275D–1276A · lemmata 7–8 · Cant. V, 7–8 — Invenerunt me custodes qui circumeunt civitatem; percusserunt me, et vulneraverunt me, tulerunt pallium meum mihi custodes murorum. Adjuro vos, filiae Jerusalem, si inveneritis Dilectum meum, ut nuntietis ei quia amore langueo. = Vulgate, including the capital Dilectum, which is Migne's own (the Vulgate prints lowercase) and is mirrored. ⭐ custodes = "keepers" here and at every one of the ~10 occurrences in the gloss (custodes Ecclesiae, custodes murorum, custodes Ecclesiarum) and again at 1302B; circumeunt = "go about", with circumeuntes in the 1277A gloss → "going about", so the gloss visibly picks the lemma's verb back up. ✅ Reconciled at the merge against 1247A's vigiles qui custodiunt = "watchmen … keep": two different verses, both agreeing with the Vulgate, and both glossed differently by Foliot.
• @1276A — ei vestra obstet gratia — [cj:] DECLINED. The sentence is third-person about God throughout (Clamat … ad Deum omnis anima electa), and then a second-person plural possessive appears: "that your grace may withstand him." nostra is the obvious conjecture. Declined because the printed word is real, is grammatical, and has an intelligible reading as it stands: the sentence has just quoted a direct address to God (« Deus, in adjutorium meum intende ») and slides into continuing that address, which is a thing this author does elsewhere. A pronoun shift mid-sentence is odd, not false. Rendered as printed, with the plural kept plural. Named because a blind reader will stop at it, and if the lower bar is preferred the fix is one [cj:].
• @1276D — muro tuo inexpugnabili circumcinge nos, Domine: NOT scripture. This is the antiphon / collect formula (Circumcinge nos, Domine, muro tuo inexpugnabili), and Migne prints it without a locator. No [var:], and none was considered — a liturgical composition is not a scripture citation, and where he prints no locator a "received reading" would rest on our identification. The 1160C precedent.
• @1277B · lemmata 9–10 · Cant. V, 9–10 — Qualis est Dilectus tuus ex Dilecto, o pulcherrima mulierum? … / Dilectus meus candidus et rubicundus, electus ex millibus. = Vulgate except that Migne capitalizes all four Dilectus/Dilecto where the Vulgate prints them lowercase. Mirrored exactly ("thy Beloved from the Beloved"); capitalization is Migne's editorial hand, the parallel Latin column preserves it, so nothing is concealed. ⚠ Dilectus tuus ex Dilecto was deliberately NOT smoothed. It is opaque in English and opaque in Latin, and every modern version paraphrases it ("more than another beloved") — but Foliot treats the doubling as the passage's difficulty and answers it with the two natures (geminam in Christo … naturam), so a smoothed English would remove the thing he is explaining. Rendered literally, twice, identically.
• @1277B–1277D — the re-quotations, all held, including « O pulcherrima mulierum! » twice with Migne's exclamation marks, an abridged re-quotation stopping before ex Dilecto (carried as printed, no marker), and « Dilectus meus candidus, et rubicundus, » with a comma the lemma lacks.
• @1277C · Prov. XXXI, 30 and its [n: (ibid.)] — both agree with the Vulgate, and the ibid. correctly resolves the second half of the same verse. Recorded as a clean case, so a later sweep does not mistake it for the pattern above.
• @1277D — Migne prints a FULL STOP, not a question mark, on a direct question. Quo signo nobis elucidas … quem et tu diligis. Carried as printed, in company with the nunquid questions at 1153C–1158A. Do not "fix" it.
• @1277D — duodenario … numero. Foliot says the description is comprehended "under the number twelve" and then names two things. Rendered as printed: he is counting the twelve members of the whole descriptio that runs from V, 10 to V, 16, not the two natures.
• @1278C · lemma 11 · Cant. V, 11 — Caput ejus aurum optimum. Comae ejus sicut elatae palmarum, nigrae quasi corvus. = Vulgate (the Vulgate sets a semicolon where Migne sets a full stop after optimum; carried as printed). ⭐ nigrae → "black", the same English word as 1183C's Nigra sum, sed formosa. Here it is hair and not the Bride, and Foliot's own gloss makes the echo the point: quia ii etsi virtute proficiant, nigros se ut corvos fore protestantur. A different English word here (swarthy, dark) would hide the connection from exactly the reader who would notice it. ⭐ elatae palmarum → "the topmost boughs of palm trees", not Douay's "branches", because Foliot glosses the word: Elatae palmarum partes in palma superiores dicuntur. "Branches" would leave that sentence saying nothing.
• @1279A · lemma 12 · Cant. V, 12 — Oculi ejus sicut columbae super rivulos aquarum, quae lacte sunt lotae, et resident juxta fluenta plenissima. = Vulgate. ⭐ fluenta plenissima → "the fullest streams" rather than "plentiful streams", because the gloss puns the superlative straight across: juxta antiquorum Patrum scripta plenissima. Keeping one English word in both places is what makes the gloss legible.
• @1279A — colombae carried in italics INSIDE a guillemet quotation. sunt « sicut colombae quae lacte sunt lotae. » For columbae, printed correctly in the lemma four lines above. ⚠ It takes no second italic marker — nested italics are not expressible. Sense is unaffected: the doves are named in the lemma immediately above and the gloss's subject (viri spirituales) is intact.
• @1279B · lemma 13 · Cant. V, 13 — Genae illius sicut areolae aromatum consitae a pigmentariis. = Vulgate apart from Migne dropping the comma before consitae. ⚠ The gloss quotes the lemma in three separate guillemet fragments OUT OF ORDER — « Areolae » … « a pigmentariis, » … « consitae sunt, » — with the verb turned from participle to finite. All three preserved as three units; the English word order follows Migne's, not English's. areolae aromatum → "beds of aromatical spices", the rendering held again at 1282A.
• @1279C · lemma · Cant. V, 13b — Labia ejus lilia distillantia myrrham primam. = Vulgate. The gloss re-quotes with stillantia for distillantia; both render "dropping" — there is no English exponent for the prefix here. Recorded, no marker.
• @1279D · lemma 14 · Cant. V, 14 — Manus illius tornatiles aureae, plenae hyacinthis. = Vulgate apart from the Vulgate's comma after tornatiles. tornatilis = "turned", and the lathe (tornus) gloss runs on into 1289D, which is what fixes the word across the chapter boundary.
• @1279D — expreverunt carried in italics, and the conjecture deliberately kept OUT of the text. et illis peccantibus … id expreverunt, et se Deo … devoventes. Not a Latin word; almost certainly spreverunt ("they spurned it"), which is what the sentence requires — the good angels rejected the rebels' course. Unlike the discerptionis case at 1273A, this is a NON-word, so [cj:] is unavailable: it requires the printed word to be real. ⚠ The quarantined run is the clause's only verb, but the clause's force survives in the two participial limbs on either side (peccantibus … volentibus before, devoventes … calcaverunt after), so the English does not assert anything false. Flagged rather than repaired.
• @1268A–1279D — the apple sense of mala does not occur in this chapter at all. The Song's mala belong to chapters II, IV and VIII. The malus, -a, -um forms here (malignos spiritus, ad malum … inclinare, male habentes, qui malignus est) are all evil, decided from case and governing verb each time. A negative result is worth as much as a positive one.

Caput VI — 1280A–1288B

• @1280C · lemma · Cant. V, 14b — Venter ejus eburneus, distinctus sapphiris. = Vulgate. Re-quotations held: « eburneus, » and Quae quidem Sapphiris distincta est ("set with sapphires" both times). ⚠ Gender: Quae quidem … distincta est … adornatum at 1280D–1281A is a feminine relative and participle for the masculine venter, then a neuter participle in the same sentence. No English exponent; English untouched, logged so a later sweep does not churn it.
• @1280C · Isa. LIII, 4 · Psal. XLIV, 3 · @1280D · Psal. XV, 10 — checked word by word against the Vulgate; all agree apart from ordinary truncation at the end of the quoted run and orthography (foenum, charitas, seculi). No markers.
• @1281A · lemma 15 · Cant. V, 15 — Crura illius columnae marmoreae, quae fundatae sunt super bases aureas. = Vulgate. The gloss's basibus aureis → "bases of gold", matching.
• @1281A · lemma · Cant. V, 15b — Species ejus ut Libani, electus ut cedri. = Vulgate. ⭐ species = "fairness", on 1210B's precedent (Haec forma, haec species, haec pulchritudo), governing also the gloss Haec Deo species attribuitur, the habitationes speciosae two lines above, and Ps. XLIV, 3 Speciosus forma prae filiis hominum → "Fair in form above the sons of men." This is the decision that frees "comely" for decora without stranding speciosa — see the conventions. electus = "chosen" throughout, never Douay's "excellent" or "bright", because Foliot glosses the election itself at 1284C.
• @1281C ⭐ · lemma 16 · Cant. V, 16 — the masculine twin of the amica ruling, and both words stand in one line. Guttur illius suavissimum, et totus desiderabilis: talis est Dilectus meus, et ipse est amicus meus, filiae Jerusalem. = Vulgate; capital Dilectus mirrored. Rendered "such is my Beloved, and he is my friend", so amicus takes the same English as the ratified amica and the two Song terms stay visibly distinct in the one sentence where they meet. Re-quotation held verbatim at 1281D, capital again. guttur = "throat", the same at 1216A and 1293D.
• @1281C — Quid hoc dulcius? quid suavius esse potest? — a DECLARED collision. Migne stacks two comparatives of the same idea in one breath, and both Latin words are fixed to "sweet" elsewhere in the work. Rendered "What is sweeter than this? what can be more delightful?" Declared rather than left silent: English has no second word at the same register, and flattening both to "sweeter" would make Foliot's doublet look like a stammer.
• @1281C · Exod. XXI, 24 — [var: Vulg. reddet animam pro anima, without a subject "man"]. Migne prints Reddet homo animam pro anima, supplying a subject the Vulgate does not have. The added noun changes who is bound by the lex talionis, and Foliot's whole antithesis is Old-Testament homo against New-Testament si quis.
• @1281C ⭐ · Matth. V, 39 — [var: Vulg. in dexteram maxillam tuam, "on thy right cheek"]. Migne prints Si quis percusserit te in maxilla una, praebe ei maxillam alteram. ⚑ Material in THIS work specifically: Foliot builds a sustained argument on dextera against laeva/sinistra, and the plate has removed a dextera from a proof-text.
• @1281C · Matth. V, 40 — dimitte ei et pallium for remitte ei et pallium. Declined: a prefix-synonym; both mean "let go to him," and no gloss touches it.
• @1281D · lemma 17 · Cant. VI, 1 — Quo abiit Dilectus tuus, o pulcherrima mulierum? Quo declinavit Dilectus tuus, et quaeremus eum tecum? = Vulgate. pulcherrima = "most beautiful"; "of women" (genitive) keeps it distinct from 1194A's in mulieribus "among women". Re-quotations held, all three with the capital Dilectus Migne prints.
• @1282A — qui haec pulcherrima mulierum est. Masculine qui with a feminine referent (quae expected). Rendered "since she is this most beautiful of women." Latin gender has no English exponent; nothing to emend, nothing to mark.
• @1282A · lemma 1 · Cant. VI, 1(2) — Dilectus meus descendit in hortum suum ad areolam aromatum, ut pascatur in hortis, et lilia colligat. = Vulgate. ⭐ areola aromatum = "bed of aromatical spices", following 1279B — the diminutive is NOT argued by Foliot here, so it is not forced into "little bed"; contrast lectulus and virgula, which he does argue. Re-quotations held: in areola aromatum · hujus aromaticae areolae · aromatum areola · in his pascatur hortis · haec vere lilia … colligat.
• @1282B — Abel's clause, declined. virgo pariter et martyr existens, sacerdotem et hostiam in se sacerdos pariter et hostia, confitens virgo pariter et martyr, hostia et sacerdos occubuit repeats virgo/martyr and sacerdos/hostia twice each in twenty words, with the participle confitens stranded between them. Rendered clause for clause as printed, in Migne's order. No emendation, no marker — nothing here is a non-word or a false assertion; it is just rhetoric the plate may have doubled. Flagged so the blind read looks.
• @1282C · Deut. XVIII, 18 — [var: Vulg. suscitabo eis de medio fratrum suorum, "unto them, from the midst of their brethren"]. Migne prints Prophetam suscitabo vobis de medio fratrum vestrorum. The addressee changes from Israel-in-the-third-person to Israel addressed directly — which is what lets Foliot introduce it as Moysen … ait. ⭐ This is also the third witness the merge asked for on 1258C: the verse stands here in the first person (suscitabo … ego ultor existam), agreeing with 1261A and against 1258C's third-person suscitabit … ipse ultor existet. Two first-person printings against one third confirms 1258C as a single-site plate divergence. Do not "correct" it.
• @1283A · lemma 2 · Cant. VI, 2(3) — Ego Dilecto meo, et Dilectus meus mihi, qui pascitur inter lilia. = Vulgate; both Dilectus capitalized here, against 1294A's lowercase re-quotation of the same formula. Mirrored on both sides.
• @1283A · lemma 3 · Cant. VI, 3(4) — Pulchra es, amica mea, suavis et decora sicut Jerusalem: terribilis ut castrorum acies ordinata. = Vulgate. ⭐ decora = "comely", which is Douay's word, 1194C's precedent inside this work, and what the gloss here depends on (unde patientiae decorem obtinet at 1283B, and again at 1292D). ⭐ castrorum acies ordinata = "an ordered line of battle of the camp." Douay's "an army set in array" was rejected because Foliot DEFINES the phrase at 1283B–C (Acies enim ordinata castrorum est, cum qui pugnaturi sunt … in castris) and glosses castra again at 1288D (choros castrorum → "the companies of the camps"). An English that drops the camp leaves two columns of gloss expounding a word not on the page. suavis = "sweet".
• @1283B — « terribilis est ut castrorum acies ordinata. », where his own lemma four lines above does not print est. Migne against himself: logged, no [var:] — the witness list has no internal term. (Standing Item 2; a supplied copula does not earn a marker on materiality.)
• @1283B · Psal. LXXXV, 5 — checked word by word; agrees.
• @1283C · lemma 4 · Cant. VI, 4(5) — Averte oculos tuos a me; quia ipsi me avolare fecerunt. = Vulgate. Re-quotations held.
• @1283D ⛔ — [cj: velle; read nolle, "not to will"]: A LOST NEGATION. Migne prints … nec in cor hominis ascendit, ante communem hominum resurrectionem, et cum suis spiritibus coadunationem, velle cuiquam ostendere. Rendered literally the clause says the Bridegroom WILLS to show his glory to someone before the general resurrection — which is the opposite of the whole paragraph, whose point is that the eyes made him fly away and that he remains incomprehensibilis adhuc homini. velle is a real word, so nothing is broken to wrap, and incoherence is not a defence where the word that went wrong is a negation. Taken as to will [cj: velle; read nolle, "not to will"]. ⚠ The alternative reading — that a negative is carried over by ellipsis from the three necs — is possible, and is why the gloss is additive rather than a substitution. A plate read would settle it; a vote would not.
• @1284A · lemma 8 · Cant. VI, 8 — Una est columba mea, perfecta mea, una est matris suae, electa genitrici suae. = Vulgate. ⭐ mater and genitrix must stay two words, because Foliot ARGUES the distinction at 1284C: Haec vero genitrix, eadem quae et ejus mater est: diversitate tamen verborum, diversa denotat. Fixed as "mother" and "her that bore her" (Douay's phrase), so the sentence that says the two words denote different things is readable. ⭐ racemus = "berry" against botrus = "cluster" — ex multis granis acervus unus, ex multis racemis botrus unus, "out of many berries one cluster", which is what the argument needs (many make one) and keeps botrus free for 1292D's ubera tua botris. Re-quotations held, the last with ad modum genitricis suae glossed "after the manner of her that bore her."
• @1284A — "sixty" and "eighty" spelled plainly, holding 1250A's ruling.
• @1284C · lemma · Cant. VI, 8b — Viderunt eam filiae, et beatissimam praedicaverunt: reginae et concubinae, et laudaverunt eam. = Vulgate. ⚠ Migne's gloss expands filiae to filiae Sion at 1284D and again at 1285A — internal expansion, not a Vulgate divergence; logged, no marker.
• @1285A — « reginae, » qui summus in Ecclesia ordo. Masculine relative for the feminine quoted noun, because the antecedent is the ordo. No English exponent; English untouched.
• @1285A · lemma 9 · Cant. VI, 9 — Quae est ista quae progreditur quasi aurora consurgens, pulchra ut luna, electa ut sol, terribilis ut castrorum acies ordinata? = Vulgate.
• @1285B ⛔ — [sic: progeditur], and it is a self-divergence as well as a defect. Migne's own re-quotation eleven lines below prints « Quae est ista quae progeditur quasi aurora consurgens? » for the lemma's progreditur. progeditur is not a Latin word (there is no progedior), so it is carried. ⚠ The quarantined run is the clause's only verb, so the English supplies the sense beside the marker — "Who is she that goeth forth [sic: progeditur*] as the dawn rising?" — rather than leaving a hole the size of the predicate.
• @1285B ⭐ — aurora: the etymology IS the gloss, so the Latin stays visible. Aurora dicitur, quasi hora aurea is rendered "It is called aurora, as it were hora aurea, the golden hour," and aurora is "the dawn" everywhere else. Without the two Latin words on the page the pun — reinforced three words later by auri more rutilans — explains nothing.
• @1285B — capitalization mirrored INSIDE one sentence. quae Solem praeveniens solem affuturum denunciat → "which, going before the Sun, announces that the sun is about to come." Migne's own inconsistency, reproduced, not normalized.
• @1285C · Joan. I, 11–12 — checked word by word; agrees. Re-quotation « Pulchra ut luna. » held.
• @1286A · Matth. XXV, 34 — [var: Vulg. possidete paratum vobis regnum a constitutione mundi, "from the foundation of the world"]. Migne prints possidete regnum quod vobis paratum est ab origine mundi.
• @1286B — [cj: lux carnis; read lex carnis, "the law of the flesh"]. Migne prints et quia lux carnis legem mentis percutit, one clause after Cum ergo lex peccati cum lege mentis contendit — the Romans VII antithesis, stated correctly, then restated with lex become lux. Rendered literally the page says the flesh has a LIGHT that strikes the law of the mind, in a paragraph whose entire scheme is light = faith against darkness = flesh (fides lux est, nox infidelitas, eight lines below). The type is not broken, so [sic:] cannot reach it, and the faithful English asserts something Foliot did not. Single competing conjecture, so the call was taken rather than declined.
• @1286C · lemma 10 · Cant. VI, 10 — Descendi in hortum nucum, ut viderem poma convallium, et inspicerem si floruisset vinea, et germinassent mala punica. = Vulgate. ⛔ This is the mala lemma, and poma and mala punica sit inside it — apples and pomegranates in one line, kept apart throughout the gloss (ut videat poma convallium · de malis … punicis). At 1287B Foliot reproduces 1265B's contrast exactly: nuts = the hearts of the saints, poma = the confessors who are NOT made equal to the martyrs, mala punica = those who bud into suffering for Christ.
• @1287B — Poma enim dicuntur fructus omnes qui molli teguntur cortice = APPLES, and it must be. This is Foliot's own definition of the word: rendering it "fruits" would make the sentence tautologous and break the contrast with the pomegranates two lines later.
• @1286C · Isa. LV, 9 and @1288A · Luc. II, 10–11 — checked word by word; agree.
• @1286D — venti fiant et irruunt in domum ejus — [cj:] DECLINED. The Matt. VII, 25 echo has flaverunt venti, so fiant for flant is a plausible one-letter slip. Declined because fiant renders perfectly well ("the winds arise") and asserts nothing false. Logged so the blind read does not treat it as unseen.
• @1287A · II Cor. IV, 7 — [var: Vulg. habemus autem thesaurum istum, without "hidden"]. Migne prints Habemus hunc thesaurum absconditum in vasis fictilibus. The Vulgate has no "hidden," and Foliot's gloss depends on it (the kernel interius, the shell exterius).
• @1287A · Rom. VIII, 38 — [var: Vulg. Certus sum enim quia, "I am sure that"]. Migne prints Scio quod, with the intervening members compressed. All five printed negatives carried (neque × 4 + nec ulla alia creatura). ⚑ The same verse takes the same call at 1244B; the two entries agree, deliberately.
• @1287C · lemma 11 · Cant. VI, 11 — Nescivi: anima mea conturbavit me propter quadrigas Aminadab. = Vulgate. Negation carried: Nescivi → "I knew not." Re-quotations held.
• @1288A — the Aminadab etymology. « Spontaneus populi mei » → « The willing one of my people, », and Foliot's five sponte in one sentence are all "willingly," so the name-gloss keeps its force.
• @1288B · lemma 12 · Cant. VI, 12 — Revertere, revertere, Sulamitis: revertere, revertere, ut intueamur te. = Vulgate. ⭐ The gloss COUNTS the imperatives (quater reverti monetur, "four times") and the count in the English matches the four Revertere the lemma prints. Sulamitis = "Sulamitess", held across the chapter boundary.

Caput VII — 1288C–1296C

• @1288C · lemma 1 · Cant. VII, 1 — Quid videbis in Sulamite, nisi choros castrorum? = Vulgate. nisi carried ("but") in both the lemma and the verbatim re-quotation at 1288D. castra = "camp(s)" here and in acies ordinata castrorum at 1283A/1285A, so Foliot's one image reads as one image.
• @1288D · lemma 2 · Cant. VII, 1b — Quam pulchri sunt gressus tui in calceamentis, filia Principis! = Vulgate apart from Migne's capital Principis, mirrored. ⚠ The ! is printed on the lemma, again on the full re-quotation at 1289A, and a THIRD time on the split re-quotation « quam pulchri » nunc « sunt gressus tui! ». All three carried; none supplied, none deleted. English ! count = Latin ! count = 3.
• @1289B · lemma · Cant. VII, 1c — Juncturae femorum tuorum sicut monilia, quae fabricata sunt manu artificis. = Vulgate. ⭐ monilia kept in ITALIC LATIN, on the murenulae precedent already set inside this work at 1199B — Foliot's gloss IS the etymology, Monilia quidem a monendo dicta sunt, and it then runs for a dozen lines on what they admonish (monent, four times). "Necklaces are so called from admonishing" explains nothing in English; the Latin noun beside the English verb "admonish" explains everything. ⚠ Contrast 1197D, where monile is rendered "necklace" because no etymology is argued there. Declared here rather than left silent.
• @1289C — Femina vel femora. Foliot names both plurals of femur; both kept in italic Latin so the doublet survives, since English has one word.
• @1289C and @1289D — two internal divergences in the gloss, both Migne against himself, both logged and neither marked. « Junctura femorum tuorum sicut monilia. » (singular, for the lemma's plural Juncturae) and « manu artificis fabrefacta sunt, » (for the lemma's fabricata). Number and synonym; neither earns a marker on materiality.
• @1289D · lemma 3 · Cant. VII, 2 — Umbilicus tuus crater tornatilis, nunquam indigens poculis. = Vulgate. nunquam carried, and the gloss's answering double negative at 1290B–C — qui nec hic indiget poculis, nec in futuro aliquatenus indigebit — carries both. tornatilis = "turned", holding 1279D's rendering; the lathe gloss runs through both chunks and is what fixes the word.
• @1290A · Isa. XL, 6–7 and @1290A · Psal. LXXXIX, 6 — checked word by word; agree.
• @1290B · I Cor. XV, 44 — DECLINED. Migne prints seminatur corpus animale, surget spirituale, dropping the second corpus. An elided repeated noun is not a changed reading; nothing in the gloss turns on it.
• @1290C · lemma · Cant. VII, 2b — Venter tuus sicut acervus tritici, vallatus liliis. = Vulgate. ⚠ Internal divergence: the gloss at 1290D re-quotes acervus hic « vallatur liliis » — present passive for the lemma's participle. Logged, not marked.
• @1290D · lemmata 4 and 5 · Cant. VII, 3–4 — Duo ubera tua, sicut duo hinnuli gemelli capreae. 5. Collum tuum sicut turris eburnea. = Vulgate. ⚠ Migne numbers TWO lemmata "5." in this chunk — this one and Caput tuum ut Carmelus at 1291D. His numbering, reproduced, not renumbered. caprea = "roe"; ubera = "breasts" per the declared mammae/ubera flattening. Re-quotations held.
• @1291A · lemma · Cant. VII, 4 — Oculi tui sicut piscinae in Hesebon, quae sunt in porta filiae multitudinis. = Vulgate. Re-quotations held, including « cingulus moeroris, » twice.
• @1291B — se nil mali machinaturos = evil (genitive with nil; no fruit in the sentence).
• @1291C · lemma · Cant. VII, 4b — Nasus tuus sicut turris Libani, quae respicit contra Damascum. = Vulgate.
• @1291D — Naso deputantur Ecclesiae, quae … sentiunt, resumed three words later by masculine Hi in Libano.** Gender mismatch with no English exponent; rendered "those of the Church who perceive."
• @1291D — mala multa illata sunt Israeli = "evils" (neuter plural with multa, of Damascus and Cain).
• @1291D / @1292A · lemma 5 · Cant. VII, 5 — Caput tuum ut Carmelus: et comae capitis tui sicut purpura Regis vincta canalibus. = Vulgate apart from Migne's capital Regis, mirrored. The lemma closes one chunk and is expounded across the seam; its re-quotation at 1292B and the two bare re-quotations at 1292C are all held to the lemma rendering word for word.
• @1292A — pluviam septem annis clausam, and it is Foliot's own arithmetic. III Reg. XVIII has the drought at three years and six months. No [var:]: this is not a scripture quotation and Migne prints no locator, so a marker would rest on our identification rather than on his — the 1160C principle. Logged as a factual oddity for Wilson's eye, alongside the twelve loaves and nine thousand at 1176D.
• @1292C · lemma 6 · Cant. VII, 6 — Quam pulchra es et quam decora, charissima, in deliciis! = Vulgate (charissima is orthography). The ! is Migne's and is carried; it is the only one in the chunk, and the split re-quotation at 1292D correctly carries none. decora = "comely"; charus / charior / charissima = "dear / dearer / dearest", so the gloss's three-step climb survives.
• @1292C — Canales quidem sunt emissiones gratiarum Dei, quas modis pluribus in suos quos eligit, ipsosque martyrii gratia dignos efficit ipse et suscipit — a MISSING VERB, and how it is shown. The relative quas has no verb; efficit and suscipit both take ipsos, not quas. Rendered with a bracketed supply — "which in many manners he [sends] into his own whom he chooses" — using the corpus's existing convention for supplied English words rather than inventing a marker. No [ed:]: nothing is missing from OUR digitization. No [cj:]: the missing item is a whole verb, not a wrong word, so there is nothing to gloss additively.
• @1292D · lemma 7 · Cant. VII, 7 — Statura tua assimilata est palmae, et ubera tua botris. = Vulgate. botrus = "cluster", per 1284A. palma here is the palm tree, never a fruit.
• @1293A · lemma 8 · Cant. VII, 8 — Dixi: Ascendam in palmam, et apprehendam fructus ejus: et erunt ubera tua sicut botri vineae, et odor oris tui sicut malorum. = Vulgate. ⛔ malorum here is APPLES — odor governs, and the gloss's ut malorum odores, gratissimi at 1293C is apples too. Re-quotations held.
• @1293D · II Cor. II, 15 — checked word by word; agrees.
• @1293D · lemma 9 · Cant. VII, 9–10 — Guttur tuum sicut vinum optimum, dignum Dilecto meo ad potandum, labiisque et dentibus illius ad ruminandum. Ego Dilecto meo, et ad me conversio ejus. = Vulgate; Migne capitalizes both Dilecto, mirrored. guttur = "throat", holding 1281C.
• @1294A ⭐ — CAPITALIZATION DIVERGENCE INSIDE ONE CHUNK, mirrored on both sides. The lemma prints capital Dilecto twice; the gloss re-quotes « Ego dilecto meo, et ad me conversio ejus; » with a LOWERCASE d, and then eight lines later writes Ego Dilecto meo digna et chara sum with the capital again. The English follows each letter as printed — "my beloved" in the re-quotation, "my Beloved" in the lemma and in the paraphrase. This is the one member of the "Migne against himself" family that English CAN carry, which is why it needs no marker.
• @1294A — a further internal divergence, logged not marked. « labiis et dentibus ruminandum » drops the lemma's -que, illius and ad.
• @1294A — ruminare = "to chew upon" in the lemma and "to chew the cud" in the Levitical gloss, which is the one place the two senses have to be legible at once; the noun animal makes the shift the author's, not ours.
• @1294C · lemma 10 · Cant. VII, 11 — Veni, Dilecte mi, egrediamur in agrum. = Vulgate; capital Dilecte mirrored. Re-quotations held.
• @1294D ⭐ — campus "plain" against ager "field", and the etymology kept visible. The whole gloss at 1294D–1295A turns on the untilled campus becoming the tilled ager, and on the etymology Ager … ex hoc, quod agi coepit vomere — for which ager and agi are both kept in italic Latin beside the English "driven," on the same ground as monilia.
• @1294D · Gen. III, 17 — [var: Vulg. quia audisti vocem uxoris tuae, "because thou hast hearkened to the voice of thy wife"]. Migne opens the curse at Quia comedisti de ligno, dropping audisti vocem uxoris tuae, et. Eve disappears from the sentence that curses the ground — in a gloss about how the campus came to bear thorns.
• @1294D · Matth. XXVIII, 20 — DECLINED. Ecce ego vobiscum sum usque ad consummationem seculi, without omnibus diebus. A truncation at a natural break, of a kind Migne makes constantly in this work, with no argument resting on the dropped words.
• @1295B · lemma · Cant. VII, 11b — Commoremur in villis. = Vulgate. villa = "village", against oppidum = "town" and civitas = "city", all three of which Foliot ranks in one sentence here.
• @1295C · lemma 11 · Cant. VII, 12 — Mane surgamus ad vineas: videamus si floruit vinea, si flores fructus parturiunt. = Vulgate as far as Migne's lemma runs; expounded across the chunk seam. ✅ Reconciled at the merge: the lemma here and its re-quotation « Videamus si floruit vinea » at 1295D both read "let us see if the vineyard has flourished," and flores / fructus are "flowers" / "fruits" on both sides of the seam.

Caput VIII — 1295D–1304D

• @1295D · lemma · Cant. VII, 12 — Si floruerint mala punica. The verse runs si floruerunt mala punica; Migne prints the future-perfect floruerint, rendered "shall have flourished". The tense stands as printed. No [var:]: Migne prints no locator for the Song lemmata anywhere in the work (they are the text being expounded), and the divergence is one letter of tense with no argumentative weight — the gloss turns on the pomegranate, not on the tense. ⛔ mala punica is the FRUIT, as at 1264D and 1286C: In malis quidem punicis amor indicatur martyrii confirms it — exactly 1265B's contrast of pomegranates = martyrs against apples = confessors.
• @1296A · lemma · Cant. VII, 12b — Ibi dabo tibi ubera mea. = Vulgate. Re-quoted once at 1296B, held word for word.
• @1296B · lemma · Cant. VII, 13 — Mandragorae dederunt odorem. In portis nostris omnia poma: nova et vetera, Dilecte mi, servavi tibi. = Vulgate, including the punctuation. Migne capitalizes Dilecte mi, so the English reads "my Beloved". ⛔ omnia poma = APPLES, and the pomegranates sit ten lines above in the same chunk.
• @1296B — [cj: difficilem; read non difficilem, "not difficult"]: A DROPPED NEGATION. Migne prints Ad hoc quidem quod expeto, difficilem te invenire non ambigo, quia mandragoras in portis nostris odorem tui jam dedisse intelligo. Rendered literally, the Bride says she does not doubt that she will find him DIFFICULT — in a sentence whose quia-clause gives the reason he is NEAR (the mandrakes have already given his odour), and in a paragraph whose whole burden is that his coming is assured. Every word is real Latin and correctly set, so [sic:] has nothing to wrap; the English asserts something the passage denies. Marked additively: difficult [cj: difficilem; read non difficilem, "not difficult"]. ⚠ The competing construal — that ambigo with an accusative-and-infinitive here means "I hesitate to find thee difficult" — was considered and does not rescue the sentence either, so the call is not between two good readings and the decline-and-log clause does not apply.
• @1296D · lemmata 1–2 · Cant. VIII, 1–2 — Quis mihi det te fratrem meum sugentem ubera matris meae… et mustum malorum granatorum meorum. = Vulgate word for word, with ONE punctuation divergence: the Vulgate closes v. 1 with ? (it is a wish-question) and Migne prints a FULL STOP. Carried as printed; the English ends on a period. No [var:] — a mark is not a reading. Re-quotations all held.
• @1297D ⛔ — THE HOMOGRAPH COLLIDES INSIDE ONE ELEVEN-WORD CLAUSE, and the call is stated rather than smoothed. Migne prints, of the martyrs, qui sibi sic in amoris dulcedine juncti sunt, ut sibi mala junguntur in malis, et nullis a te potuerunt malis averti — immediately after « Et mustum malorum granatorum meorum, », which it glosses. Rendered "as apples are joined to one another among apples." The referent is plainly the pomegranates of the lemma (the grains packed inside the fruit), and "as the grains are joined in the pomegranates" is what Foliot means. ⚠ It is not what he WROTE. He wrote bare mala, and this work's own system — fixed at 1264D and 1286C and unbroken — is poma / mala = apples, mala punica / mala granata = pomegranates. Importing granata into the English would be a silent emendation, and it would conceal an elision that is itself evidence about how Foliot handled the word. ⛔ The third malis in the same clause is "evils" (nullis … malis, ablative with nullis). The clause therefore reads apples–apples–evils in eleven words. If a later reader overturns the first two, the third must not move with them.
• @1297D — amorem … qui tot modis formatus est, quod modis vina formari solent. The correlative to tot is quot; Migne prints quod. Rendered "formed in so many ways, as wines are wont to be formed", which is as close as the printed words allow. No [cj:]: the English asserts nothing false, only slightly loose. ⚠ Borderline — it IS a correlative, and a stricter reading of the rule would mark it. Recorded so the decision is visible.
• @1297C–D and @1297D — two re-quotations that are NOT the lemma. (1) « Ibi tu docebis me, » for the lemma's ibi me docebis — word order reversed and tu added; English has one rendering, so nothing is visible to the reader. (2) « in cubiculum genitricis meae, » — this phrase is not in the lemma at all: it is Cant. III, 4, imported into the exposition of Cant. VIII, 2 and expounded there (antiquum illud amoris habitaculum). Not a variant of anything printed above it, so no marker; but it is the second time in this chapter that Foliot glosses words he did not quote.
• @1297A · Luc. I, 35 — Ideoque quod nascetur for the Vulgate's Ideoque et quod nascetur. A dropped et, no sense change, no gloss on it. Declined.
• @1297D ⭐ · lemmata 3–4 · Cant. VIII, 3–4 — THE REPRISE, and the two divergences got opposite treatments ON PURPOSE. Laeva ejus sub capite meo; et dextera illius amplexabitur me. 4. Adjuro vos, filiae Jerusalem, ne suscitetis, neque evigilare faciatis Dilectam, donec ipsa velit. Foliot's own gloss here is one sentence long (Haec quia jam dicta sunt, non repetimus), so the lemma carries the whole weight, and both verses were checked against their earlier englishings at 1221C and 1248B before rendering.
• The ; for 1221C's , after capite meo — the English differs; the semicolon is carried.
• dextera ILLIUS for 1221C's dextera EJUS — the English does NOT differ, and this is a decision, not an oversight. Both are simply "his" in English; there is no exponent for the distinction between the two demonstratives, and manufacturing one ("that one's right hand") would invent a distinction of sense the Latin does not make. Two standing precedents govern: a Latin word with no English exponent is logged as a crux and the English is left alone, and the Adjuro vos conjunctions took no marker for exactly the same reason. No [var:] — and a [var:] was impossible here regardless, because the divergence is Migne against Migne. ⚑ Note that against the Vulgate both printings agree: dextera illius is in fact the Vulgate's own reading at Cant. VIII, 3, so it is 1221C's ejus, not this one's illius, that is the odd printing.
• The Adjuro vos verse, sixth printing — donec, Dilectam capitalized, adjuration clause absent. The capital IS mirrored ("nor make the Beloved to awake"), against 1223A's and 1248B's "the beloved"; everything else is identical to the earlier English, deliberately. Two printed negations, two in the English, none supplied. ⚑ If the merge disagrees with either call, it should change the ejus/illius one, not the capital — English has an exponent for the first and none for the second.
• @1298A · lemma 5 · Cant. VIII, 5 — Quae est ista, quae ascendit de deserto, deliciis affluens, innixa super Dilectum suum? = Vulgate. ⚠ A DIFFERENT verse from 1248C's Quae est ista, quae ascendit per desertum? (Cant. III, 6). The two must not be harmonized: per desertum = "by the desert", de deserto = "from the desert". The frame is deliberately identical because Migne's frame is identical.
• @1298B — « potens in deliciis, », which is no verse of the Song. Foliot re-quotes his own lemma's deliciis affluens as potens in deliciis (Cant. VII, 6 has carissima in deliciis). Migne against himself again, and again no [var:] is possible. Rendered as printed, "mighty in delights"; the gloss then reverts to affluens and stays with it for fifteen lines.
• @1298C–D — the Gregory excerpt is unattributed by any [n:] and is carried as printed.
• @1298D · lemma · Cant. VIII, 5b — Sub arbore malo suscitavi te: ibi corrupta est mater tua: ibi violata est genitrix tua. = Vulgate apart from Migne's colons where the Vulgate punctuates ; and , — carried as printed, and the re-quotation at 1299A punctuates the same clause with SEMICOLONS, and that too is carried.
• @1298D ⛔–1299A — THE HARDEST mala SITE IN THE WORK, and it is a genuine ambiguity in Foliot, not a defect. arbor is feminine, so arbore malo cannot be adjectival ("evil tree") — it is malus, -i f., the apple tree, and Foliot says so explicitly: Dicunt enim vetitam illam arborem … malum fuisse. He then slides across the homograph six times in four lines:
• sub arbore malo (lemma, and the 1298D re-quote) = apple tree
• quod arbore malo contractum est = apple tree
• vetitam illam arborem … malum fuisse = apple tree
• Ab hoc quidem inobedientiae malo = evil — "from this evil of disobedience", with inobedientiae fixing it as the neuter noun
• seipsam sub malo, id est, sub peccato quod ex malo contractum est = apple, twice — rendered "under the apple … contracted from the apple", not "under the evil": Foliot's whole point in the sentence is that the tree of the lemma is the sin's origin, and the preceding clause has just identified the tree as an apple. ⚠ This is the one place in the work where the two senses could each be defended; the reading is recorded here rather than smoothed, and a reader who disagrees can see exactly which word was decided which way.
• @1298D — pulchrum oculis → "beautiful to the eyes", per the standing ruling; not "fair".
• @1299A — mater / genitrix kept visibly distinct, because the line ARGUES the distinction: matrem vocans generationem omnem … genitricem vero illam nominans — mother against she that bore thee. Collapsing them would leave that sentence expounding a difference that is not on the page.
• @1299B · lemma 6 · Cant. VIII, 6 — Pone me ut signaculum super cor tuum, ut signaculum super brachium tuum. = Vulgate. Re-quoted at 1299C, held.
• @1299C–1300B — Migne BREAKS the verse into three lemma-blocks, and adds a join. The Vulgate runs straight on into quia fortis est ut mors dilectio, dura sicut infernus aemulatio; Migne prints three blocks (1299C, 1299D, 1300B) and prints Et dura sicut infernus aemulatio for the Vulgate's unconnected dura. [var:] declined: this is SEGMENTATION, the same class as the curremus seam at 1178A/1180A, and the added Et is Migne's join between his own two lemma lines, not a reading. ⚠ Note also that the third block, Quia fortis est ut mors dilectio, is printed with no terminal punctuation at all — carried as printed, the English line ends bare.
• @1299C ⭐ — A THIRD SENTENCE WITH NO APODOSIS. This is what makes it a class. Haec enim si cohaerentia sunt, cor deliberans, et brachium id quod corde deliberatur exsequens. The protasis is complete, the two participial limbs are complete, and there is no main clause — while the next sentence starts a fresh conditional (Quae si ambo memoriale … tenent … cohibebunt) which does have its apodosis. Rendered ending where Migne ends, deliberately unmarked: the TEI, the Latin chunk and the plate all stop there, so nothing is lost by us. ⭐ Pairs with 1159D and 1269D — three sentences, three parts of the work, three readers who could not see each other. See Standing Item 1.
• @1299D · Prov. VI, 16–19 — quoted inside the gloss; verbatim Vulgate, so no marker. ⛔ pedes veloces ad currendum in malum = evil — the homograph a third time in the same chunk, this time inside a proof-text.
• @1300B — [sic: marytrum], with the sense kept outside the quarantine. se quasi sequaces marytrum et ipsorum imitatores praeferebant. A non-word (metathesis for martyrum). Rendered "followers of the martyrs [sic: marytrum]" — the marker isolates the defective type and not the force it carries, and marytrum is the clause's only object noun; deleting it would leave "followers of the — and imitators of them." The same sentence prints martyres correctly twice, so the type is certainly a slip.
• @1300D — illam in the lemma, illum in the re-quotation. The lemma prints nec flumina obruent illam (feminine, agreeing with charitatem); eight lines below Migne re-quotes « flumina non obruent illum, » in the masculine, of the same charity. English has "it" for both and no exponent for the gender, so the English cannot show it. Recorded. ⚠ Note also that the lemma's nec becomes non in the re-quotation — both are negations, both carried ("neither shall … overwhelm" and "shall not overwhelm").
• @1301A — the lemma's . against the re-quotation's ?. The lemma prints Quid faciemus sorori nostrae in die quando alloquenda est. with a full stop (the Vulgate has ?); the gloss re-quotes « Quid faciemus sorori nostrae, in die quando alloquenda est? » with a question mark and an added comma. Both carried exactly as printed — the English lemma ends on a period and the English re-quotation on a question mark. No [var:] on the Vulgate's mark: a mark is not a reading.
• @1301A · Rom. XI, 25–26 — DECLINED. Migne prints Cum plenitudo Gentium intraverit, tunc omnis Israel salvus fiet for the Vulgate's donec plenitudo gentium intraret, et sic omnis Israel salvus fieret. A recast of tense and connective that says the same thing, of a verse universally quoted in this form; the gloss turns on the fact, not the wording.
• @1301A — an unclosed guillemet, mirrored. Migne opens « before Cum plenitudo Gentium intraverit and never closes it; the quotation runs into the note and then straight on into Foliot's own prose (Hujus itaque synagogae reminiscitur…), which is unmistakably not the Apostle. The English opens in the same place and likewise does not close: 7 open / 6 close on both sides. Do not "fix" this.
• @1301B — Hanc … sanctum apostolum … convertendam: a missing function word, shown by bracketing. Hanc itaque sanctum apostolum in novissimis convertendam, et Eliam, et Henoch ad hos convertendos a Christo reservatos praesciens… The gerundive convertendam agrees with Hanc, but sanctum apostolum is a bare accusative with no preposition and no verb to govern it; the sense requires an agent (per sanctum apostolum, or a nominative sanctus apostolus). Rendered "to be converted in the last times [by] the holy apostle", with the supplied preposition in square brackets per the corpus convention, so the reader can see it is ours. No [cj:] — the defect is a missing function word, not a real word asserting something false, and the bracketing already exposes it.
• @1301C and @1302A — two column anchors printed INSIDE italics, on their own lines. Migne's [1301C] and [1302A] stand alone as [1301C] and [1302A] in the Latin chunk, while [1301A] stands bare. All three reproduced exactly as they stand, italics included. This looks like an artefact of the lemma-italics run rather than anything on the plate, but it is in the Latin twin and marker positions are matched positionally, so it is not ours to normalize. (The same shape occurs at 1272C.)
• @1302B · lemma 11 · Cant. VIII, 11 — Vinea fuit pacifico in ea, quae habet populos: tradidit eam custodibus. = Vulgate apart from Migne's comma after in ea, which the Vulgate does not print and which his own re-quotation eleven lines below (« In ea quae habet populos, ») also does not print. Both carried as printed. ⭐ pacifico rendered "the peaceable one" and held so through all four re-quotations, because the gloss identifies him — pacifico, id est, Jesu Christo Domino nostro, qui pacificavit omnia — and a rendering that lost the root would leave pacificavit explaining nothing. ⭐ custodibus = "keepers", per the ruling at 1275D, and this chapter needs it: the gloss runs custodienda commissa est and qui custodiunt fructus ejus within the same paragraph.
• @1302D · lemma 12 · Cant. VIII, 11b–12 — Vir affert pro fructu ejus mille argenteos. 12. Vinea mea coram me est. Mille tui pacifici, et ducenti his qui custodiunt fructus ejus. = Vulgate.
• @1302D · Psal. I, 1 — [var: Vulg. opens the verse with a beatitude, "Blessed is the man who hath not walked…"]. Migne introduces the verse as Vir ille bonus, cujus meminit David, dicens, « Vir qui non abiit in consilio impiorum… » under the note [n: (Psal. I, 1, 2)]. The Psalm opens Beatus vir qui non abiit, and the dropped word converts a beatitude into a bare definition of "the man." ⭐ Material, and the argument runs on it: Foliot has just asked Quis iste vir est? and answers by identifying the vir of Cant. VIII, 11 with the vir of Psalm 1. With Beatus present the proof-text is a blessing; without it, it is a description of the man the question asked after — and the reader who knows the Psalm will supply "Blessed" silently and never see what Migne's page actually offers as the proof. Content written as a phrase in English, per the mirror-test finding. The rest of the citation is verbatim Vulgate, including the three nons, all carried. ⚠ Contrast 1198C, where the same psalm is compressed and the marker is declined because nothing there turns on the word.
• @1303B — Foliot resolves his own lemma's ambiguity, in the opposite direction to the English. The lemma's Mille tui pacifici is either a vocative singular or a nominative plural; the gloss re-quotes it as « per pacificos mille » — an accusative plural, "the thousand peaceable ones." The lemma is rendered "A thousand are thine, O peaceable one" (the Douay construal, and the one tui supports) and the gloss is rendered as HE printed it. Migne against himself once more; recorded, not reconciled.
• @1303C ⛔ — AN ARITHMETIC SELF-DIVERGENCE, and the best plate-read candidate in the work. The lemma prints ducenti, two hundred, and the gloss's first limb correctly renders it ("by the two hundred, who keep the fruits thereof") — but the very next sentence divides them as Hi namque perfecti illi sunt … and then « Centum vero qui supersunt », "the HUNDRED that remain." Two hundred split into a first group and a hundred remaining. Rendered as printed; the conjecture is that Centum stands for a second hundred, or that the plate has dropped a C. No marker taken: this is a number, not a negation or a correlative, so no pattern reaches it, and [sic:] cannot wrap a word that is perfectly good type. See Standing Item 3.
• @1303C · lemma 13 · Cant. VIII, 13 — Quae habitas in hortis, amici auscultant: fac me audire vocem tuam. = Vulgate. ⭐ amici = "the friends" — the amica/amicus ruling arriving at its last site in the work, and here it is doubly right: the gloss glosses the word with fidelis amicitiae at 1304A, so the friendship sense is what Foliot expounds. habitas, tuam are singular — "thou"**, throughout the lemma and every re-quotation.
• @1303D–1304A — a re-quotation printed BARE, and left bare. The lemma is re-quoted three times, twice inside guillemets (« Quae habitas in hortis, fac me audire vocem tuam. » and the split pair « Amici » … « auscultant; ») and once with no guillemets at all. Guillemets are 1:1 and none is supplied.
• @1304A · lemma 14 · Cant. VIII, 14 — Fuge Dilecte mi, et assimilare capreae hinnuloque cervorum super montes aromatum. = Vulgate except that the Vulgate punctuates Fuge, dilecte mi and Migne prints Fuge Dilecte mi — no comma, and a capital D. Both are carried: the English lemma reads "Flee my Beloved" with no comma, while both gloss re-quotations (1304A, 1304C) print the comma and read "« Flee, my Beloved. »". No [var:] — a comma is punctuation and the capital is the standing mirroring convention, not a reading. caprea = "roe", never goat; hinnulo cervorum = "the fawn of the harts", held identical in the lemma and in both gloss re-quotations.
• @1304B · Matth. XXV, 34 (Venite, benedicti Patris mei…) carries no [n:] at all, where the Ecce Sponsus venit two lines above does. Verbatim Vulgate. Not a defect — Migne simply does not note it — and recorded only so a later indexer knows the omission is his.
• @1304C — « aromatum. » is a ONE-WORD guillemet unit. Migne quotes the single word he is about to gloss. Kept as its own unit, "« of spices. »", not merged into the surrounding sentence.
• @1304D ⛔ — THE WORK ENDS HERE, on conquiescit., with no colophon, no closing formula, no Explicit, no Amen and no editorial note. ✅ Verified against the TEI, not inferred from the chunk, precisely because a work's last chunk is where closing apparatus is most often dropped by a chunker: sources/pl/tei/11613.xml closes …certissima Sponsi sui promissione certificata conquiescit. </p></div2></div1></body></text></TEI> — no trailing <note>, no <trailer>, no <closer>, no further <pb/>. Nothing was lost between the source and the chunk, and 1304D is the last column of the work as Migne prints it. ⚠ What this does NOT establish is whether Migne's PLATE carries a colophon that Corpus Corporum's transcription dropped. Our files cannot answer that; a plate read of PL 202, 1304 can. Flagged rather than assumed, since the absence of any closing formula in a twelfth-century commentary on the Song is unusual enough to be worth one look. See Standing Item 3. — The final sentence ends the whole work on the Church at rest, awaiting the kiss of Cant. I, 1: Foliot closing the circle back on his own first lemma. Nothing is missing from the sense, which is itself evidence that the plate ends where we end.
• @1296D–1304D — the eleven [n:] in these columns were checked against the text they mark and ALL ELEVEN ARE CORRECT: Luc. I, 28, 35 · Matth. XXVIII, 10 · Prov. VI, 16, 17, 18, 19 · Rom. XI, 25, 26 · Matth. V, 3 et seq. · Matth. XX, 16 · Psal. I, 1, 2 · Psal. XI, 7 · Matth. XXV, 6. Nothing for citation-corrections.json.

Checked and deliberately NOT reported

So that a later reader knows these were looked at rather than missed.

• Column-band arithmetic. The bands are not four per column, and single-band jumps (1192D → 1194A, 1197A → 1197B, 1206C → 1206D → 1207A, 1253C → 1268A) are normal. Anchor sequence — not merely count — was diffed against the Latin twin at every chunk boundary in the work; no text is broken mid-word or mid-construction at any anchor.
• Marker parity, both directions, counted rather than eyeballed. Column anchors, [n:] notes, « », ?, ! and ## heads were counted against each Latin twin before and after writing, including the eleven deliberately mirrored guillemet imbalances (1180C, 1184B, 1205B, 1213C, 1217C, 1217A, 1227A, 1227C, 1240A, 1243A, 1244D, 1251A, 1259D, 1266C, 1268B, 1271A, 1272B, 1301A). Where an English chunk's guillemet count looks unbalanced against the body, the extra « is in the incipit: frontmatter field, copied verbatim from the Latin twin — a known measurement artefact, not a defect.
• No [f: …] tags anywhere in the work. This is a lemma-and-gloss commentary, not a florilegium. The four constructions that might have tempted a tag — the inline italic (Joan. XV, 5) at 1217B, (Hebr IV, 12) at 1230D, (Sirac III, 20) at 1265C and (Matth. IV, 17) at 1274C — are each reproduced verbatim, untranslated, in place, on the [n: …] principle.
• (cont.) on every repeated chapter head is our chunk division, carried untranslated. Migne's numbered section heads (## 5., ## 6., ## 10. and the rest) pass through unchanged.
• Medieval spellings — lethale, charitas, coelum, seculi, foenum, thymiamata, Bersabe, Haram, seraphin, abcessum, abcessisse, adolescentulae, exultabimus, contepescat, conviventia — translated normally, no italics, no crux.
• Paragraphing follows the plate exactly throughout, including where a paragraph break falls mid-argument.
• Em-dashes: none in the work. Migne prints none; commas, colons and full stops carry the sense everywhere it was tested.
• Negation fidelity was checked clause by clause across the work, and the count holds: every non / nec / neque / nisi / nullus / nihil / nunquam the Latin prints appears in the English, and none appears that the Latin does not print. The dense sites are named where they fall — six nons in one sentence at 1242A, five nullas at 1252A, five neques at 1231C, four in the gloss at 1248B, five negatives in Rom. VIII at 1287A, the three nons of Psalm 1 at 1302D, and the two apparent polarity defects at 1151A and 1241D, both of which were examined and both of which construe.

⭐ MIGNE'S OWN CONJECTURE NOTES — 72 of them, recovered from the plate 2026-08-18

Migne prints numbered textual notes at the foot of his pages — editorial conjectures, set as Forte word. — and our source does not contain a single one of them. Measured across all 5,276 PL TEI files and 498,774 <note> elements: zero. They were read off the plate for this work (archive.org patrologiaecurs139unkngoog, cross-checked against a second copy, patrologiaecurs114unkngoog) and are printed here in full.

Why they belong on this page. These are the editor's own suggestions about his own text, and they are exactly the apparatus that settles a crux. At 1254B this work prints quanam « Aaron » filiabus Sion consulitur egredi, which does not construe; our translator weighed two conjectures and rightly declined for want of a witness. Migne's witness was printed at the foot of that very page — Forte earum — and quanam earum construes. That entry is below.

How to read them. The column is the column of the MARKER in Migne's running text, which is not always the column the note is printed under. The English of this work renders what Migne printed, not what he conjectured — the conjecture is offered to the reader here, and never silently adopted into the text.

⚠ Two entries are not plain conjectures. (14) is recorded ILLEGIBLE: a reading was offered with a real letterform-and-sense case and was overruled at 900 dpi, where the letter shows no ascender though the f. beside it does. (64) is legible but truncated on Migne's own plate — both copies read F. spr with clean white space after, so recording it as illegible would claim a limitation in our reading that does not exist.

The sequence is complete, and its irregularities are Migne's: (12-13) is one note serving two markers · (59) is never printed · (62) is used twice, the second set as *(62\)**, which is to say he caught his own duplicate and starred it rather than renumbering.

⛔ PLATE SWEEP, 2026-08-29 — every [sic:] in this work read at Migne's plate

This work's [sic:] markers stood on columns nobody had read. All were read this session from the on-disk PL 202 scan at 300 dpi (500 dpi where a single letter decided it), full page width, with the volume's page map re-calibrated against the printed corner numbers before the read.

A [sic:] asserts a defect in MIGNE'S TYPE, and our Latin is not the printed page — it is Corpus Corporum's transcription of it. Where the plate disagrees with our file, the fault is ours and the accusation is withdrawn.

Withdrawn — Corpus Corporum's error, not Migne's

The Latin is corrected in data/tei-patches/11613.json and each marker is removed from the English.

Confirmed — Migne's own defect; the marker stands

Full ledger for all 23 works swept: data/plate-sweep/2026-08-29-sic-sweep.tsv.