These are the edition's working notes, published as they were written. They record, work by work, where Migne's plate is defective or the reading uncertain: the printed form, what the English does with it, and why. They are not a groomed apparatus. Where a note argues with itself, withdraws an earlier decision, or leaves a question open, that is the record of how the reading was actually reached, and it is left standing — a conjecture that is visible can be checked, and one that has been tidied away cannot.
The English never silently repairs the page. A construable but wrong reading is rendered literally and the conjecture is recorded here, never applied to the text. How the English is produced and checked →
Translated 2026-08-18, one translator, 3 chunks (2,686 Latin words → 3,957 English, 1.47×), on Opus. Structure: ARGUMENTUM, CAPUT PRIMUM (chunk 0); CAPUT II (chunk 1); CAPUT III, CAPUT IV (chunk 2).
Apparatus: 8 [var:] · 1 [sic:] · 1 [cj:] · 0 [d:] · 0 [ed:] · 0 [nt:] · 2 [n:] (exact [n:] parity with the Latin twin). Marker parity verified after writing: 28 column anchors (11 + 10 + 7, verbatim and in order, including the two anchors Migne sets hard against a verse address with no space — [0613D]VERS. 23.--, [0615A]VERS. 17.--, [0616B]VERS. 16.--), 126 italic spans (41 + 57 + 28 — exactly the 126 of the lemma census, marker contents excluded), 2 question marks (1 + 1 + 0), 5 ## heads. Zero guillemets in either language. Zero em-dashes in either language.
Conventions follow GLOSSA-NT-AGENT-BRIEF.md: VERS. n.-- passes through verbatim; the lemma is Englished; notes ride as ordinary [n:], contents untranslated. Pattern 17: every second person in this work is grammatically plural (vos, vobis, estis, servatis, facitis) — the Colossians as a body — so the English is "you" throughout and carries no thou. Checked; there is no singular address anywhere in the work.
§0 · PLATE STATUS — A CHECKED ZERO, NOT AN UNCHECKED WORK
Foot-of-page apparatus for cols 0609–0616: a CHECKED ZERO. Read by the orchestrator this session, not by the translator. Method: archive.org patrologiaecurs06saingoog, leaf n = (column + 9) / 2, verified against the running head. Leaves n309, n310, n311 and n312 read at full page width across the whole foot. No foot rule, no numbered notes — the columns simply end. n309 was read at the mid-page work division (Philippians ends, EPISTOLA AD COLOSSENSES begins below the rule, which is where this work opens) — the configuration that lost three notes on PL 202, and it is clean here. n312 was additionally read in a previous batch and agrees.
Tier 3, coverage full, notes_recovered: 0. Migne answered none of the cruces below. The single [cj:] of §3 and the single [sic:] of §4 therefore stand on our reading alone — they are not conjectures Migne had already settled on the page.
⚑ The two plate readings I flagged at 0611A were referred to the orchestrator and are now ADJUDICATED. He re-read n310 at full resolution; the result split two ways and both halves are recorded here because the method is the point.
- receperunt quae per Moysen — our twin is right and the crop was wrong. An earlier reading receperunt eum per Moy… came from a compressed strip that had overlapped two lines. ⚑ Same mechanism as the jubentur error of the previous batch: a plate reading taken from a crop that was not seen whole. No change to the English.
- concilio — the plate is right and our twin is wrong. See the section below.
⭐ The general lesson, and it is the reason this entry is long: a translator does not emend the committed twin from a partial crop, and does not adopt the twin against a plate reading seen whole. Flagging both and rendering the twin was correct in the interval; acting on either without the adjudication would have been wrong in one direction or the other.
§0a · A DIGITIZATION ALTERATION OF MIGNE'S SPELLING — 0611A, concilio → consilio
A class distinct from anything the marker set was built for, met twice in one session and named here because this book is where it first surfaced.
| reads | |
|---|---|
| Migne's plate, n310, full resolution | Alienati et inimici concilio ejus, quod idem valet |
| Our twin as Corpus Corporum delivered it (pre-patch) | Alienati et inimici consilio ejus, quod idem valet |
| ✅ Our twin now, patched | Alienati et inimici concilio ejus, quod idem valet |
Corpus Corporum silently altered Migne's spelling — and not by corrupting it, but by "correcting" what it took to be an error. ⚑ Cross-reference: Esther (8948) @0739D, the batch's other instance, where the plate prints convivum and our twin prints convivium. Two books, two instances, one session: a pattern, not a curiosity.
Why it is neither of the two things the apparatus already marks. [sic:] says Migne's type is defective; [ed:] says our digitization lost text Migne carries. This is a third thing: our digitization changed text Migne carries, leaving a well-formed word that reads perfectly. It is invisible from the Latin alone, and no marker in the English can carry it, because the fix belongs upstream in the TEI. ⭐ The orchestrator is applying a data/tei-patches entry restoring concilio after these files close (patches run before chunking, so it could not be applied while they were open). Migne's plate is untouched; the patch corrects our transcription back to what he set.
What I did about it, and it is a rendering decision, not a marker decision. The English now renders concilio, the word Migne printed:
Alienated and enemies to his council, which comes to the same thing, because they did not receive the things which God commanded through Moses, inasmuch as they were devoted to their own idols.
✅ SETTLED, and the columns agree. The plate reads concilio; the twin has been restored to concilio; the English renders concilium's own sense. The patch is recorded at data/tei-patches/8990.json — it corrects Corpus Corporum's transcription only, and Migne's plate is untouched.
No [sic: concilio], and here is the reasoning. Pattern 12 marks type that cannot be rendered in place; concilio renders in place perfectly — it is a real word (concilium, an assembly or council), it occupies a real grammatical slot (dative with inimici, which governs the dative), and the sentence it yields is coherent: they were alienated from and hostile to his assembly, and quod idem valet says that being alienated and being an enemy come to the same thing. I concluded the printed word is SOUND, not defective, so it is rendered plainly under Pattern 7 and the marker would be noise.
⚑ A second reason to decline, independent of the first: concilium/consilium is an attested medieval orthographic interchange, so even a reader who wants consilium here is not misled by the facing column — Pattern 9's exemption and the runbook's false-positive item 6 both point the same way. I record it as a rendering of concilium's own sense because that is what the plate prints and I have no warrant to read it as a spelling of the other word.
⭐ The general point, which I am the first in this book to hit and which belongs in the brief: a word that reads perfectly well may read well because someone upstream fixed it. That is undetectable from the Latin alone. The practical rule that follows is cheap: where a form strikes you as suspiciously tidy in a passage that is otherwise rough, say so and ask for a plate read. Both of this batch's instances surfaced exactly that way.
§1 · THE LEMMA DIVERGENCE LIST — ALL 126 SPANS CHECKED AGAINST THE VULGATE
Every italic span in the work — the verse lemmata, the mid-paragraph second lemmata, and the four scripture quotations drawn from OTHER books inside the glosses — was checked word by word against the Clementine Vulgate before the English was written, and each rendering then held fixed against every re-quotation of it in the gloss below. Agreements are listed too, so this can be told apart from a report by someone who only listed findings.
1a · The eight that diverge — all marked [var: …]
| col | Migne prints | Vulgate | marker + reasoning |
|---|---|---|---|
| 0610D | Sicut Pater habet vitam in se, sic et Filius qui suscitat et vivificat quos vult, | John 5:26 Sicut enim Pater habet vitam in semetipso, sic dedit et Filio habere vitam in semetipso | ⭐ The single most dangerous span in the work, and one of the four flagged in the launch as high-risk ground. A conflation: the first half is John 5:26 with in se for in semetipso and no enim; the second half is not 5:26 at all but a recast of John 5:21 (sic et Filius quos vult vivificat), with suscitat imported from 5:21's Pater suscitat mortuos. Rendered as printed: so also the Son, who raises up and quickens whom he will,. ⛔ The reach here is for the familiar "so hath he given to the Son also to have life in himself" — which would delete suscitat et vivificat, the two verbs the gloss's conclusion (sic ostenditur perfectus Deus esse) is built on. |
| 0611B | mihi data est in vobis, | Col. 1:25 quae data est mihi in vos | Accusative → ablative: in vos is direction ("toward you"), in vobis is location ("among you"). Material, and the gloss is written on the ablative — Migne introduces it with Vel as a second construal precisely because the case has changed, and closes tantum est enim quod a meis quaeritur, quantum si a meipso. Note the FIRST lemma of the same verse, Data est mihi., agrees with the Vulgate exactly; the divergence appears only in the re-quotation. |
| 0612A | secundum philosophiam et inanem, | Col. 2:8 per philosophiam, et inanem fallaciam | ⭐ per → secundum, and the whole gloss turns on it: ideo cum diceret secundum philosophiam et inanem, addit, secundum elementa mundi — the glossator's argument is the repetition of secundum, which does not exist in the Vulgate's text of this verse. Restoring per would destroy the argument in the sentence that makes it. Rendered according to philosophy and vain,. |
| 0612C | Decretum, | Col. 2:14 chirographum decreti | Genitive → nominative. Not scribal headword-normalization: this glossator preserves case elsewhere (Sortis. at 0610A stays genitive, Redemptionem. stays accusative), and here the gloss makes Decretum the subject of est — Decretum, est sententia in cunctis hominibus adversa. Rendered The decree, with the comma Migne prints. |
| 0612C | Morte moriemini, etc. | Gen. 2:17 morte morieris (singular) | ⭐ Migne prints the plural. The plural morte moriemini is a real Vulgate string — but it stands at Gen. 3:4, in the serpent's mouth and negated there (Nequaquam morte moriemini). So the decretum which the gloss calls sententia in cunctis hominibus adversa is quoted in the form the serpent used to deny it. Rendered You shall die the death, — plural, per Pattern 17, not "thou shalt". |
| 0614B | Qui enim scit voluntatem Domini, et non facit, plagis vapulabit multis [n: (Luc. XII)] | Luke 12:47 Ille autem servus qui cognovit voluntatem domini sui, et non praeparavit, et non fecit secundum voluntatem ejus, vapulabit multis | Four divergences at once: scit for cognovit, Domini for domini sui, the whole clause et non praeparavit dropped, and plagis added, which the Vulgate does not have anywhere in the verse. Rendered as printed. ⚑ The note [n: (Luc. XII)] sits inside the italic span in Migne; reproduced inside it. |
| 0614C | secundum imaginem Dei, | Col. 3:10 secundum imaginem ejus qui creavit illum | ⭐ The gloss cites this lemma explicitly as the reading of THIS book — vel secundum imaginem Dei, ut hic legitur, against secundum Deum "as is read in the epistle to the Ephesians" (Eph. 4:24). The whole comparison is between two printed readings, so conforming this one to the Vulgate would leave the glossator contrasting Ephesians with itself. |
| 0614D | Pax Dei. | Col. 3:15 pax Christi | The gloss distinguishes pax Dei from a peace that may be so called and quae non habet charitatem — a distinction that needs the divine name it has. Marked though the Old Latin and part of the Vulgate tradition read pax Dei here; Pattern 14's witness is named as Vulg. and the Clementine reads Christi. |
1b · Divergences examined and deliberately NOT marked, with the reason
Real differences from the Clementine text, each weighed against Pattern 14's bar ("differs … in a way that changes the sense") and declined. Recorded because a blind reader will meet them and must be able to tell a decision from an oversight.
- 0609C · VERS. 9 Impleamini in omni. — Col. 1:9 ut impleamini agnitione voluntatis ejus, in omni sapientia et intellectu spirituali. The lemma splices two runs three words apart. No word is altered; the gloss knows the full verse (plene scire quid Deus velit answers agnitione voluntatis ejus, in omnibus rebus activae vitae et contemplativae answers in omni sapientia et intellectu). Abridgment, not variance. Rendered That you may be filled in all.
- 0611D · VERS. 5 Sed et spiritu. — Col. 2:5 sed spiritu vobiscum sum. An added et; a connective, no content word, and nothing in the gloss touches it. Rendered But also in spirit.
- 0612D · Exspolians principatus et potestates. — Clementine expolians. Orthographic (exsp-/exp-), Pattern 9. No marker, no italic carry.
- 0613C · VERS. 22 Quae omnia ipso usu, etc. — Col. 2:22 quae sunt omnia in interitum ipso usu. Internal elision inside a lemma Migne himself closes with etc., and the gloss restores the elided phrase in its own words (ducunt in mortem). Abridgment, not variance.
- 0614A · VERS. 3 Vita vestra abscondita. — Col. 3:3 vita vestra est abscondita. Dropped copula in a fragmentary lemma. Nothing at stake.
- Orthographic throughout, all Pattern 9, none marked: charitate/caritate, quodcunque/quodcumque, caeteris/ceteris, coelum/caelum, Tichicum (8982's spelling, and Migne's here).
- Capitalization is not variance. Migne capitalizes In Spiritu. at 0609C where the Vulgate prints in spiritu, and Regnum Filii. where it prints in regnum filii. English capitalization of "Spirit" and "Son" carries no exponent for the difference and the gloss does not turn on it.
1c · The four scripture quotations from OTHER books, inside the glosses — the launch's named risk
These were checked first and hardest, because this is where the reach for a remembered English Bible happens. Three of the four agree with the Vulgate exactly and are recorded as agreements so the check is visible; the fourth is the conflation of §1a.
- 0610A–B · Regnum meum non est de hoc mundo; (John 18:36) — agrees verbatim, semicolon included. Rendered My kingdom is not of this world;
- 0610B · non est hinc, (John 18:36, the verse's own close) — agrees verbatim. Rendered it is not hence, ⚑ And the contrasting phrase beside it, Non est hic, is ROMAN in Migne, not italic — it is the glossator's counterfactual, not a quotation, and it is not in John. The English keeps it roman: he did not say: It is not here, but, it is not hence. Three negatives in that one clause (non ait · Non est hic · non est hinc), all present, each on its own verb.
- 0610B · De mundo non estis, sed ego elegi vos de mundo. (John 15:19) — agrees verbatim with the Vulgate's quia vero de mundo non estis, sed ego elegi vos de mundo, propterea odit vos mundus; the lemma simply opens after quia vero. Rendered You are not of the world, but I have chosen you out of the world. — plural estis, plural "you", Pattern 17.
- 0612B · Verbum caro factum est [n: (Joan. I)] (John 1:14) — agrees verbatim. The note sits inside the italic span and is reproduced inside it.
1d · Scripture the gloss uses WITHOUT italics — checked, and correctly unmarked
Roman-set scriptural matter is the glossator's prose, not a quotation, and Pattern 14 does not reach it. Named here so a blind reader can see they were noticed and not silently conformed:
- 0610A apud quem non est iniquitas (after 2 Chr. 19:7 / Rom. 9:14) — roman, rendered as prose.
- 0614B radix omnium malorum (1 Tim. 6:10) — roman.
- 0614A tempus terrena amantium semper est paratum — a re-cast of John 7:6 (tempus autem vestrum semper est paratum), with the possessive replaced by a genitive phrase. Roman, so no
[var:]; rendered as the glossator recast it, not as John. - 0614D vestis Domini inconsutilis desuper contexta (John 19:23) and eminentior via (1 Cor. 12:31, where the Vulgate has excellentiorem viam) — both roman, both rendered as printed.
§2 · THE UNEXPRESSED SUBJECT — every supplied pronoun, named
Fifteen places where English had to print a pronoun the Latin does not. Each is named with what decides it. Where two readings construe, both are set out — and the one adopted is stated, because a flagged passage is not a checked passage.
| col | Latin | supplied | what decides it |
|---|---|---|---|
| 0609C | coepit Archippum adjuvare | "he began to help Archippus" | The lemma above decides it, not the sentence. Pro vobis, opens Col. 1:7, whose subject is Epaphras (sicut didicistis ab Epaphra … qui est fidelis pro vobis minister). The rival reading is Paul, and it is not absurd — the ARGUMENTUM makes Archippus the one who preached, and Paul the one who writes. Adopted: Epaphras, because the verse the lemma opens supplies him and the glossator's Quorum salutem desiderans answers pro vobis in the same verse. English prints only "he", so the ambiguity survives intact either way; nothing was silently chosen in the text. |
| 0609D | Quod bene potest ille qui est claritas Patris | "he can well do who is the brightness of the Father" | Migne supplies the antecedent himself: id est Filius. No judgment needed. |
| 0609D | Vel secundum quod claram cognitionem vobis dedit | "he has given you clear knowledge" | The Son, carried from the immediately preceding id est Filius. |
| 0610B | hos per eum merito dimitteret | "he might justly let these go" | The devil. The clause is quia diabolus illos merito tenebat … hos … merito dimitteret: the same subject holds and releases, which is the whole symmetry of the sentence (merito … merito, reos … reum, conditione mortis … poena mortis). Reading it of God or Christ breaks the antithesis. |
| 0610D | sic ostenditur perfectus Deus esse | "he is shown to be perfect God" | The Son, subject of the John 5 quotation immediately above and of Quia omnia potest ex se immediately before. ostenditur … esse is nominative-with-infinitive, so the subject is the person, not "it is shown that". |
| 0610C | quia super fidem humanitatis ejus fundata est | "it is founded upon the faith of his humanity" | ⚑ The ejus case the launch names. ejus, not sua: the possessor is not the subject of fundata est. Subject = Ecclesia ("it"); possessor = Christ ("his"). Rendering sua here would have made the Church founded on the faith of its own humanity. |
| 0611A | quia non receperunt quae per Moysen mandavit Deus | "they did not receive" | The alienati et inimici of the lemma — i.e. the addressees in their former state. Migne shifts from the verse's second person (Et vos cum essetis aliquando alienati) to the third; the English keeps his third person and does not repair it to "you". |
| 0611A | in ipso auctore vivunt | "live in him their author" | Christ; Omnia per eum facta is the subject and ipso the agent of per eum. |
| 0612B | in templo manufacto … habitavit | "he dwelt" | God — Non quia corporeus sit Deus two clauses earlier. |
| 0612D | longe ab hoc regno ad coelum duxit | "he led them far from this kingdom" | Object supplied, not subject: the just just named (Abraham, Isaac et caeteros justos, the object of Auferendo in the previous sentence). The subject is Christ, from Traduxit. |
| 0613A | interior quae nos tenebat potestas ejus occisa est | "his inner power … was slain" | ⚑ A second ejus of the same class, and it decides the sentence. Subject of occisa est is potestas; possessor is the devil, named as subject of accepit in the same period. "Its own inner power was slain" would lose the devil entirely. |
| 0613B | amans ea quae non vidit … quarum rationem ignorat | "loving those things which he has not seen … he does not know" | The quis of Col. 2:18 (Nemo vos seducat, volens in humilitate), third singular throughout. |
| 0612D | Ut jam cesset | "That it may now cease" | illud of the lemma Affigens illud. — the chirographum/decretum. |
| 0615A | ne coarctati ab illis, delinquant in illos, et Deum offendant | "lest the sons, hemmed in by them, transgress against them" | ⭐ REPAIRED after the blind read (§8, repair 3) — this entry previously claimed the English required no choice, and that was wrong. The subject of delinquant is the sons: they are the ones coarctati ab illis (hemmed in by the fathers) and they transgress in illos (against the fathers). The antecedent sits in the lemma's verse, Col. 3:21, and in Modestos praecipit patres esse filiis two words earlier — not in the ne clause. ⛔ Why the first rendering failed: a bare "they" sitting between "by them" and "against them" resolves to the fathers on the default English parse, which makes the gloss incoherent as counsel (fathers told to be moderate lest the fathers transgress). Latin morphology leaves coarctati free to take the remoter noun; English word order does not. The English now names the sons, which is recovery of a printed antecedent, not a supplied one. |
| 0616A | Apertum est, quando multa et magna dicit | "It is open, when it speaks" | os (neuter), carried from Apertum est which is itself neuter and agrees with os. The rival reading is "when he speaks" (Paul); it is possible, and declined only because the neuter Apertum in the same clause fixes os as the running subject. Logged rather than settled. |
| 0616A | quod rediit ad me | "because he has returned to me" | Mark — Col. 4:10 Marcus consobrinus Barnabae, de quo accepistis mandata: si venerit ad vos, and the gloss's own et ideo si venerit quotes that clause. |
§3 · THE ONE [cj: …] — 0613A, Ego for Ergo
Migne prints Ego Scriptura quae tunc fuit exactrix operum significantium, nunc testis est rerum significatarum. Ego is a real, perfectly good Latin word, so Pattern 12 has nothing to wrap — but it has no grammatical slot: every verb in its clause is third singular (fuit, est), and the sentence next to it is third singular too (quae tunc observabatur … nunc recitatur). Read as Ergo the period construes without remainder and gives the gloss the connective its argument needs (it is drawing the conclusion from Tunc fuerunt hujusmodi praecepta … nunc testimonia).
Applied under Pattern 18 additively, Migne's word keeping its English:
I
[cj: Ego; read Ergo, "therefore"]the Scripture which then was the exactor of the works that signify, is now the witness of the things signified:
Why the bar is met, and the precedent. The nearest ruled case is 10703 @1126B — Oro ("I pray") for Pro, leaving a stranded ablative — ✅ on Pattern 18a's list. This is the same shape: a real word substituted for a short function word, leaving a syntactic stranding that no English can carry, and the false English ("I, Scripture, …") asserts a first-person speech act the glossator never makes. The verbatim-in-the-Latin-twin guard is satisfied by Ego.
⚠ Weighed against 18a's decline classes and it survives them: it is not inside a marker (nothing is defective enough for [sic:]), it is not inside a quoted scripture (that would be Pattern 14), and Migne glosses nothing here himself — §0 records the foot of these columns as a checked zero, so this conjecture is ours and is disclosed as ours. The nearest decline is 11064 @1154C (suscepit for suscepi, a person slip, declined); the difference is that 11154C's sentence has no main clause in either language, while this one has a complete main clause with a pronoun stranded outside it.
⛔ The [sic: Ego] alternative was considered and rejected: a [sic:] would take the connective out of the English altogether, which is exactly the 7a⁗ swallowing this edition has spent a corpus-wide sweep undoing. Pattern 18's additive form keeps both the printed word and the conjecture on the page.
§4 · THE ONE [sic: …] — 0614C, me morat for memorat (word division, Pattern 10)
Migne prints Hoc me morat ut se evasisse gaudeant et de caetero caveant. me morat will not construe: morat (late for moratur, "delays") with me leaves a first-person object governing a purpose clause about a third-person plural, and the clause has no other finite verb.
This is the split direction of Pattern 10's word-division class — the shape 7a⁗-b names as the most dangerous, because quarantining a split run removes a whole lexeme and this lexeme is the sentence's only predicate. me morat is memorat, the Glossa's commonest formula for "he mentions/recalls this, so that …".
In which you also walked. He mentions
[sic: me morat]this, that they may rejoice that they have escaped, and may beware for the rest.
Test 1 run on it, all four questions. Strip the marker and read: "He mentions this, that they may rejoice that they have escaped, and may beware for the rest." — (1) the clause still has its verb ("mentions", supplied in plain words beside the marker per 7a⁗, not left inside it); (2) no negation was in the run; (3) no interrogative or conjunction was in the run; (4) ⭐ the fourth question, the one 7a⁗ does not ask — the clause still has its object ("this"). Nothing is stranded on punctuation and nothing is doubled.
⚑ Rival readings weighed and rejected: monet ("he admonishes") gives the same sense but is not a word division and would be a bare emendation; movet ("this moves me") leaves ut … gaudeant unattached. memorat is preferred because it accounts for the printed spacing, which the other two do not.
§5 · Sections read clean, and cruces declined with reasons
Read clean and reported as clean (a checked zero is not a skip): the ARGUMENTUM; the whole of CAPUT IV; VERS. 3, 8, 10, 15, 22 of CAPUT PRIMUM; VERS. 1, 2, 4, 12, 19 of CAPUT II; VERS. 1, 2, 8, 16, 23 of CAPUT III. Verse addresses run in ascending order in all four chapters with no dislocation — 1, 3, 6–15, 18–25, 28/29 · 1, 2, 4, 5, 7–9, 12, 14, 15, 17–20, 22, 23 · 1–3, 5, 7, 8, 10, 14–17, 21, 23 · 3, 5, 8, 10, 16, 18. Nothing renumbered, nothing reordered; there is no Jude-type dislocation in this work, and its absence was checked, not assumed.
Declined, each with its reason:
- 0609B · Nota omnibus dignitate recte hic apostolum se nominat. The obvious conjecture is Notus omnibus dignitate — "known to all by his dignity" — which construes the whole sentence and answers the quia clause perfectly. Declined for a
[cj:]*: Nota as an imperative is the Glossa's single commonest opening formula (Nota quod …), so the printed word is not merely a real word but the expected word in this position, and the sentence it yields is elliptical rather than false — 18a's ❌ "a merely awkward or elliptical sentence". Rendered Note that to all in dignity he here rightly names himself an apostle, with "that" as the English complementizer Nota + indicative takes, and no punctuation supplied (Migne prints none between Nota and omnibus*). - 0611A · concilio ejus, quod idem valet. ⚑ Plate-confirmed as concilio; see §0a for the digitization-alteration class and for why no marker was taken. No
[var:]either: the italic lemma is Alienati et inimici alone and it agrees with the Vulgate exactly — the Vulgate's sensu is not in the lemma at all, and concilio ejus is roman, i.e. Migne's own typography already marks it as gloss rather than quotation. So Pattern 14 has nothing to fire on. The ejus has no antecedent in its own sentence and is rendered "his" without resolution. - 0612A · quasi admonens ne qualescunque adoratores simulacrorum, sed et quasi doctores … cavendos esse. The construction wants non solum where Migne prints ne, and the temptation is to render "not only any sort of worshippers". The "only" was not supplied. Rendered as if admonishing that not worshippers of images of whatever kind, but also, as it were, teachers and interpreters of the signs, are to be guarded against — the negative sits on adoratores, where Migne's ne sits, and the sentence is left as broken as he prints it. Declined for a
[cj:]because the correlative's answering member (sed et) IS printed and the shortfall is a mood particle, not a lost negation: nothing is asserted falsely, the sentence is merely un-Latin. - 0612C · non autem solum alia. Elliptical: alia has no verb and the non … solum has no answering sed. Rendered yet not that alone, others as well — "as well" is the English exponent of the asyndeton, and it is the only word in this work supplied to a Latin ellipsis. Flagged so a blind reader can weigh it. The negative sits on solum, and the following quia nisi hoc et alia tolleret confirms that both Adam's sin and the others are meant.
- 0613D · In honore Dei, cujus consilium, irritum dicitur. ⚑ Migne prints a comma between consilium and irritum dicitur — subject severed from its verb. Under Pattern 8 it is preserved, and the English reads whose counsel, is called void. This will look like our typo and it is not; the Latin twin carries it in the facing column. Not marked (a comma is not carried type), logged here because it is the one place in this work where a printed mark makes the English read wrongly.
- 0613D · the sunt / non sunt series. ⭐ The single passage most likely to be mis-rendered in this work, and it is not marked because it is not defective. Migne alternates et sunt: Non ad parcendum corpori (affirmative sunt, the negative living inside the quoted lemma) with et non sunt: In honore and et non sunt: Ad saturitatem carnis (negative sunt). All three preserved exactly as printed. Collapsing the first into a negative — which reads more smoothly, since the eye expects three parallel denials — would reverse the glossator's claim about non ad parcendum corpori, which is that the observances do answer that phrase of the verse.
- 0613D–0614A · Sagina enim carnalis est sensus traditio humana. est stands between carnalis and sensus, which is hyperbaton, not corruption: carnalis is the genitive agreeing with sensus. Rendered For the fattening of the carnal sense is human tradition. No marker.
- 0610C · usque ad. This work's only occurrence (qui ab Abel usque ad ultimum justum generantur) is ordinary prose, not the abridgment formula: it is roman in Migne, sits inside a gloss rather than a lemma, and elides nothing. Rendered as far as — the settled English — but roman, with no italics added, because adding italics Migne does not print would break span parity to honour a convention that does not apply here.
- 0616B · VERS. 18 Antequam salutationem ponat, praemittit exhortationem. In the Vulgate the salutation (Salutatio, mea manu Pauli) stands before Memores estote in the same verse, so the glossator's "before he sets down the salutation" reverses the printed order of Col. 4:18. His reading, reproduced; not corrected.
§6 · THE THREE TESTS FROM THE BRIEF, RUN ON THESE FILES
Test 1 — a [sic:] must not swallow the word it marks. Both markers stripped and the residue read aloud, plus a mechanical scan for stranded articles/prepositions sitting on punctuation and for doubled marks. Clean. The [sic:] site keeps its verb and its object (§4, where the fourth question is asked explicitly); the [cj:] site is additive by construction and keeps Migne's own word in the text. The one doubled-mark hit was etc., not of God at 0613C, which is Migne's own etc., non Dei.
Test 2 — a negation can be RELOCATED. All 90 negation-bearing tokens in the Latin (28 + 40 + 22) were listed, and for each the single Latin word it attaches to was named and then found in the English with the negative on it. Counts also compared band by band and match everywhere. The correlative sites — the 2 Thessalonians shape — were checked one at a time and all six hold: Non solum dico … sed ut (0611D), Non solum vobis gentibus … sed et nobis Judaeis (0612C), non autem solum alia (0612C), Non solum quae facitis … sed omnia (0615A), Non enim minus … quam (0611D-open), Non minus … quam (0609C). The scope traps checked individually: quae non servare peccatum erat (0613A — the negative is on the infinitive, not on "was sin"); qui timore non peccat (0615B — on peccat, not on timore, so "he who, through fear, does not sin"); ne solus nomen deitatis habeat (0614B — ⚑ corrected after the blind read, §8 repair 2: the negative is on solus, NOT on habeat — idolatry's aim is that God should not have the title alone, and the answering limb ut solus usurpet affirms of the covetous man exactly what ne solus denies of God); ne haec putentur non periculosa (0614C) and ne videatur salus non esse promissa (0611B) — two true double negatives, both carried double, where dropping one yields the smoother and false "lest these be thought harmless" / "lest salvation seem to have been promised only to the Jews". Not one negative was pressed into double duty.
Test 3 — punctuation compared BY COLUMN BAND, never by total. The 28 verified anchors partition both twins identically; colons, semicolons, question marks, exclamation marks, guillemets, quotation marks and parentheses were counted in each of the 31 bands. The first pass returned 19 mismatched bands and every one was a real site — the totals would have hidden most of them, since supplied and destroyed marks were cancelling across bands exactly as the brief warns. All 19 fixed: 14 semicolons and 2 colons supplied where Migne prints a comma or nothing (0609A, 0609B, 0610C, 0611D, 0612A, 0612B, 0612C, 0613A, 0613B, 0613C, 0613D×2, 0614-open, 0614C, 0614D, 0615A), and 3 colons destroyed where he prints one (0609D confugerunt:, 0610A a diabolo:, 0613A exstinxit: — the last needing the English clause reordered so the colon fell on the correct side of the [0613A] anchor, as it does in the Latin). Final state: zero mismatched bands, all marks, all chunks.
⚠ Per the brief, no comma→semicolon sweep was attempted: this book prints semicolons freely (five in the Latin), so Tobit's narrow exception does not apply and was not invoked.
Also verified after writing: 28 column anchors 1:1 and in order; 126 italic spans 1:1 and in place; 2 [n:] in the same positions with contents untranslated (both are citations — (Joan. I), (Luc. XII) — so neither takes Pattern 15's [nt:]); 2 question marks, both Migne's (quomodo caput eorum? at 0610C, a direct question needing no recasting; quare legem servatis? at 0613C, rendered as the direct question Migne's dicite introduces — but do you say to others, Why do you keep the law? — so the mark rides with a comma rather than Pattern 8b's colon, which was not needed here and would have been a supplied mark); 5 ## heads; zero em-dashes on either side.
§7 · What a blind reader should look at hardest
- 0610D, the John 5 conflation. The
[var:]is long and the rendering is deliberately un-familiar. Check suscitat et vivificat survives in the English, since the remembered verse has neither verb. - 0613D, the sunt / non sunt series. Three parallel clauses of which only the first is affirmative. If one of the three has drifted, it is here.
- 0613A, the
[cj:]on Ego. The most arguable judgment in the work; §3 sets out the precedent it rests on and the two decline classes it had to survive. - 0609C, coepit Archippum adjuvare. The one unexpressed subject where the rival reading (Paul rather than Epaphras) is genuinely live and the English cannot show which was meant.
- 0612B, quia ille caput in quo omnes sensus, insistit; quasi solus tactus. The hardest sentence in the work to construe, and repaired after the blind read (§8, repair 1). insistit governs omnes sensus, and the verbless second limb gaps it. No verb is supplied anywhere in these three chunks — the English gaps in the second limb exactly as the Latin does.
- 0611A, concilio ejus — the digitization-alteration site of §0a, now settled: plate, twin and English all read concilio / "council", patch at
data/tei-patches/8990.json. Worth a reader's eye only as the class, not as an open question — a word that reads perfectly may read perfectly because someone upstream tidied it.
§8 · THE BLIND POLARITY READ — 4 sites, 3 repaired, 1 declined
A blind reader that never saw this cruces file read all three chunks. It returned 4 sites and independently decided all five of the sites flagged above. Three repairs were made; one finding was declined on Pattern 18 grounds. All repairs are folded into the sections above and into the running text; this section records what changed and why, so a later reader is not left comparing versions.
⭐ What the read confirms about the method, and it is the most reusable line here. The reader named five separate opportunities to conform a quotation to a familiar English Bible — John 5:26, Col. 2:8, Gen. 2:17, Luke 12:47, Col. 3:15 — and found all five logged rather than silently corrected (§1a). It then said that is precisely why it looked hardest where no marker appears at all. That is the 7a″ apparatus working as designed: a divergence list does not merely record findings, it tells the next reader where the translator's attention has already been, and so redirects the hunt to the gloss prose between the lemmata. All four of its sites are in gloss prose. None is in a lemma.
Repair 1 — 0612B, insistit mis-parsed, and a number defect smoothed away
ille caput in quo omnes sensus, insistit; quasi solus tactus, quibus datus est spiritus ad mensuram.
| was | "because he, the head in which are all the senses, stands firm; as it were touch alone is in those to whom…" |
| now | "because he is the head in which all the senses, reside; as it were touch alone, in those to whom…" |
The first rendering gave insistit to ille, supplied an unmarked "are" for omnes sensus, and then supplied a second, independent verb for the verbless second limb. insistit belongs to omnes sensus: the Augustinian figure is a two-term comparison — the head has all the senses, the members touch alone — and quasi solus tactus is verbless precisely because it gaps insistit. Read the old way, "stands firm" had no work to do and no counterpart in the second limb, which then had to be given a verb out of nothing. The English now gaps in the second limb exactly as the Latin does, so the count of supplied verbs in this work is zero (only the copula of ille caput, which Latin gaps routinely).
⚠ Why the wrong reading was tempting, and it is the tell: insistit is singular against plural omnes sensus. Re-assigning the subject made the English read perfectly and left no trace that anything in the Latin was odd — a sentence that reads too well where the plate is rough. That is the same signature as §0a's concilio, arriving from the opposite direction: there an upstream hand had tidied the Latin; here the translator's own parse tidied the English.
Marker: DECLINED, and the reasoning, since the orchestrator left the call to me.
- Not
[sic:]. Pattern 12 marks type that cannot render in place; insistit has a real slot and a real sense, and only its number is at issue. Pattern 12's second failure test — "it renders only by HIDING the defect", the 11632 @0304A case — was the live argument for marking, since any grammatical English must pick a number and thereby conceal the mismatch. - ⭐ It was declined because the singular is EXPLICABLE, not broken. Under the gapped reading the one verb serves two subjects: plural omnes sensus in the first limb and singular solus tactus in the second. insistit stands between them and agrees with the nearer — ordinary Latin nearest-subject agreement, and it is the very structure the corrected parse posits. Marking it would assert that Migne's type is defective about a form I judge sound, which is the same conclusion, on the same grounds, that declined
[sic: concilio]at §0a. Consistency between those two calls is deliberate. - The mismatch is not thereby concealed: the singular form stands in the facing Latin column, and this entry states it.
Repair 2 — 0614B, the negative floated off solus
ut ne solus nomen deitatis habeat.
| was | "that he alone may not have the name of deity" |
| now | "lest he alone should have the name of deity" |
⭐ The 2 Thessalonians / Esther class exactly: nothing added, nothing dropped, every count matching, and the false reading grammatical. The ne attaches to solus, not to habeat. Idolatry's aim is not that God should fail to have the name of deity — it is that he should not have it alone, that others should share the title. The proof is inside the sentence: the gloss is a sicut … ita parallel pivoting on solus, affirmed of the covetous man (ut solus usurpet quae Deus omnibus fecit) and denied of God. The old word order also read most naturally as "he, and he only, may not have it" — a third sense, and the wrong one.
⚑ This site had been checked under test 2 and passed. The negative was present, on a plausible host, and the band count matched. Naming the host word is not sufficient if the wrong host is named — the check must be run against the sentence's own argument, which here is the sicut … ita pivot. Recorded because test 2's own statement ("name the single Latin word it attaches to") assumes the naming is right.
Repair 3 — 0615A, wrong antecedent
Folded into §2's table above, which previously claimed the English required no choice and was wrong. "lest the sons, hemmed in by them, transgress against them." A bare "they" between "by them" and "against them" resolves to the fathers on the default English parse; Latin morphology permits the remoter noun and English word order does not.
Repair 4 — 0611D, same mechanism as repair 2
quia non omnibus patent, sed his qui petunt et pulsant.
| was | "because they do not lie open to all, but to those who ask and knock" |
| now | "because they lie open not to all, but to those who ask and knock" |
The non attaches to omnibus, and the sed his contrast proves it. The sense survived the old rendering because the but-clause recovered it, which is why the reader rated it minor — but it is repair 2's class, and the two together say something worth keeping: this work's characteristic polarity risk is not a dropped negative but a negative resting on the verb when it belongs on a quantifier or a correlative. Both of this book's polarity repairs are that shape.
DECLINED — the [cj:] on Ego at 0613A stands unchanged
The reader argued the running text should carry the accepted conjecture ("Therefore") with Ego moved into the bracket, on the ground that stripping the marker leaves "I the Scripture … is", which is not English. Pattern 18 rules the other way, explicitly, and the treatment is not changed:
"The rule: the gloss is ADDITIVE. Migne's word keeps its English, and our conjecture stands beside it. Nothing is removed, nothing is replaced, and the reader sees both."
Its worked example is the angels announce peace purely [cj: munde; read mundo, "to the world"] — printed word in the running text, conjecture in the bracket. Putting Ergo in the text would replace Migne's word with our conjecture, which is the one thing the pattern forbids; and verify-english's substring check on the bracketed word exists precisely to make that substitution impossible. The un-English residue is not a fault in the rendering — it is the defect, visible.
✅ The conjecture Ergo is confirmed correct by the reader's own independent reasoning: the verb is est, third person; the Tunc … nunc contrast supplies the connective the sentence needs; and there is no first-person prosopopoeia of Scripture anywhere else in the commentary. That third argument is new and is worth recording — it is a survey of the whole work, which §3 did not have.
⭐ THE [sic:] / [cj:] BOUNDARY — why this work carries one of each
The reader called 0614C's [sic: me morat] "the better model" and read the two markers as two conventions for one operation. They are not. They are two conventions because they are two situations, and the distinction is stated here because this is the second time in two batches that the boundary has been queried and the answer should be findable in the work itself:
| 0614C, me morat | 0613A, Ego | |
|---|---|---|
| what is on the plate | broken type — a word-division defect; me morat is not a Latin word | a real Latin word, correctly formed |
| can it render in place? | No — there is nothing to render | Yes — "I" is its English, exactly |
| governing pattern | Pattern 10 (carry the whole run) + Pattern 12 ([sic:]) | Pattern 18 ([cj:]) |
| what the convention requires | 7a⁗: the sense in plain English beside the marker, because the quarantine took the clause's only verb | additive gloss: Migne's word keeps its English, conjecture beside it |
| strip the marker and you get | "He mentions this, that they may rejoice…" — complete | "I the Scripture … is" — broken, and meant to be |
The test that separates them is not how wrong the Latin looks; it is whether what the reader sees is our translation or his type. Where the type is broken, the marker holds the type and the English must carry the sense beside it. Where the word is real, the English must carry the word, and the conjecture goes in the marker. Reversing either one hides something: a [sic:] on a real word deletes a word Migne printed, and a [cj:] that puts the conjecture in the text deletes it just as surely while looking tidier.
Confirmed clean by the blind read, and not to be churned
0610D (suscitat et vivificat survives; the conflation logged, not conformed) · 0613D's et sunt: / et non sunt: series — the trap was avoided, one affirmative followed by two denials reproduced exactly where the eye wants three parallel denials · 0614C's [sic:] passing the swallow test · both true double negatives, at 0611B and 0614C, with correct scope · the John 18:36 sequence at 0610A–B, both negatives placed and the Non est hic / non est hinc distinction preserved word for word · 0613D's comma between subject and verb (Migne's, carried deliberately — his marks ride as printed; raised by the reader only as typographic and settled house practice) · 0609C, where the reader independently confirms the unexpressed subject is Epaphras (from Col. 1:7, the verse the lemma Pro vobis, opens; Archippus is accusative and is therefore the one helped, not the helper) and finds no defect, since the Latin leaves the subject unexpressed and the English leaves it unexpressed identically. Nothing was decided wrongly there because nothing was decided.